Recent News

April 2014  
Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30      

Last changed: Jun 16, 2009 12:51 by Justin Deoliveira

0) What's up
1) Road map

dwins: aaime, jdeolive: meeting time?
[10:02am] jdeolive: yup
[10:05am] jdeolive: agenda items?
[10:05am] You changed the topic to "0) what's up 1) road map".
[10:06am] dwins: i'd like to propose a community module, but maybe that should go on the mailing list
[10:06am] jdeolive: dwins: as you see fit, it might be nice to do it on the mailing list so everybody sees teh approval
[10:07am] dwins: okay. email seems better so I'll do that
[10:08am] jdeolive: ok... no more agenda items lets start
[10:08am] jdeolive: 0) what's up
[10:08am] jdeolive has been fixing 2.0 bugs
[10:09am] dwins has been off in js land, but spent some time helping agerber wrap up some spring issues yesterday
[10:10am] aaime is working on 3d kml and wicket ui
[10:10am] groldan is doing arcsde improvements: JNDI, multiversion support and non-spatial tables support
[10:10am] groldan is doing arcsde improvements: JNDI, multiversion and non-spatial tables support
[10:10am] agerber is about to commit printing
[10:11am] jdeolive: cool
[10:11am] jdeolive: 1) road map
[10:11am] jdeolive:
[10:11am] sigq: Title: Roadmap - GeoServer (at
[10:12am] jdeolive: we are working through the blocker and priority fixes for 2.0-beta2
[10:12am] jdeolive: i have fixed for GEOS-3076, and GEOS-3118 pending so those should be gone soon as well
[10:12am] jdeolive: GEOS-2705 can probably be pushed back
[10:13am] jdeolive: since it is more of a meta issue
[10:13am] mcrmcr: What about migrating some plugins/extensions to 2.0
[10:13am] jdeolive: mcrmcr: it has been on the plate for a while now... but i pushed it back to 2.0-RC1
[10:13am] jdeolive: that said
[10:13am] jdeolive: folks are free to do it now
[10:14am] jdeolive: and if people think it is important we have all the extensions for beta2, we can promote it as well
[10:14am] jdeolive: thoughts?
[10:14am] mcrmcr: hmm, every time I tried to include an extension into 1.7.x I did a minor mistake
[10:15am] mcrmcr: I think, for 2.0, the situation will be the same
[10:15am] jdeolive: mcrmcr: third time will be charm, i promise
[10:16am] FerMendez joined the chat room.
[10:16am] jdeolive: mcrmcr: which of yorus need to come forward to trunk still, featuregeneralized, and imagemosaic-jdbc?
[10:16am] mcrmcr: and db2-ng would make me happy
[10:18am] jdeolive: i thought the db2 extension on trunk was db2-ng... maybe i am wrong
[10:18am] jdeolive: yup.. i am wrong
[10:19am] mcrmcr: I want to get rid of the old db2 module, nobody supports it
[10:19am] jdeolive: but i think we more or less agreed that the old jdbc extensions will go bye bye, and we will ship only the ng ones
[10:19am] aaime: yup
[10:19am] mcrmcr: code sprint ?
[10:20am] ahocevar left the chat room. (Nick collision from services.)
[10:21am] jdeolive: mcrmcr: for porting all the jdbc-ng extensions? or to fix jdbc-ng issues?
[10:21am] jdeolive: or both?
[10:21am] ahocevar joined the chat room.
[10:21am] mcrmcr: first fix the issues to get in supported land and if we have a good run, then both
[10:22am] jdeolive: have we isolated the issues that are preventing us to moving jdbc-ng to supported?
[10:23am] mcrmcr: I think, jody had some arguments (aggregate functions)
[10:23am] aaime: aggreation and jgarnett was very pro fixing the code duplication before
[10:23am] jdeolive: the only objections i heard last time around from jody was events and aggregation
[10:23am] aaime: turning that into a supported module
[10:24am] jdeolive: events are there, aggregation is ready to go given that people agree on the compromise i proposed, which i believe matches the functionality of the old datastores
[10:24am] jdeolive: as for code duplication, i have yet to see an acceptable patch for an alternative, i stated some quite serious issues with jody's approach which he has not responded on
[10:25am] aaime: his position seems to be that despites where the patch comes from, that should be fixed before going into supported land
[10:25am] aaime: if I'm interpreting it right, it's a bit of a hard position, yet it's going to be a bigg-ish change
[10:25am] jdeolive: how do others feel... i feel that is an unreasonable blocking issue
[10:26am] aaime: so it would be nice to make that before turning the module in supported land
[10:26am] aaime: (but I see that as a very good move, not as a blocker)
[10:26am] jdeolive: agreed, i see it as an improvements (nice to have)
[10:27am] jdeolive: but if you look at the datstores that are currently "supported" there is duplication everywhere, far more than the jdbc-ng ones imho
[10:27am] mcrmcr: yeah, lets finish jdbc-ng, otherwise jdbc-ng will finish us
[10:27am] aaime: jdeolive, one thing I did not check is how the modules fares against the 5 stars for graduation
[10:27am] jdeolive: haha, well put
[10:27am] groldan: as long as there's the will to make it, and provided the unit tests are there to avoid regressions, I don't think it should be blocker
[10:27am] mcrmcr: nice
[10:29am] jdeolive: groldan: well... even that is in question
[10:29am] jdeolive: on the issue in question, it involves adding another level of indirection
[10:29am] jdeolive: which i think was one of the major flaws of the old datastore
[10:30am] jdeolive: having to drill down 4-5 classes to actually get to the code that creates the SQL
[10:30am] jdeolive: anyways... perhaps we are starting to get off topic
[10:32am] jeremy-wrk joined the chat room.
[10:33am] tenzochris is now known as tenzo|away.
[10:33am] jeremy-wrk is now known as Pe.
[10:35am] jdeolive: so... getting back to the road map
[10:35am] jdeolive: any proposed changed?
[10:35am] jdeolive: changes
[10:36am] cabraham left the chat room.
[10:37am] aaime: none in particular
[10:37am] aaime: one thin
[10:37am] aaime: thing
[10:37am] aaime: next week during the Bolsena sprint I'll try to fix as many beta2 issues as I can
[10:37am] aaime: any suggestions/preferences on which I should concentrate myself?
[10:38am] cabraham joined the chat room.
[10:39am] jdeolive: well... i think are focusing on bugs rather than UI improvements?
[10:39am] jdeolive: so i guess any bugs that represent the biggest or most serious regressions

Posted at 16 Jun @ 12:48 PM by Justin Deoliveira | 0 comments
Last changed: Jun 09, 2009 15:40 by Andrea Aime

1) what's up
2) release update
3) road map
4) it's a 3d world

bmmpxf: 1) what's up
jdeolive: whoops
***aaime is doing the 1.7.5 release and is setting up for hacking (very) partial 3d support in gt2 and gs
***groldan is on arcsde related paid work for the whole week
***jdeolive is working on resource pub stuff and getting ready to close to 2.0 related bugs
***bmmpxf is going through old doc tasks, and doing misc doc curation things
mcrmcr: mcrmcr resumes work on geoxacml
tschaub1 [] è entrato nel canale.
tschaub ha abbandonato il canale (quit: Read error: 110 (Connection timed out)).
jdeolive: next topic?
aaime: yup
ahocevar2 [] è entrato nel canale.
jdeolive: 2) release update
jdeolive: aaime?
tschaub1 è ora conosciuto come tschaub
aaime: I'm finishing both the gt2 and gs releases
aaime: gs artifacts are mostly fully uploaded
aaime: but I miss someone to make the osx installer
aaime: if you can give a whirl at the artifacts it would be great
aaime: I'm still fighting with gt2
aaime: the souce build won't build with an empty repo because some people did not do what I asked last month
bmmpxf: aaime: Don't forget that there is a new Windows installer that should be posted alone side the old for testing.
aaime: ???
aaime: this is completely new to me
***aaime dies at the thought of uploading another 40MB
aaime: did we have two installers in the previous releases?
aaime: bmmpxf, jdeolive?
bmmpxf: the winv2 was uploaded last month, after the 1.7.4 release. We discussed on the ML to have both up there so people could test, since it had Vista probs.
ahocevar [] è entrato nel canale.
aaime: are there instructions on how to build it?
jdeolive: bmmpxf: do you want to handle that?
jdeolive: i believe you did last time no?
***aaime would like that
bmmpxf: I will handle it. Eek, there are no instructions, although it is notionally identical to the old windows installer.
bmmpxf: I can build both v1 and v2 if you'd like.
ahocevar2 ha abbandonato il canale (quit: Read error: 104 (Connection reset by peer)).
aaime: already built v1 and uploaded it
bmmpxf: And I can call the new one ng if everyone one would like.
aaime: sure
aaime: someone you should also answer the user dilemma "why would I want to download this one"
aaime: (like with oracle ng)
bmmpxf: definitely
bmmpxf: thanks for that, btw
bmmpxf: if you have anyt roubles with the PDF doc ouptu too, let me know.
aaime: that's not part of the standard release guide either...
aaime: (I was wondering about it)
bmmpxf: ah crud
bmmpxf: okay, filing a JIRA for me now
aaime: the release guide instructs only to build the html modules
aaime: sorry, the html output
bmmpxf: sorry, aaime
aaime: should we add the pdf version too?
bmmpxf: yes
bmmpxf: oh wait
bmmpxf: only for download, not for live posting on docs
aaime: right, for downloads
mcrmcr: pdf would not be bad, had read some mails from my university about pdf documentation for geoserver/geotools
bmmpxf: we started pdf output as of 1.7.4
aaime: for the record, the pdf generation fails on my machine
awright ha abbandonato il canale (quit: Read error: 104 (Connection reset by peer)).
bmmpxf: error?
aaime: it complains about a missing font
bmmpxf: hmmm
bmmpxf: let's talk after the meetings
bmmpxf: I'll try it too.
awright [] è entrato nel canale.
aaime: Ok. I guess that is it?
jdeolive: 3) road map
sigq: Title: Roadmap - GeoServer (at
jdeolive: soo... i guess first thing is date for beta2?
jdeolive: now it is set to the 22nd?
aaime: Who's going to make the release?
bmmpxf: Q about 2.0. Is the UI stable enough to start creating screenshots about it?
bmmpxf: or is it still subject to lots of change?
aaime: I will be in Bolsena for the european OSGEO sprint, so my internet connectivity might not be good
jdeolive: aaime: i will also be in australia at the complex features workshop
jdeolive: so will groldan
jdeolive: i suggest we push it back unless anyone else steps up
aaime: ok, so for that date we seem short of releasers
jdeolive: letting sit a little longer allows to fix more issues anyhow
aaime: either that, or make it mid of the week before
chippy ha abbandonato il canale (quit: Remote closed the connection).
jdeolive: we could... what are preferences of the people here?
jdeolive: actually before we answer that
sophiap: what would not done if pushed to the week before?
jdeolive: i have another related question
jdeolive: sophiap: probably just fewer lower priority bugs fixed
jdeolive: related question is
sophiap: ok
jdeolive: do we need a beta3?
aaime: hmmm... probably not? It all depends on the answer to bmmpxf question
aaime: how many other big UI changes are planned?
aaime: or any other big change in general, for that matter
jdeolive: good question
jdeolive: there is stuff i want to make nicer... but practicality tells me it is probably good enough now since it is such a huge step forward from the 1.7.x UI
bmmpxf: I feel like I want the UI to be at 96% stable before I start thinking about 2.0 docs. Program stability is another matter.
ahocevar ha abbandonato il canale (quit: Read error: 60 (Operation timed out)).
jdeolive: aaime, groldan: what are your thoughts?
ahocevar1 ha abbandonato il canale (quit: Success).
aaime: hmmm... not sure
aaime: I think the UI is good enough
aaime: and now we're at liberty to improve it more, and more easily
aaime: (ok, bmmpxf won't like that very much )
jdeolive: agreed
groldan: not sure neither, there might be a couple things to review in order to tell its ok to start screenshotting the ui
bmmpxf: ha!
jdeolive: i think we can continue to incrementally improve it
aaime: and incrementally improve the screenshots as well
groldan: the first I can think of is jdeolive wanted the order of table columns to be inverted
groldan: but all in all, it's looking quite solid
groldan: so I think I agree with aiime, it's also something that can be incermentally improved
aaime: anything else?
aaime: 1.7.6 wise... it's quite doubtful there will be such a release, right?
aaime: jdeolive?
jdeolive_ [] è entrato nel canale.
bmmpxf: Well, we didn't plan a 1.6.5, but that came out why? (I forget)
aaime: I don't remember exactly
aaime: jdeolive_, we were talking about 1.7.6
aaime: and the fact that it's not really planned
jdeolive_: aaime: i think 1.7.6 is probably likely, especially if more paid work comes in
aaime: right, but it is on the hook for sponsored work that includes making the release in the lot
bmmpxf: actually, I think paid work was what created 1.6.5
jdeolive_: true
aaime: the thing is, it's not like we'll do it "no matter what"
jdeolive_: but I still think we can plan it accordingly with the community
jdeolive_: and hence throw it on the road map... maybe just not attachning a release date just yet
aaime: unless someone pops up and actually wants to do it, that is
sophiap: what would be the benefit of 1.7.6?
aaime: the same as 1.7.5
jdeolive_: sophiap: well 1.7.x is the only branch suitable to stable paid work at the moment
aaime: and other previous releases I guess
jdeolive_: and given there are a number of people doing stable work right now... it is probably the only option for that imo
aaime: trunk is nowhere near stable enough
sophiap: how stable does the trunk have to be to release 2.0?
aaime: if experience teaches something, stable, but not that much
jdeolive_: i guess no blockers in jira
aaime: we usually release when we run out of issues we can find on our own and from the users that actually do test RC
aaime: but usually x.y.1 and x.y.2 are still a bit shaky
sophiap: ok
jdeolive_: agreed, we usually don't get stable for a couple of release cycles
aaime: (as in, really stable, really solid)
aaime: it's not like we release stuff we don't believe in, it's just when the user community actually hits the release
sophiap: so users don't migrate until after .1 or .3 releases?
aaime: (usually, only early birds use RC)
sophiap: .2
aaime: most people start really testing gs with x.y.0 and that results in new issues we did not know popping up
aaime: and for cautious people, well, we've seen people upgrading from 1.6.x straight to 1.7.4
bmmpxf: sophiap: On the mailing list there seems to be a solid continuum of people, from the crazies using nightlies to those who wait until .3 .
mcrmcr ha abbandonato il canale (quit: " ajax IRC Client").
aaime: right
aaime: and well.. geoserver has such a huge testing surface that it needs time for the user community to hit it all
aaime: we have a ton of tests, but not nearly as big as what the user community testing can actually do
aaime: if you think about just the CITE tests I run today, almost 1000 tests did hit GeoServer
aaime: and we pass every single one of them every time
aaime: plus we have hundreds of tests in gs and thousands in gt2
aaime: yet, the real world is a different beast
jdeolive_: the biggest hole in our testing is with data directories
jdeolive_: starting up with old data dirs, various aspects of peoples dd's in the wild, etc...
jdeolive_: it is too bad funky_c's call for people to supply their dd's did not get more uptake
ahocevar [] è entrato nel canale.
bmmpxf: yeah, that is a drag
aaime: yeah
aaime: I guess the issue is that most people refer to local resources that cannot be shared
aaime: databases, network diskss
aaime: it's the usual thing in all OS I guess, we do our best to deliver, but in the end we don't have a QA department
aaime: and GS usually sees a lot of changes, big ones, every time we release a x.y.0 release
jdeolive_: on a related note one thing i want to do for this years GeoServer past present future talk is graph our major releases against new features + bug fixes against those of other projects, i think it would be an interesting comparison
jdeolive ha abbandonato il canale (quit: Read error: 60 (Operation timed out)).
aaime: yeah, I guess compared to other projects
aaime: our number of new features and bug fixes goes out of the roof
bmmpxf: yeah, no kidding
bmmpxf: We're like the anti-QuarkXPress
iwillig [] è entrato nel canale.
aaime: ha ha
aaime: ok
aaime: let's move to the next topic?
aaime: jdeolive, anything else for the road map?
jdeolive_: well
jdeolive_: i don't think we came to a consensus on if we need a beta3, and when we should do beta2
groldan: right
aaime: imho we don't need a beta3, as for beta2, I'm open to doing the release end of the week before the 22nd
groldan: I would personally like to postpone beta2 for a wheek, and go for rc1 after that
groldan: week, I mean
jdeolive_: ok. i am for RC after beta2 as well, as for when beta2
groldan: rationale being there's a couple things I'd like to get fixed for beta2 that I don't think I will really have time for before the australia workshop
jdeolive_: if someone wants to do it that week (early) i am ok with that
aaime: jdeolive_, which week?
aaime: date please
groldan: we should make first sure there's no features missing after beta2
jdeolive_: the week before the 22nd
groldan: it'd be nice to run for rc1 as bugfixing only
aaime: early in the week before the 22nd? .... too early?
groldan: did anybody make a run through that feature-parity bunch of issues?
aaime: groldan, nope
jdeolive_: sorry, end of week before the 22nd, what you said
groldan: how close are we to feature parity?
aaime: no idea
aaime: I think we're pretty close, thougth we lack many of the demos
sigq: Title: GEOS-2134 Bring the new UI to feature parity with the old UI - (at
groldan: says we're there?
jdeolive_ è ora conosciuto come jdeolive
aaime: I guess it does not take into account the demos
aaime: rss, wfs-t
aaime: and so on
groldan: but I still doubt it
groldan: right
aaime: I know bmmpxf also wanted to change the demo data?
jdeolive: i sort of think those should be seperate downloads imo... but that is aside
groldan: layer group editor?
aaime: don't know
aaime: I think I made some fixes there, but I don't remember what its status is
aaime: Oh, one thing I remember that is missing: it's not possible to choose a different style for a layer in the group
sigq: Title: GEOS-2705 Wicket UI styling and usability improvements - (at
aaime: right.. there's quite a bit of stuff there... for some there are some hard problems to be solved
groldan: how much of that do we need for a beta2
aaime: pretty much everything if we want a pure bug fix RC
groldan: if there's something that's a feature rather than a cosmetic issue that'd be priority
aaime: if you look at some issues they ask for new panels in the UI
aaime: with that in mind, and given some of us are working on unrelated stuff
jdeolive: imo things under that issue can be done in future releases
aaime: all of them?
jdeolive: maybe not all... not sure have to look through again
aaime: like, you seemed to be pretty interested in having the preview in the layer editor
chippy [] è entrato nel canale.
jdeolive: but that issue was supposed to be more or less container for all not crucial stuff
jdeolive: yeah... but again, i would not call it a blocker for 2.0
aaime: ok, so there is a need to re-prioritize issues for beta2
jdeolive: agreed
aaime: I'm a bit worried about it thought
jdeolive: that is one of my tasks for this week
aaime: rationale: many of us are on paid work this week, unrelated to beta2
aaime: and the week of the 22 we are away
aaime: so that leaves only one week for making the changes for beta2
aaime: and moving it to the 29 would not change things a bit
jdeolive: yeah... so should we push for the week after the 22nd?
aaime: right, but at the end of that week
aaime: because if we shoot for the beginning the time to do fixes is still just one week
aaime: unless I spend all of the Bolsena sprint working on the wicket UI (deja vu)
jdeolive: haha, ok... well i am open to pushing back
jdeolive: and it sounds like groldan is as well
groldan: that sounds good to me
aaime: jdeolive, do you have time to go through the beta2 issues and try to re-prioritize them
groldan: pity, cause I'd like it sooner, but more close to reality imho
jdeolive: aaime: yup, it is on my task list for this week
aaime: this way we can have an idea of what should actually be done
aaime: nice
jdeolive: will try to get to it today
aaime: I guess we can talk about the effective date after we have that
aaime: and for the record, I'm not against working on beta2 during the Bolsena sprint
jdeolive: sounds good
groldan: it's 15 mins past meeting time, I'm going back to hack mode, ping me if needed
aaime: I just want to keep some time to exploit networking opportunities that might arise there
aaime: (with geonetwork or with other projects)
jdeolive: aaime: yeah. i think you should focus on that, it would be a pity if you had to hack the whole time
aaime: indeed
aaime: ok, meeting done?
jdeolive: ok... so i will reorganize jira, and send the road map update
jdeolive: 4) its a 3d world?
aaime: oh right
aaime: I just wanted to give a heads up
aaime: about the 2d+1 work I'm about to start
aaime: it will end up giving us 3D kml output from postgis data sources
aaime: but it won't cover other datastores or gml for the moment
aaime: thought I hope to add that later
aaime: for curious people, see the proposal on the geotools mailing list
aaime: (most of the work wil be done there, the kml producer should not need deep changes)
aaime: that's it

Posted at 09 Jun @ 3:36 PM by Andrea Aime | 0 comments
Last changed: Jun 02, 2009 12:56 by Justin Deoliveira

0) What's up
1) Roadmap
2) MVN site generation
3) Chart extension

jdeolive: 0) what's up
[10:15am] lasma is away: Gone away for now
[10:15am] aaime went to the zoo with family and he's not relaxing by fixing a few bugs
[10:15am] jdeolive fixing some restconfig bugs
[10:16am] jdeolive thinks it is quiet... a bit too quiet
[10:17am] aaime agrees
[10:18am] groldan joined the chat room.
[10:18am] aaime: groldan, meeting just started
[10:19am] aaime: "what's up" is the current topic
[10:19am] groldan: Hi sorry I'm late. Stayed woke up late lastnight helping sophia with some arcsde stuff
[10:19am] funky_c needs to research more things to test; figure out what's changed.
[10:20am] groldan is looking for what wicket UI issues to tackle today, have been helping with some arcsde export stuff and fixing UI bugs
[10:20am] groldan is also gonna be back to contract work this week
[10:21am] simboss: simboss: playing with modules inherited on gt-trunk
[10:21am] jdeolive: cool
[10:21am] jdeolive: 1) road map
[10:21am] jdeolive:
[10:21am] sigq: Title: Roadmap - GeoServer (at
[10:21am] jdeolive: so as aaime mentioned on the mailing list there are lots of issues for 1.7.5
[10:21am] aaime: yeah
[10:21am] jdeolive: so we shoud probably try to prioritize the important ones
[10:21am] jdeolive: and push back the rest
[10:21am] aaime: indeed
[10:22am] jdeolive: so... anyone want to bring up issues they think are important and/or they want to fix for 1.7.5
[10:22am] aaime: (important, or very quick to fix )
[10:22am] jdeolive:
[10:22am] sigq: Title: Issue Navigator - (at
[10:22am] jdeolive: aaime: good point, low hanging fruit is good too
[10:23am] aaime: GESO-2316 and GEOS-2011 do not look very critical to me
[10:23am] jdeolive: oops, wrong list
[10:23am] jdeolive:
[10:23am] sigq: Title: Issue Navigator - (at
[10:24am] groldan: that seems to be 2.0beta2
[10:24am] jdeolive: aaime: agreed, should be push back to 2.x?
[10:24am] aaime: yeah
[10:24am] aaime: jdeolive, shouldn't the list be:
[10:24am] sigq: Title: Issue Navigator - (at
[10:25am] jdeolive: doh... sorry, tried to paste that one
[10:26am] aaime: Other issues I'd like to fix:,,,
[10:26am] aaime:
[10:26am] sigq: Title: GEOS-2782 MaxFeatures equal to Integer.MAX_VALUE not disgarded when requesting multiple layers - (at
[10:26am] aaime: This one looks annoying pesky as well:
[10:26am] sigq: Title: GEOS-2770 FeatureType folder does not get deleted if it contains schema.xml and/or schema.xsd - (at
[10:27am] aaime: (annoying was meant to replace pesky... whatever)
[10:27am] aaime: is related to some recent Tike fixes, isn't it?
[10:27am] sigq: Title: GEOS-2269 exception should be thrown when user tries to submit a gml:Surface or gml:Curve in a WFS transaction - (at
[10:27am] jdeolive: cool, aaime want to up the priority on those to critical in order to prioritize?
[10:28am] aaime: yeah, thought some of the above fall more in the "ashamed to be there and does not look hard to fix" category
[10:28am] aaime: more than in "the world will stop if we don't fix those" one
[10:28am] aaime: but yeah, we just have priority to categorize
[10:29am] jdeolive: yeah
[10:30am] jdeolive: so perhaps people should go through and up the priority of things they want to fix?
[10:30am] jdeolive: and we can go from there?
[10:30am] groldan: this one seems to fall into andrea's category too:
[10:30am] sigq: Title: GEOS-2703 Completely wipe out the "schema base" drop down from the feature type editor - (at
[10:30am] aaime: jdeolive, works for me
[10:30am] jdeolive: maybe reply to teh road map update mail with the ones that have been upped?
[10:30am] aaime: groldan, yeah, there are two html things that are still lingering
[10:30am] aaime: that should not be visiable
[10:30am] CIA-21: aaime * r12536 /trunk/src/ (3 files in 3 dirs): GEOS-2790, Cannot download WCS outputs coverages generated with store=true
[10:31am] aaime: (and they are out of proper layout too, which makes them more visible)
[10:31am] groldan: what's the other one?
[10:31am] aaime: two html elements I mean
[10:31am] Bojan left the chat room.
[10:34am] aaime: jdeolive, shall we move on
[10:34am] aaime: or we re-prioritize issues now?
[10:34am] jdeolive: sure, so if everyone could scan through jira at some point and prioritize that would be great
[10:34am] jdeolive: moving on
[10:34am] jdeolive: 2) mvn site
[10:34am] jdeolive: aaime?
[10:34am] aaime: yeah, I was wondering if
[10:35am] aaime: people had a look at the site I geneated for GeoTools, and if the same could be useful for geoserver as well
[10:35am] aaime: (and if so, if we should schedule site generation in hudson)
[10:35am] jdeolive: aaime: do you have teh link handy? scanning through the gt devel list...
[10:36am] aaime:
[10:36am] sigq: Title: Geotools 2 - About (at
[10:36am] sonicx_ left the chat room. (Remote closed the connection)
[10:36am] jdeolive: thanks
[10:36am] aaime: and here is a sample set of reports:
[10:36am] sigq: Title: Geotools 2 - Generated Reports (at
[10:37am] jdeolive: yeah, i think this would be great for geoserver
[10:37am] aaime: which is the (only imho) interesting part of the site
[10:37am] aaime: if we do that, we should setup a place where the generated site is going to be published
[10:37am] aaime: and then I can give a crack to site generation
[10:38am] jdeolive: sounds good... maybe we can talk to arne about setting up a or something
[10:39am] aaime: or we can just copy everytihng into the nightly section
[10:39am] aaime: the site is going to be useful to devs mostly anyways no?
[10:39am] aaime: (this would make for a simple local file copy, wouldn't it?)
[10:40am] groldan left the chat room. (Read error: 60 (Operation timed out))
[10:40am] jdeolive: sure, that works fine too
[10:40am] aaime: I see there are quite a bit of documentation tasks open for 1.7.5 too
[10:41am] Topic changed to "0) what's up 1) road map 2) mvn site 3) chart module" by aaime.
[10:41am] aaime:
[10:41am] jdeolive: yeah... i was actually thinking i would put a couple of days toward doc tasks
[10:41am] aaime: a couple of days.. this week
[10:41am] aaime: (release is scheduled for Monday)
[10:42am] jdeolive: yeah, this week
[10:42am] jdeolive: but i guess if some docs slide in late that won't be a show stopper, can always update hte doc artifact
[10:43am] groldan joined the chat room.
[10:43am] aaime: well, site wise, I guess I'll ask on the list too
[10:44am] aaime: to see how other devs feel (for the mometn we have two +1)
[10:44am] jdeolive: cool
[10:45am] aaime: let's move to the next topic?
[10:45am] jdeolive: sounds good
[10:45am] jdeolive: 3) chart module
[10:45am] aaime: wondering how to include it in the release
[10:46am] aaime: official package, or we just link it from the blog
[10:46am] aaime: or... what?
[10:46am] jdeolive: is there an extension for it?
[10:46am] aaime: what is the way to include in the release a package that is not an extension
[10:47am] aaime: jdeolive, you suggested to make it distributed as unsupported or something
[10:47am] aaime: that's why I'm asking
[10:47am] jdeolive: right, sorry just trying to jog my memory
[10:47am] jdeolive: so how does it live right now? community module?
[10:47am] aaime: it's a build profile
[10:47am] aaime: there is no need for a module
[10:47am] aaime: since it provides a SPI extension
[10:48am] aaime: so it's more or less like a datastore dependency
[10:48am] jdeolive: so only a geotools lib?
[10:48am] aaime: yeah
[10:48am] jdeolive: yeah, i would say still an extension like we do for other ones
[10:48am] jdeolive: but for this time around
[10:48am] jdeolive: we can just do it up manually if we don't want to do up the full extension, we have done it in teh past
[10:49am] jdeolive: but i would say eventually an extension which contains the proper profile woudl be the way to go
[10:49am] aaime: I'm ok to make the proper extension module
[10:49am] aaime: which in fact would just contain deps
[10:49am] jdeolive: cool, i would say lets go that route
[10:49am] aaime: ok
[10:50am] aaime: I guess this is it?
[10:50am] aaime: anything else?
[10:50am] jdeolive: cool
[10:51am] jdeolive: someone on the logs?
[10:51am] groldan: about a deps only extension
[10:51am] groldan: what hapens if we set the pom's packaging to pom instead of jar?
[10:52am] groldan: does that still work and we avoid having an empty jar for the extension itself?
[10:52am] jdeolive: i don't think that works, but would have to verify
[10:52am] jdeolive: what i suggest doing instead is just not picking up the empty jar in the release artifact desciriptor

Posted at 02 Jun @ 12:55 PM by Justin Deoliveira | 0 comments
Last changed: Jun 01, 2009 09:27 by Andrea Aime

0) What's up
1) Road map update
2) Dual meeting

jdeolive: 0) what's up
***jdeolive has been catching up on outstanding jdbc-ng work and patches
***aaime has been putting togheter a relase and is know fiddling with wms and memory consumption
***funky_c needs to file some bugs on style and store deletion.
***groldan working on various arcsde and gt-wfs patches to apply, right now figuring out why it takes so long for a wfs request to start streaming, catching up on some UI stuff, going to stress test arcsde gce over the week
jdeolive: groldan: i have some thoughts about the high setup cost, i have been looking at encoder performance recently
groldan: btw, the UI does not have a setting to change the output strategy, does it?
aaime: never had, no
groldan: jdeolive, cool, can we catch up after the meeting?
jdeolive: groldan: sounds good
groldan: thanks
jdeolive: moving on
jdeolive: 1) road map
jdeolive: i just upated for the road map for beta 2
jdeolive: Andrea made some additions for beta2, and for 1.7.5
jdeolive: as discussed on the mailing list
jdeolive: any other major additoins?
sigq: Title: Roadmap - GeoServer (at
aaime: nope
aaime: are all of those new feature really scheduled?
aaime: (
sigq: Title: Issue Navigator - (at
funky_c: jdeolive: I'll have some deletion bugs in the next couple hours.
jdeolive: aaime: probably not
jdeolive: beta2 will need to be pruned down
aaime: that too
jdeolive: funky_c: cool, sounds good, feel free to schedule them against beta2 and make them blockers
jdeolive: unless you think its not appropriate
funky_c: Ya, def.
jdeolive: cool
jdeolive: so jira for beta2 needs some love
jdeolive: i can try to get to that today or tomorrow
jdeolive: unless of course someone else wants to
aaime: be my guest
jdeolive: agerber, dwins: does the stuff you guys are working on need to show up on the road map?
jdeolive: or is it already one
jdeolive: aaime: thanks
jdeolive: already one = already on
dwins: it is not on the roadmap. however, it will probably need to remain a community module for a while
dwins: having it integrate properly with geoserver will be a fairly tall order
groldan: che, jdeolive, the arcsde gce fix may worth being mentioned
sigq: Title: GEOS-2748 SDE Raster issue - image shifted - (at
groldan: (sorry for getting back to 1.7.5, was lurking around issues)
jdeolive: dwins: cool, thanks for the update
jdeolive: groldan: cool
aaime: out of curiosity, what is it that dwins is working on?
dwins: i am not working on geoserver this week, it is all javascript
aaime: You mentioned a community module above?
jdeolive: dwins: i think aaime is referring to the community you speak of
dwins: aaime: I am talking about the printing service that we discussed at last week's meeting
aaime: mapfish integration
dwins: yes
aaime: ok, I see
aaime: I'm courious to see how that is evolving
aaime: any way I can be kept in the loop?
dwins: yes, i've been meaning to talk to alan about sending status updates to the respective mailing lists
jdeolive: agerber: feel free to chime in with an update of what you have been up to
jdeolive: continuing on
jdeolive: a date for beta2 needs to be set
jdeolive: thoughts on that?
jdeolive: 2 weeks?
agerber: jdeolive: nominally i've been working on putting together mapfish printing functionality as a module for geoserver
aaime: jdeolive, wans't beta2 already scheduled?
agerber: jdeolive: but much of my time the last week has been getting a grip on various gis concepts and how geoserver's put together and the infrastructure supporting it
jdeolive: agerber: cool, how is it progressing? anything interesting to report, or are you still tackling that far too steep geoserver learning curve?
jdeolive: agerber: gotcha
jdeolive: aaime: in jira... let me check
aaime: I used that date to discuss complex features integration, tutorial and so on
aaime: I believe it was June 15
jdeolive: aaime: yup
jdeolive: how does that sound to anyone any objections?
agerber: jdeolive: i think i'm nearer the top of the learning curve, and hopefully might make something build in the next couple days
aaime: works for me
bj0rn2 [] è entrato nel canale.
aaime: errr.... 1.7.5 next week?
aaime: ouch
jdeolive: aaime: that was a date just thrown out by me, there was noone with a hard deadline to have that
jdeolive: so we can easily push that back
jdeolive: agerber: cool, thanks for the update
aaime: well, if we want to keep the WMS changes in it would be better to have the reelase date pushed away at least one week
aaime: so that people other than me can check them
jdeolive: sounds good, lets propose a push back
juliatorti [] è entrato nel canale.
aaime: otherwise, let's keep the release date and let's integrate them in 1.7.6 (if we have that release in mind)
jdeolive: either or for me
jdeolive: since 1.7.x is pretty stable at this point, i think we can lower the frequency of releases
jdeolive: to every couple of months
jdeolive: but that is just me
aaime: agreed
aaime: let's propose that on the mailing list?
jdeolive: sure, i will roll that into the road map update
aaime: look at this:
sigq: Title: Browse Version - (at
aaime: atm we have only 11 fixes for 1.7.5
jdeolive: haha
jdeolive: then yeah, i would say lets push back indeed
aaime: anyways, let's hear the PSC at a large on the ml
jdeolive: sounds good
jdeolive: is that it for the road map?
aaime: I guess so
jdeolive: 2) Dual meeting
jdeolive: aaime: ?
aaime: Right, jgarnett has been proposing to have an earlier meeting
aaime: where australian people can participate
aaime: I wanted to check how people feel about this
jdeolive: cool, taht will work well for me in july
aaime: lol
jdeolive: for june that will be hard as I will back to MST
jdeolive: but depends on what time
aaime: the time was not setup
aaime: I can be on both meetings if the eaarly one is 8-10 hours before this one
aaime: anyways, whoever is there
aaime: the idea was to save the logs and pick up the topics again in the second meeting
aaime: to coordinate them better it may be a good idea to decide by mail the day before what the topics would be
aaime: so that the two meetings can discuss the same
jdeolive: sounds good to me
aaime: Ok, I'll send a mail to the ml and see how this works out
jdeolive: sounds good
jdeolive: anything else for meeting topics?
aaime: hmmm... nope?

Posted at 01 Jun @ 9:23 AM by Andrea Aime | 0 comments
Last changed: Apr 28, 2009 13:05 by Gabriel Roldán

0) What's up
1) Roadmap update
2) Docs going live
3) 1.7.4 Release (Without OS X artifact?)

 <dwins> okay, meeting time?
 <arneke> yeah
 <dwins> aaime: groldan, bmmpxf?
 <groldan> yup
 <aaime> yo
 <arneke> do we have any topics?
 <groldan> 0) What's up :)
 <bmmpxf> 1) docs going live
 <bmmpxf> does someone want to do roadmap in lieu of jdeolive?
 * arneke has changed the topic to: 0) what's up 1) Docs going live
 <aaime> yeah
 <arneke> that would be more of a 2.0 roadmap, wouldnt it ?
 * aaime has changed the topic to: 0) what's up 1) Docs going live 2) roadmap update
 <bmmpxf> there is 1.7.5 still, though, IIRC
 <groldan> arneke, roadmap should be topic 1) I guess?
 * groldan has changed the topic to: 0) what's up 1) roadmap update 2)  Docs going live
 <groldan> lets go?
 <arneke> +!
 <arneke> +1
 <arneke> :)
 * arneke finished up gs 1.7.4 release, OS X and Windows artifacts are coming. Will release tonight or tomorrow
 * groldan was settling down some 2.0-beta1 issues with aaime and rpenate and UI bugfixing
 * awright (n=awright@ has joined #geoserver
 * dwins is debugging some javascript stuff as usual
 * aaime is looking into some estimates
 <aaime> (and looking into further UI fixes)
 <groldan> ok move on, 1) roadmap update
 <groldan> the page is not recognizing the macros
 <sigq> Title: Roadmap - GeoServer (at
 * bmmpxf is making screencasts, building a new Windows installer, and fixing up docs, and being late for 0)
  bmmpxf brandon
 <sigq> Title: Issue Navigator - (at
 <aaime> Well, as far as I know 1.7.x is entering deep freeze as far as OpenGeo is concerned
 <arneke> two jdeolive , two afabiani
 <arneke> for 2.0 though,, there are the improvements
 <aaime> so all bugfixes and new features are on the spare time of developers, or on developers from other organisations?
 <arneke> but first and foremost lots of bugs
 <aaime> for 2.0.x afaik we're pushing hard to make the finishing touches to the 2.0.x UI
 <arneke> depends on who is reporting them, right?
 <aaime> so that we can release 2.0 final
 <bmmpxf> When do we branch 2.0?
 <aaime> arneke?
 * bmmpxf is feeling the crushing weight of having to update the docs for 2.0
 <aaime> when it's stable enough I guess, I don't know any deadlines
 <aaime> but I know OpenGeo is moving all resources available to finishing up 2.0
 <arneke> aaime: if an enterprise customer needs a fix, we've committed to fixing it, I think
 <aaime> and leaving 1.7.x to critical/blocker bugfixes only
 <aaime> arneke, yeah, and customer related fixes
 <aaime> but community related fixes are frozen
 <arneke> sure
 <aaime> as far as OpenGeo is concerned, that is
 <aaime> other developers, or OpenGeo developers in their spare time, can still do stuff :-)
 <groldan> right
 <arneke> I'm a big +1 on pushing 2.0, it only gets harder the longer we run both
 <aaime> (see my charting thing)
 <arneke> anyway, doesnt matter
 <groldan> that's for 1.7.x
 <groldan> for 2.0-beta1, what's the intended release date, if any?
 <aaime> not sure
 <aaime> either
 <groldan> and does anyone know why do we get "Unknown macro: {jiraissues}" on the roadmap page?
 <aaime> as soon as possible  afaik
 * vheurteaux (n=vheurtea@ has joined #geoserver
 <aaime> groldan, because the jira plugin was making confluence oom once a day
 <arneke> groldan: it's the jira plugin (maybe RSS too) that fouled up confluence constantly
 <groldan> ok, that's cool, there's lot of work to do anyway
 <sigq> Title: Issue Navigator - (at
 * vheurteaux (n=vheurtea@ has left #geoserver
 <arneke> those are the improvements,, but I dont think that's where the meat is
 <aaime> arneke, no, it's not
 <arneke> sorry, saw afterwards the resolution filter is wrong anyway
 <aaime> the next raft of activity is here:
 <sigq> Title: [#GEOS-2705] Wicket UI styling and usability improvements - (at
 <aaime> (at least from OpenGeo)
 <aaime> not sure how things are going with the appschema development
 <aaime> as in, will it be ready for 2.0.0 or delayed?
 <aaime> and without a deadline is also hard to ask
 <groldan> as far as I knew things were going quite well, like in they were finally able to encode gml, except a couple minor exceptions
 <aaime> so, is it just a matter of documentation and testing?
 <groldan> so I'll cross fingers and wish them the best of lucks :)
 <aaime> especially , document how to configure the thing
 <aaime> (and see how the UI will react to that)
 <groldan> ah good point
 <groldan> I guess I was missing that (fundamental) part
 <arneke> hm, are they running with the old UI now ?
 <groldan> but who knows, it'd be good to have ben around, we can ask on the ml though
 <aaime> the are runing without UI afaik ;)
 <groldan> meaning they're running community/web2
 <groldan> but have no way to configure an app-schema ds
 <groldan> yet
 <aaime> ideally I would like to have the complex stuff ui-less 
 <aaime> like, you go and hack xml files to configure it
 <aaime> but at the same time, it would be nice for the ui not to break solid in that case
 <groldan> the simplest way of doing so is gonna be:
 <aaime> (when trying to open details about a complex type)
 <groldan> manually set up an xml mappings file, and then provide the datastore factory with the file url for the config file
 <groldan> just like for any other file based datastore
 <aaime> groldan, agreed, yet the UI might still balk at a real complex feature
 <groldan> like in freezing until the beast wakes up?
 <groldan> or even worse... OOM'ing...?
 <groldan> I don't know, I guess even if available for 2.0, it will still need specialized manual setup, including data config and -XmxALOT
 <groldan> still, that's app-schema datastore specific
 <groldan> geoserver's ability to serve complex stuff is an orthogonal concern
 <aaime> groldan, just to name a concern among others
 <groldan> but who am I to assess if they'll be ready or not for 2.0...? shutting up here :)
 <aaime> attributes listing like in the old ui
 <groldan> do we list attributes on the new ui at all?
 <aaime> and no, I was not thinking about the OOM thing, that's a sin they bring on them by going complex features
 <aaime> groldan, we do, but there is a bug that prevents it from working properly
 <aaime> on any newly configured layer
 <groldan> ok I see it now at the data page
 <groldan> hmmm so your concern is how to handle that for nested attribtues
 * juliatorti has quit (No route to host)
 <aaime> yeah, and where code making that assumption may be located
 * juliatorti (n=juliator@ has joined #geoserver
 <groldan> funny enough, in complex land there are properties of an abstract type, so you don't know but at runtime the actual attribute instance type
 <groldan> lot of fun
 <aaime> anyways, a topic for the ml
 <groldan> yup
 <aaime> let's move to the next one?
 <aaime> 2) Docs going live
 <arneke> yeah
 <aaime> bmmpxf?
 <bmmpxf> hi
 <bmmpxf> well, not much to say here, other than that the new Sphinx documentation is going live, bascially now.
 <bmmpxf> LInks are changing...
 <bmmpxf> ...the DOCS artifact for 1.7.4 will be Sphinx and not Confluence
 <bmmpxf> ...and the wiki now has a deprecated header
 * arneke has changed the topic to: 0) what's up 1) roadmap update 2) Docs going live 3) Release (without OS X artifact)?
 <arneke> looks good :)
 <aaime> hmmm... ok, provided the new docs stil point to the old ones as a "further information" thin
 <aaime> (thing)
 <bmmpxf> aaime: That can easily be added to the main page at
 <aaime> given we all know the new ones only cover part of what the old ones did
 <bmmpxf> (which is still being themed)
 <aaime> bmmpxf, I would like to have that in the PDF as well
 <aaime> you download the thing, but you know that online there's more, even if not as well organized
 * sonicx has quit (Read error: 104 (Connection reset by peer))
 <aaime> so, I would really put the links in the first page of the sphinx doc
 <aaime> with a little explanation of why there are two docs sets and what the online one is still going to be used for
 <bmmpxf> aaime: Well, it's matter of timing.  The artifacts are already created.  I'd rather create a README file with the download for now, and then add the content to the 1.7.x so we don't need to do that wya again/
 <aaime> "content"?
 <bmmpxf> "content" = "a little explanation of why there are two docs sets etc"
 <aaime> works for me
 <bmmpxf> aaime: I'm not trying to shirk off duties, but I don't want this to push back the release date.  :(
 <aaime> no no, I understand
 <arneke> I can just edit it into the HTML and repack too
 <aaime> bmmpxf, you're going live with a search box that does not work ;-)
 <aaime> (
 <sigq> Title: GeoServer User Manual GeoServer User Manual v1.7.4 documentation (at
 <bmmpxf> aaime: Not the worst thing we've ever gone live with.  :)
 * cbriancon has quit ("Leaving.")
 <aaime> can we at least hide it?
 <aaime> (or is that too hard?)
 <bmmpxf> But thanks for pointing it out, I actually forgot that didn't work.
 <arneke> yeah, I can hide it in CSS if I am going to do the edit anyway
 * bmmpxf defers to others for this decision.
 * arneke is +1 to adding a small blurb about more info at , and +1 for hiding search box in css
 * bmmpxf is +1 on arneke making the changes. :)
 <aaime> ok, we're all a happy family 
 <aaime> let's move to the next point then? ;-)
 <arneke> yeah
 <arneke> 3)
 <bmmpxf> anyway, that's all I wanted to say for 1).  It's not the end of the story, but the beginning.  :)
 <arneke> so all artifacts are ready to go
 <arneke> but Mac OS X dmg file requires OS X
 <arneke> which nobody on the geoserver team except jdeolive has
 <groldan> I do
 <arneke> aha :)
 <groldan> but have no idea how to create it
 <groldan> I can volunteer if you have some clues though
 <aaime> are there instructions?
 * arneke didnt think your monster was a mac ;)
 <sigq> Title: Release Guide GeoServer Developer Manual v1.7.4 documentation (at
 <groldan> I changed my monster
 <groldan> though I miss clising it with a hit... 
 <arneke> :) hehe
 <aaime> does not requires anymore a punch to close it eh?
 <groldan> from time to time I open it just to close it 
 * sonicx ( has joined #geoserver
 <arneke> ok
 <arneke> so ... problem is, I think this assume you have the complete checkout
 <arneke> built etc
 <arneke> let me see how big my directory is
 <arneke> 655 mbyte
 <groldan> its gonna be bigger than the checkout thanks to the build artifiacts
 <arneke> yeah, by quite a bit 
 <arneke> that would take me two hours to upload
 <groldan> what's the svn tag?
 <arneke> 1.7.4
 <arneke> and everything is comitted
 <arneke> so the catch is you have to build GT 2.5.5 too
 <groldan> ok, I can get things going on the background until having to build the macos artifact
 <groldan> np
 <arneke> cool :)
 <arneke> thanks
 <arneke> so 
 <groldan> wait, isnt 2.5.5 deployed already?
 <arneke> i'll be on the road parts of tomorrow ,, i'll try to get it out tonight
 <aaime> I deployed it
 <arneke> ok, nvm then
 <aaime> could not release it cause the source packages did not build without -Dall
 <groldan> so I don't need to build it at all
 <aaime> on a clean repo
 <aaime> groldan, you have
 <aaime> because I deployed only the supported packages
 <groldan> ok
 <aaime> but you can build it quickly using "mvn clean install -Dall -DskipTests"
 <groldan> skipTests is my allee anyway
 <groldan> does it build a macos universal installer?
 * arneke has no idea
 <arneke> i guess ... .dmg ?
 <arneke> can check the 1.7.3 release
 <groldan> yup, universal means a dmg with an intel and a ppc version inside afaik
 <groldan> but probably we are doing intel only
 <groldan> dunno
 <groldan> arneke, so I build it, upload it to.. say... atlas for you to pick it up?
 <arneke> yes please :)
 <arneke> just shoot me an email
 <groldan> np, hope it goes alright, keep you posted
 <arneke> thanks
 <groldan> you're welcome
 <groldan> meeting over?
 <aaime> yep
 <arneke> think so :)
Posted at 28 Apr @ 1:01 PM by Gabriel Roldán | 0 comments
Last changed: Apr 21, 2009 17:14 by Mike Pumphrey

0) What's up

1) In a brave new Oracle world

2) roadmap

16:07 aaime: arneke, dwins, jdeolive, simboss, bmmpxf, iwillig
16:07 aaime: meeting time?
16:07 iwillig: sure
16:07 jdeolive: howdy
16:08 dwins: hi there
16:08 arneke: hi
16:09 bmmpxf: hi
16:09 aaime: topics?
16:09 arneke: 1.7.4 release?
16:10 aaime: 1.7.4 part of roadmap maybe?+
16:10 arneke: sure
16:11 arneke: 0) ?
16:12 jdeolive: is working on fleshing out the resource pub work
16:12 aaime: is working on the security subsytem ui and hiding a secret
16:12 dwins: is working on javascript stuff
16:12 arneke: has done about as much i18n as he can do. Switching to 1.7.4 release tomorrow, or GWC on trunk integration
16:12 bmmpxf: is training new team member awright (please say hello), cleaning up docs, and working on a new Windows installer for GeoServer.
16:12 awright: hello!
16:13 dwins: bmmpxf: looks like you've trained her well
16:13 jdeolive: hi awright, welcome to your first geoserver IRC meeting
16:14 bmmpxf: now no hazing, please.
16:14 aaime: welcome awright
16:14 awright: thank you. its been very interesting so far (even with the hazing. mike "made" me eat 100 malt balls.)
16:15 bmmpxf: claims innocence
16:15 iwillig: is still working on alachua contract work
16:15 dwins: anyway... time to move on to 1) ?
16:16 aaime: please
16:16 aaime: 1) In a brave new Oracle world
16:16 aaime: eh... "all base are belong to... Oracle" it seems
16:16 bmmpxf: I guess we can call this topic a Sun-set.
16:16 aaime: lol
16:16 jdeolive: haha
16:16 arneke: woot finally we can rely on solid vendor for all our stack needs
16:17 aaime:
16:17 jdeolive: loses his lunch
16:17 bmmpxf: As a practical measure, what does this mean for us?
16:17 arneke: I am joking
16:17 aaime: paints himself more Oracle work and starts crying out loud
16:17 jdeolive: not sure it means all that much... unless oracle decides to make some big moves
16:18 jdeolive: if anything it could be beneficial
16:18 aaime: yeah, we'll see
16:18 jdeolive: since the binding between oracle and java will undoubtely get tighter, we could be in a better position to offer support than others
16:18 dwins: gets ready to start improving OpenJDK
16:18 arneke: MySQL is dead (who cares) , Java will either be in trouble, or maybe OpenJDK will get a big push from IBM who also depends heavily on Java
16:18 bmmpxf: Someone online theorized that Oracle would deprecate MySQL in favor of Oracle Express.
16:19 arneke: Java = Java for anyone but Oracle
16:19 jdeolive: dwins: that probably brings up a good point, we may want to start looking at seriously considering running on other JDK's
16:19 aaime: there is already a full community maintained fork of it going: MariaDB
16:20 dwins: jdeolive: that would improve our story for installing on linux distros too
16:20 aaime: yeah, been twisted on the whole "run on openjdk" thing for a while
16:20 dwins: so i'd be interested in at least investigating
16:20 arneke: I donth think it matters much,,, postgres is just too hard to admin, but the LAMP stack can easily replace the M
16:20 aaime: make us look good because we play with openjdk, look bad because openjdk graphic performance is absymal
16:20 dwins: it's also kind of wrong in some cases.
16:21 dwins: had some "fun" debugging wms decorations when his path got messed up
16:21 aaime: he
16:21 arneke: dwins: the fonts?
16:21 dwins: yeah
16:21 bmmpxf: I think "other" JDK support is best for the long run, don't you?
16:21 arneke: phew those scared me a bit
16:22 aaime: bmmpxf, agreed
16:22 bmmpxf: (not that it's trivial, of course.)
16:22 aaime: mcrmcr is also trying to bring the whole thing to build on IBM JDK
16:22 aaime: (apparently, it already runs fine)
16:22 arneke: what kind of license does IBM have anyway?
16:23 arneke: (obviously good enough that they could afford to not buy Sun)
16:24 aaime: don't know
16:24 aaime: it's also somewhat weird in that jdk 6 (was) only available on linux or something like that
16:24 dwins: :w
16:25 dwins: sorry, wrong window.
16:25 arneke: Ibm has an agreement to license Java until 2016
16:28 jdeolive: should we move onto the road map?
16:28 aaime: yep
16:29 jdeolive: the road map page seems down again...
16:29 jdeolive: but i think the biggest thing is probably 1.7.4
16:29 arneke: jdeolive: yeah, mentioned it in the status email
16:29 aaime: I believe they disabled the jira plugin to get back stability?
16:29 arneke: priorities email, sorry
16:29 arneke: yeah
16:29 arneke: that plugin is dead until we hire a real sysadmin
16:30 arneke: or atlassian writes a new version
16:31 jdeolive: fair enough... i guess we need to come up with a another way to do the road map then... i guess manually
16:31 aaime: arneke was consdiering the rss plugin in its place, right?
16:31 aaime: does that work?
16:31 bmmpxf: wow, without JIRA integration, there's very little use for Confluence in general, right?
16:31 arneke: aaime: yeah, something like that, havent had a chance to try yet
16:32 aaime: for the short term, what about jsut linking permlinks to the jira queries
16:32 arneke: bmmpxf: there was only one page anyway?
16:32 aaime: and we follow them up to see what's what
16:32 simboss: aaime: sorry I am back
16:32 jdeolive: aaime: good idea
16:32 arneke:
16:32 sigq: Title: Issue Navigator - (at
16:32 simboss: better later than never
16:33 aaime: well, that is a bit more than just the roadmap interesting items
16:33 aaime: it's the open issues
16:34 aaime: this is more it:
16:34 sigq: Title: Issue Navigator - (at
16:35 aaime: plus this one for the bugs:
16:35 sigq: Title: Issue Navigator - (at
16:35 aaime: is that it jdeolive?
16:35 aaime: new features, and critical/blocker bugs?
16:35 jdeolive: aaime: looks like it yeah
16:36 arneke: simboss has one in progress ?
16:36 arneke: moved to 2.0 , we've discussed it before
16:36 sigq: Title: GEOS-2515 Have KML super overlays use GWC by default - (at
16:37 aaime: and the release is being built... tomorrow?
16:38 jdeolive: arneke is going to be on point for this release
16:39 arneke: GEOS-2116, GEOS-2363,GEOS-2028 are marked as in progress?
16:39 simboss: arneke when are yuo going to cut the release?
16:39 arneke: that's jdeolive, simboss and aaime
16:39 arneke: simboss: we're goint to discuss that now I guess
16:40 aaime: uh?
16:40 arneke: so i guess I need to build a release, do the CITE testing
16:40 aaime: ah, GEOS-2028 has been in progress forever
16:40 arneke: if that works, Andrea or Justing have to make me a GeoTools release
16:40 arneke: Justin
16:41 arneke: and then I repeat CITE for the actual release?
16:41 jdeolive: arneke: by GEOS-2116 do you mean GEOS-2126?
16:41 arneke: sorry, yes
16:42 jdeolive: yeah... not really in progress, i will update it
16:42 aaime: I updated mine
16:43 simboss: for GEOS-2927 I have already committed on 1.7.x
16:43 simboss: will port to trunk afterwards, however
16:43 simboss: most part of the work I have done
16:43 simboss: was with the mosaic
16:43 simboss: which is committed as of now on 1.7.x
16:43 simboss: and 2.5.x
16:43 arneke: simboss: should be moved to 1.7.5 then ?
16:43 sigq: Title: GEOS-2363 Commit various improvements for RESTConfig - (at
16:43 simboss: I would like to document the code a bit more though
16:44 simboss: I would say
16:44 simboss: that we could just kill this group
16:44 simboss: after jdeolive work on rest
16:44 arneke: ok
16:44 simboss: about to do that
16:46 simboss: ( is it too late to add topics to the discussion?)
16:46 arneke: so I guess I'll just start running CITE tests tomorrow, and we'll see from there (I'll make a note of the revision, can always use -r when creating the tag)
16:47 arneke: we've got another 14 minutes, so should be okay
16:47 simboss: k, firsto one
16:47 simboss: thing
16:47 simboss: I am going to document a bit more the code I improved (or at least I tried to improve ) for the mosaic
16:47 simboss: between tonight and tomorrow morning
16:47 simboss: when are you going to do the final cut on the gt branch?
16:48 arneke: i can't do it, since im not a GT comitter
16:49 simboss: good answer
16:49 arneke: tomorrow afternoon at the earliest though
16:49 simboss:
16:49 simboss: k, that is fine
16:49 simboss: I actually have 1 interesting topic and 1 not interesting
16:49 simboss: which one are we going to start with?
16:49 simboss:
16:49 arneke: not interesting
16:50 simboss: I will take your advice and I will start with the interesting one
16:50 simboss: db based catalog
16:50 simboss: (sorry arneke, time was running out :-9 )
16:51 simboss: I am looking at alessio's work
16:51 simboss: in order to see if I can make good use of it in 1.7.x
16:51 simboss: we might want to port to trunk as well
16:51 simboss: with some help of course
16:51 simboss:
16:52 arneke: wonders why the room feels so empty
16:52 aaime: that looks like a very big change
16:52 simboss: I don't think we have ever discussed this topic here, so basically I am asking for thoughts/ideas from the others about how the others think that such a thing should ne implemented
16:52 arneke: meaning you're planning on staying with 1.7.x for a while ?
16:52 aaime: or is it just a matter of plugging in a different catalog?
16:53 simboss: that's part of the feedback/thoughts/ideas I am looking for
16:53 simboss: I admit I have not yet looked at 2.0 closely
16:54 aaime: simboss, I guess we need two things:
16:54 simboss: so I am trying to understand how muc close we are to something stable
16:54 aaime: - extensions points to swap the catalog implementation (jdeolive?)
16:54 simboss: but also how much things have changed catalog/configuration wise
16:54 aaime: - an assesment of if/how you changed the catalog api to serve your ND needs
16:54 jdeolive: aaime: well that has always been the holy grail
16:55 jdeolive: a catalog interface swappable for a db backed on
16:55 jdeolive: but yeah... quite a bit of effort involved
16:57 aaime: so it seems some work is needed to start assessing the amound of work needed to actually do this
16:57 simboss: I'd say that I am past that point since
16:57 simboss: we have already done a prototype
16:57 simboss: but instead of having me telling you what we have done
16:58 simboss: (even because the thing has been around since Oct on a branch, so anyone could have looked at it)
16:58 simboss: I would like to hear your thoughts about how pors and cons
16:58 simboss: problems and issues
16:58 simboss: etc. etc..
16:58 jdeolive: yeah... unfortunately one has to dedicate quite a bit of time to do a decent review
16:58 simboss: since I want to do some refactor
16:58 jdeolive: been on the todo list.. but with all the 2.0 stuff going on... hard to find the time
16:59 simboss: yeah, sure, did not mean to say that someone should have actually done that
16:59 aaime: I only have a general opinion that basically emcompasses what I state above already
17:00 aaime: that is, first thing we need enough extension points to load a different catalog
17:00 simboss: that is more or less already in place
17:00 simboss: you as an instance mentioned today
17:00 simboss: the fact of having a cache
17:00 simboss: for performances
17:00 simboss: can you elaborate a bit on that?
17:01 aaime: simbos, you mean you already made the extension points in your branch
17:01 aaime: and that you can load either catalog by just setting a flat?
17:01 aaime: (flag)
17:01 simboss: not so easy
17:02 aaime: so what do you mean by "that is more or less already in place"?
17:02 simboss: 1> we (we== alessio and groldan)
17:03 simboss: have implemented the catalog interface that in turn uses an hibernate dao
17:03 simboss: 2> I have started to look a this work today and none of the people who worked on it area around (groldan and alessio)
17:03 simboss: which means I am still exploring stuff
17:03 simboss: but my feeling is that we can simply swap them
17:03 simboss: since the old catalog is deply wired inside the geoserver architecture
17:04 aaime: simboss, not so easy
17:04 aaime: first, we need a mediator that can choose the actual impleemntation
17:04 aaime: but this is the easy part
17:05 aaime: then we need to disable the loaders
17:05 aaime: or else, have them work only if the catalog is a certain implementation
17:05 aaime: (the classes that do read the xml files and populate the catalog)
17:06 aaime: hmmm... jdeolive, anything else assuming the api is the same 1-1?
17:06 jdeolive: nope, that just about sums it up...
17:06 jdeolive: i guess the xstream persister stuff with rest
17:07 jdeolive: that makes a few assumptions about the catalog
17:07 jdeolive: so it will need to be updated... but should not be too crazy
17:07 aaime: jdeolive, the db backed catalog shoudl not be loaded from xml anyways
17:07 aaime: (besides the case of importing a old catalog)
17:07 jdeolive: right, the rest config will still searialize the objects as XML
17:07 aaime: the db backed version would make live db queries for any method called afaik
17:07 aaime: aaah
17:07 aaime: right
17:08 aaime: simboss, performance wise
17:08 aaime: each service request access the catalog multiple times
17:08 aaime: if each api call results in a db query that speaks trouble
17:08 aaime: especially if the db is not embedded
17:08 aaime: so some kind of object cachign is in order
17:08 aaime: hibernate provides some declarative cachign that could be used
17:09 simboss: that's my plan
17:09 aaime: and ResourcePool provides also some higher level caching for gt2 objects
17:09 aaime: like stores
17:09 aaime: not sure if feature types are cached
17:09 aaime: FeatureSource isntances cannot be as they are not thread safe
17:09 simboss: tha's goingo to ber around anyway
17:11 aaime: so it does not seem such a huge undertaking?
17:11 aaime: it has lots of little details to be fixed thought
17:12 simboss: I got to dig a bit more and come back with more ideas
17:12 simboss: and doubts
17:12 simboss: topic 2, real quickly
17:12 simboss: please, anyone, give a try to the new mosaic on the 1.7.x branch
17:12 simboss:
17:13 aaime: ok, will try
17:15 aaime: meeting done I guess
17:16 arneke: yup

Posted at 21 Apr @ 5:11 PM by Mike Pumphrey | 0 comments
Last changed: Apr 07, 2009 12:47 by Mike Pumphrey

0) what's up?
1) roadmap

16:03:22 dwins: jdeolive, aaime, bmmpxf, arneke, simboss: meeting time?
16:03:27 aaime: yep
16:03:27 bmmpxf: hiya
16:03:34 jdeolive: yup
16:04:07 dwins: anyone have any topics?
16:04:18 aaime: roadmap!
16:04:25 aaime: (yay, got the easy one!)
16:04:31 dwins: cheater
16:04:35 aaime: lol
16:06:26 aaime: so, let's move forward?
16:06:38 dwins: aaime: sounds good to me
*16:06:43 dwins: 0) What's up?*
16:07:18 aaime: is feeling like an octopus, doing 8 different things at a time
16:07:20 jdeolive: is catching up on email and working toward 2.0-whatever
16:07:21 arneke: is chewing through internationalization, i.e. switching between tabs and cutnpasting a lot
16:07:55 bmmpxf: is still being doc steward, but slowly unloading this so others can take over
16:08:00 dwins: is playing with different layout configurations making sure watermarks act like they used to, etc.
16:08:15 dwins: also working on a livecd this week, that's kind of neat
16:08:51 jdeolive: is pretty excited about dwins rpm build script
16:09:11 aaime: wants a .deb
16:09:37 dwins: i'll work on a .deb after the livecd is holding together properly
16:09:54 dwins: is that it for 0) ?
16:11:44 aaime: yep
*16:11:48 aaime: 1) Roadmap update*
16:11:53 jdeolive:
16:11:56 sigq: Title: Roadmap - GeoServer (at
16:12:29 jdeolive: shall we go down the list
16:12:35 dwins: let's
16:12:40 jdeolive: dwins: you want to give updates on your issues first?
16:12:50 dwins: issues?
16:13:17 dwins: ah, didn't realize the KML -> gwc hookup was on my plate
16:13:31 dwins: so, no progress on that, but it's not a tough fix
16:14:03 jdeolive: cool, and the wms decoration stuff is on track?
16:14:20 dwins: as for the wms decorations, those are working and tested; I am trying a build now with them merged into the wms module to make sure nothing breaks.
16:14:34 jdeolive: good stuff
16:15:03 dwins: so both of those are on line for the release
16:15:34 jdeolive: arneke: is GEOS-2605?
16:15:51 arneke: jdeolive: nope, have not had a chance to work on any of this since 1.7.3
16:16:01 arneke: and neither of them are small
16:16:08 aaime: there is something wrong with the roadmap, the 1.7.5 bug fixes are actually for 1.7.4?
16:16:21 jdeolive: so sounds like they are not really in scope for 1.7.4?
16:16:48 jdeolive: aaime: fixing
16:16:52 arneke: no
16:17:13 arneke: i can maybe do one if I drop translation , but my understanding is that it is higher priority, which makes sense
16:18:01 jdeolive: arneke: cool, lets move em back
16:18:10 arneke: sounds good
16:18:58 jdeolive: confluence is not responding for me...
16:19:35 arneke: yeah, I think there are some problems with it
16:20:06 arneke: i thought it was just melkjug... but confluence seems to be just as bad at hammering the disk
16:20:23 arneke: (obvious now that it lives separately)
16:20:33 arneke: will see what I can do
16:21:32 jdeolive: moves to cached page
16:21:47 jdeolive:
16:21:48 aaime:
16:21:48 sigq: Title: GEOS-2498 Integrate GeoExt based styler - (at
16:21:49 sigq: Title: roadmap.jpg (at
16:21:58 jdeolive: this issue is not really in scope for 1.7.x, will move back
16:22:28 aaime: (the above being a screenshot of my hung firefox page, may be better than nothing )
16:22:59 jdeolive: aaime, i think you are up next for update
16:23:27 aaime: so I made the changes for GetFeatureInfo
16:23:49 aaime: whilst it's not "click and get what you see" it's much closer than before
16:24:29 aaime: then we have support for "sparse" shapefile that arcmap is so eager to generate after editing
16:24:42 aaime: (shapefiles with holes inside, removed features in particular)
16:25:04 aaime: I already updated some of the roadmap to remove the png8 + alpha channel thing
16:25:33 aaime: it would take me a week to implement it, I have to study algorithms better
16:25:53 jdeolive: cool
16:25:56 aaime: the ability to turn off label conflict resolution keeps on popping up and does not seem hard so I'd like to keep it there
16:26:21 aaime: and also the bug about rendering hanging if chars are too small must be squashed, too
16:27:16 aaime: I guess that's it for me?
16:27:33 jdeolive: cool
16:27:37 jdeolive: the only other issue from simboss
16:27:50 jdeolive: not sure he is around though
16:29:36 jdeolive: so.. anything else for the meeting?
16:30:59 arneke: would like to add that he loathes confluence
16:31:24 tenzochris: arneke: +1!
16:31:43 arneke: so I guess we're done

Posted at 07 Apr @ 12:40 PM by Mike Pumphrey | 0 comments
Last changed: Mar 31, 2009 14:52 by Mike Pumphrey

0) what's up
1) roadmap update
2) wfs generalization
3) sphinx docs update/plan
4) wms decorations

16:02:41 aaime: bmmpxf, dwins, groldan, iwillig, jdeolive_, simboss (mcrmcr too if you want)
16:02:41 bmmpxf: hi
16:02:41 aaime: GeoServer dev meeting time?
16:03:01 mcrmcr: yep
16:03:01 jdeolive_: sounds good
16:03:41 simboss: k
16:03:41 groldan: k
16:04:21 dwins: hi
16:04:31 jdeolive: anyone have any agenda items?
16:04:31 bmmpxf: Sphinx docs update/plan
16:05:01 dwins: wms decorations
16:06:11 jdeolive: if there is nothing else shall we start?
16:06:21 aaime: sure
16:06:21 dwins: let's
16:06:31 bmmpxf: 0) what'
*16:06:41 bmmpxf: 0) what's up*
16:06:41 bmmpxf:
16:06:51 jdeolive: has been working on persistence bugs popping up
16:07:01 dwins: hacking on some js stuff, following up on overdue threads
16:07:11 bmmpxf: is writing docs, and fixing up READMEs
16:07:11 dwins: (mailing list threads)
16:07:11 aaime: is all over the place... wicket ui, shapefile issues, oracle patches to review, (add your favourite item)
16:07:31 simboss: simboss, improving imagemosaic
16:07:31 groldan: is evaluating the state of app schema support on trunk
16:08:11 jdeolive: good stuff
*16:08:11 jdeolive: 1) road map update*
16:08:11 jdeolive:
16:08:21 sigq: Title: Roadmap - GeoServer (at
16:08:21 jdeolive: so it has been proposed
16:08:31 aaime: has a new feature he has to develop for a customer
16:08:31 jdeolive: to remove the restriction on bug fix only on 1.7.x
16:08:51 mcrmcr: simone, some times ago you asked for a merge of our imagemosiac components, whats happening here
16:08:51 jdeolive: right, and also chris hodgson from RR has a new feature he wants to add
16:09:11 aaime: what about "high resistance against new features" as opposed to total freeze?
16:09:11 aaime: Spring is beginning
16:09:21 jdeolive:
16:09:31 simboss: jdeolive, it has been asked by who?
16:09:41 aaime: by me, by RR
16:09:51 dwins: me too
16:09:51 aaime: and I believe by OpenGeo as well? (WFS generalization)
16:10:01 simboss: should we go to 1.-8 then?
16:10:01 simboss: 1.8
16:10:11 aaime: simboss, we're talking minor new features
16:10:31 aaime: like the ones that went into 1.7.x so far
16:10:31 simboss: generalization does not look minor to me
16:10:41 jdeolive: hmmm... i guess we could... but i think since these are relatively minor it makes sense to maybe just do them on 1.7.x
16:10:41 aaime: the way mcrmcr wants to do it, not at all
16:10:51 jdeolive: at least that owuld be my preference (one less branch to manage)
16:10:51 aaime: the way we want to do it, it's an addon to wfs only
16:11:21 aaime: what we want to do is to add support for functions output in wfs 1.1
16:11:31 aaime: which is a part of the standard we don't implement
16:11:31 mcrmcr: and for wms, I hope
16:11:41 aaime: arg, no!
16:11:51 simboss: as long as we are converging to something stable I am fine
16:12:01 aaime: ours is totally different use case with tiny funding
16:12:01 simboss: functions ==?
16:12:11 aaime: we only need to avoid genering a lot of gml
16:12:21 aaime: not to leverage vector pyramids in the datastore level
16:12:41 aaime: simboss, wfs 1.1 allows to choose propertyName
16:12:41 aaime: but if you look at the schema
16:12:51 aaime: there is alos the possibility to specify function calls
16:13:01 aaime: like generalize(the_geom, 250)
16:13:11 aaime: (assuming 250m generalizatio distance, that is)
16:13:31 aaime: so in our case we'd read the data as-is from the datastore
16:13:31 jdeolive: thinks the tech discussion shoudl take place in a separate agenda item
16:13:41 aaime: and generalize it on the fly
16:13:51 aaime: whatever
16:14:01 groldan: what's strange about that capability (ie, returning an on the fly computed property)
16:14:11 groldan: is that it basically produces a view on the feature type
16:14:21 simboss: aaime, I would like to ask you a few things afterward about this
16:14:21 groldan: hence the schema should be different?
16:14:31 jdeolive: so in terms of changes we want to propose to the road map
16:14:41 groldan: I admit I should have to go through the spec in more detail but..
16:14:51 jdeolive: gives up
16:15:01 aaime: (please)
16:15:01 groldan: gml wise, you can easily get an output that does not conform to the DescribeFeatureType schema?
16:15:21 aaime: since this topic is indeed large, can we discuss it next?
16:15:31 aaime: back on topic, roadmap changes
16:15:31 aaime: (and updates)
16:15:41 groldan: jdeolive: didn't read your previous concern about the tech discussion.. do not give up
16:15:51 aaime: so for 1.7.4 we have on the plate
16:16:01 aaime: as new features
16:16:01 aaime: 1) GetFeatureInfo customizable radius
16:16:21 aaime: 2) support for functions in WFS (that we'll discuss later)
16:16:31 aaime: 3) wms decorations?
16:16:51 simboss: 4)improved imagemosaic
16:17:01 aaime: 5) support for restricting the query envelope in kml when tilted
16:17:11 jdeolive: and possibly the new feature from RR, but we can wait until thati s officially brought up
16:17:11 aaime: (when the view is titled)
16:17:41 jdeolive: so in terms of proposed changes:
16:17:51 jdeolive: 0) propose allowing new features on 1.7.x
16:18:01 jdeolive: and then propose above new features
16:18:31 jdeolive: does anyone else have anything of note for the road map?
16:18:31 groldan: not here
16:18:41 jdeolive: what about releases...
16:18:41 aaime: I'm +1 to allow new features that have backing and sponsoring
16:19:01 aaime: and are relatively small
16:19:01 jdeolive: lets leave voting for the mail list to give others a chance to chime in
16:19:11 aaime: ok
16:19:21 jdeolive: so in terms of release dates
16:19:21 jdeolive: is checking the road map
16:19:41 jdeolive: 1.7.4 is scheduled for April 15
16:19:51 jdeolive: 2.0-alpha2 is unscheduled
16:20:11 jdeolive: given that these new features will be coming down for 1.7.4, April 15 seems unrealistic
16:20:31 aaime: hmmm... not sure
16:20:41 aaime: I can do the getfeatureinfo one in a couple of days
16:20:41 jdeolive: and do we want to throw a date on 2.0-alpha2? maybe 1-2 weeks from now?
16:20:51 aaime: imagemosaic improvemetns is pretty far along too
16:21:21 jdeolive: ok, so maybe 15th is not unreasonable
16:21:31 jdeolive: we could push it back to the following monday, 20th
16:21:41 jdeolive: that should probably be fine?
16:22:21 simboss: imagemosaic is not too far,
16:22:41 simboss: it is just that I keep on getting distracted from other (better paid) work .D
16:22:51 aaime: simboss... not too far? you're committing to a stable branch for that work?
16:23:41 aaime: so you planned to revert the changes in case the release date was the 15th?
16:23:41 simboss: yeah
16:23:51 simboss: it is not a big deal
16:23:51 aaime: ah, ok
16:23:51 simboss: I need this for 1.7
16:24:11 simboss: and I don't want to keep 3 repos for the same stuff
16:24:21 simboss: if I won't make in time
16:24:31 simboss: I will revert the changes
16:24:41 jdeolive: so any objections to the 20th?
16:24:41 aaime: not from me
16:24:51 simboss: not here
16:25:01 groldan: wouldn't be better to develop on trunk and when its done backport to 1.7.x/2.5.x?
16:25:11 groldan: if you don't get it on time no need to roll back anyway
16:25:21 jdeolive: ok, and for alpha2
16:25:31 jdeolive: 1-2 weeks acceptable?
16:25:51 aaime: I believe so, yeah
16:26:11 jdeolive: ok, sounds good
16:26:21 jdeolive: unless anyone has anything else for the road map, we can move on
16:26:21 simboss: +0 here, I have not had time yet to play with trunk
16:26:31 simboss: yet
16:27:11 dwins: jdeolive: that depends on whether you want to lump the decorations discussion into roadmapping or not
16:27:31 aaime: wms decoration -> point 4)?
16:27:51 jdeolive: dwins: up to you? if you think it is relevant for the planning, then sure
16:28:11 aaime: (provided it does not turn into a big discussion like generalisation)
16:28:31 dwins: it shouldn't affect the timeline, I wouldn't even need the release date to be pushed back
16:28:41 dwins: so i'm happy leaving it to the end
16:28:41 jdeolive: sounds good
16:29:01 jdeolive: then moving on
*16:29:01 jdeolive: 2) wfs generalization*
16:29:11 aaime: resuming the discussion, two points were made
16:29:31 aaime: Gabriel noticed how calling a function can effectively change the returned schema
16:29:31 aaime: and ... he's right
16:29:41 aaime: I was looking into the spec and functions are named only in the schemas
16:29:51 aaime: but I don't see trace of them in the prose
16:30:01 aaime: weird?
16:30:11 aaime: Wondering if adding a &resolution=xxx wouldn't be better
16:30:41 groldan: if the only thing you want is generalization I guess a vendor parameter is the way to go
16:30:41 simboss: I agree
16:31:01 groldan: also, it provides for servers that don't support the vendor parameter nor functions still to work
16:31:01 aaime: does anyone remember the Vretanos repo with the wfs 2.0 schemas?
16:31:01 simboss: ( andrea I want to ask you a few question at the end about real functions, like count, sum ,etc.)
16:31:21 aaime: (simboss, later then)
16:31:41 groldan: yeah, aggregating functions may be a completely separate issue I guess
16:32:31 groldan: may be something on the lines of the "registered functions" support the app-schema guys are working on
16:32:51 aaime: (the topic is "generalisation")
16:33:11 groldan: that's the point, aggregating functions is a separate topic is what I'm saying
16:33:41 aaime: jdeolive, I'm not seeing functions in wfs 2.0 either (
16:33:51 aaime: so it seems the vendor param route is better
16:35:11 aaime: going back to "how" generalise
16:35:21 aaime: mcrmcr needs to leverage a pre-made pyramid
16:35:31 jdeolive: seems the query type is now mostly defined externally?
16:35:31 jdeolive: how recent is that schema?
16:35:31 aaime: jdeolive, right
16:35:51 aaime: feb 2009
16:36:01 aaime: we just want to reduce the amount of generated gml that a javascript client will eat
16:36:01 aaime: without having to move a finger on datastores
16:36:21 aaime: this we can get by adding a generalizing feature collection in the GetFeature code
16:36:31 aaime: pretty simple, localized change
16:36:41 aaime: that involves only WFS
16:37:01 aaime: that's why we're pushing it for 1.7.x
16:37:11 aaime: what mcrmcr wants to do involves wfs and wms
16:37:21 simboss: there is fuel for hitting wms in your work aaime?
16:37:31 jdeolive: i thought we already did something similar for wms?
16:37:41 aaime: simboss, not at all
16:37:51 simboss: k
16:37:51 aaime: jdeolive, mcrmcr wants to use
16:38:01 aaime: a pre-decimated set of geometries
16:38:01 aaime: a pyramid
16:38:11 groldan: mcmcr has his own DataStore implementation right? so a hint would be enough for him?
16:38:11 aaime: wms generalizes on the fly instead
16:38:21 aaime: groldan, see the discussion on the ml
16:38:31 aaime: a single hint works only in a subset of case
16:38:41 aaime: and I'm ok about addin git
16:38:41 jdeolive: i see
16:38:41 aaime: also
16:38:51 aaime: mcrmcr mentions a generalizing feature source
16:39:01 aaime: whilst geoserver can only plug new datastores
16:39:01 jdeolive: which is why mcrmcr wants to only pass down the scale denom?
16:39:21 aaime: jdeolive, sorry?
16:39:51 jdeolive: aaime: in his first mail, i thought mcrmcr mentioned he wanted to pass down a scale parameter as a query hint
16:40:01 aaime: yes, and I explained why that may not work in the general case
16:40:11 jdeolive: ok, missed that
16:40:11 jdeolive: sorry
16:40:11 aaime: but it's really a matter of protocol
16:40:21 simboss: since mcmcr is around he might explain himself what he wants to do
16:40:21 simboss: no?
16:40:41 aaime: we can just say that the hint will be passed down in a restricted number of cases
16:40:41 mcrmcr: I had this decimations problem with an applet (wfs) and wms clients and I did my on implementation for decimation, but this is not the right way
16:41:21 aaime: the question is, do you already know the current generalization in wms does not give you enough speed?
16:41:21 CIA-16: 03tenzochris * r11927 10/branches/1.7.x/doc/user/source/_templates/layout.html: Fix GeoServer Sphinx theme to work with Sphinx 0.6
16:41:51 mcrmcr: the idea is to have pre decimated vector data and an intelligent feature source which chooses the right generalisation dependend on the scale
16:42:11 aaime: wfs wise, just adding it at the end of the chain (just before encoding) will give you a very significant speedup
16:42:11 simboss: like we do far raster pyramids?
16:42:21 mcrmcr: yes, thats it
16:42:31 simboss: nice
16:42:41 simboss: aaime's goal is to reduce GML
16:43:01 simboss: while mcrmcr goal is to speed things up from the ground up
16:43:11 aaime: yep, much more radical approach
16:43:11 mcrmcr: I had a perfomance boosts with this architecture
16:43:11 aaime: takes a lot more time to develop
16:43:21 aaime: no question about that
16:43:21 simboss: yeah, I see
16:43:21 simboss: let's put i this way
16:43:31 simboss: what aaime wants to do
16:43:41 simboss: is more straightforward
16:43:51 simboss: yeah exact aaime
16:44:11 aaime: Hmmm... we're late in the meeting and we have other stuff to discuss
16:44:31 aaime: can I recap adn allow other poeple to discuss points 3 and 4 in the hours alloted for the meeting?
16:44:31 jdeolive: i guess the downside in this approach is that the pyramid is useless?
16:44:41 aaime: we can move details to the ml
16:45:01 aaime: jdeolive, we can have both, if the feature source supports generalization we add the hint
16:45:11 aaime: otherwise we do it in memory
16:45:21 aaime: there is no conflict, just a choice
16:45:21 mcrmcr: yep
16:46:01 aaime: So to recap:
16:46:01 groldan: mcrmcr: since you're hitting the server from applets and js code, are you sure even if you have the pre computed generalized geometries, the bottleneck will not be the network bandwidth?
16:46:11 jdeolive: ok... i thought you were arguing against the hint.. saying that it did not work in every case
16:46:21 jdeolive: anyways, i will shut up now
16:46:31 aaime: To recap:
16:46:31 aaime: - a vendor param seems the way to go
16:46:51 aaime: - mcrmcr case can be handled by a query hint provided we restrict it to a small but common context
16:47:11 aaime: - our case can live there in wfs too, and be an alternative when the feature source cannot generalize itself
16:47:11 mcrmcr: no, since it is my applet I compress anything at the server and decompress at client
16:47:11 aaime: All good?
16:47:41 aaime: (mcrmcr, geoserver does transparent gzip compression of gml since quite a bit of time)
16:47:51 groldan: - I'm wondering whether mcrmcr case is still relevant provided the vendor param is already in place
16:48:01 aaime: (provided the client supports http 1.1 compression)
16:48:11 aaime: groldan, why? We can either:
16:48:31 aaime: - turn the vendor param into a query hint if the feature source supports it
16:48:41 aaime: - wrap the resultin feature collection into a generalizing one otherwise
16:49:01 aaime: (same goes for wms, we pass down the hint, if the datastore can be more efficient there, more power to it)
16:49:21 groldan: I would adivse mcrmcr to first try how the vendor parameter performs over a normal datastore (generalizing on the wfs) first
16:49:31 aaime: (10 minutes to the end of the meeting, still 2 points to discuss)
16:49:41 groldan: may be there's no need for the pre computed generalizations?
16:49:41 mcrmcr: aimme, good idea, I want to avoid to load a lot of vector data from a remote database
16:50:21 groldan: he'll get generalized, gzip compressed geometries without all the work in developing such FeatureSource
16:50:31 aaime: groldan, that's why I asked about the wms case
16:50:41 aaime: it's already doing the generalization in memory
16:50:41 groldan: and may find the perf gain is not such cause the bottleneck is the net
16:51:01 aaime: if mcrmcr found the cpu/communcation with the database is the bottleneck then the radical approach is probably needed
16:51:21 groldan: right, so if he already did that assessment I'll shut up
16:53:21 aaime: next topic?
*16:53:31 aaime: 3) sphinx docs update/plan*
16:53:31 dwins: bmmpxf: ?
16:53:41 bmmpxf: hi
16:54:01 bmmpxf: So, the documentation is coming along quite nicely, and I'd like to discuss a plan for moving over to the new system...
16:54:21 bmmpxf:
16:54:21 sigq: Title: Home GeoServer v1.7 documentation (at
16:54:51 bmmpxf: I've mainly been working on the User Guide (due to comfort level), but I'd like to start leveraging the benefit of this big ol' open source community thing.
16:55:31 bmmpxf: Does anyone have any thoughts on how this should be accomplished?
16:55:51 bmmpxf: I know we're going to need to deprecate GEOSDOC at some point, I was thinking with a global header.
16:56:01 bmmpxf: And obviously a note to the mailing lists won't go amiss.
16:56:11 simboss: not sure I am getting your point
16:57:51 bmmpxf: isn't sure how to explain himself
16:58:11 aaime: (back)
16:58:21 dwins: bmmpxf: seems like you have two questions
16:58:31 dwins: a) how to deprecate the old wiki
16:58:41 dwins: b) how to encourage users to contribute to the sphinx docs
16:58:51 bmmpxf: thanks dwins
16:58:51 bmmpxf: right
16:59:01 aaime: b) seems a hard proposition to me?
16:59:11 aaime: users barely contributed to the wiki to start with?
16:59:21 bmmpxf: Also, there is the outstanding question of packaging the docs for releases. I'm working with tenzochris on this now as we speak.
16:59:21 dwins: aaime++
16:59:31 aaime: wants PDF
16:59:41 dwins: what is the question about those?
16:59:41 aaime: dreams of dead tree format too
17:00:01 bmmpxf: aaime: Regarding b), I'm not so worried about everyone contributing. I just want to make sure that WE (meaning those who normally write docs) write them in the new place.
17:00:11 bmmpxf: aaime: PDF is coming.
17:01:01 bmmpxf: At the very least, I'd like people who care about GeoServer docs to take a look at what is currently in Sphinx, and make comments/suggestions.
17:01:21 aaime: generally speaking, how do you feel the new docs cover the old?
17:01:41 groldan: wants eclipse help like docs. Searchable, bookmarkable, throwable in a new window/tab from the admin ui
17:01:41 groldan:
17:02:01 bmmpxf: Well, I can only speak on the User Guide, but I'd say we're at 80%. That is a ballpark figure.
17:02:01 aaime: the html complied for of shinx is actually somewhat close I believe?
17:02:11 aaime: not sure about search thought
17:02:21 aaime: (close to a eclipse like thing)
17:02:31 dwins: it has search support, but you need php on the web server
17:02:31 dwins: dunno if we're using it
17:02:41 groldan: yeah it'd be awesome to open the user guide pages right from inside the ui
17:02:51 bmmpxf: search is something I haven't looked into, but I can do that
17:03:11 aaime: but the html is definitely packgeable with geoserevr... just not sure we want that?
17:03:21 groldan: (not to tell to link ui anchors to help topics...)
17:03:21 aaime: I mean, we have an online version of it anyways?
17:03:51 bmmpxf: aaime: Packaging HTML is way less optimal, we just need to find a way to customize the PDF output.
17:03:51 dwins: it'd be a lot tougher to have the search work within geoserver
17:04:21 aaime: I would be +1 on having a downloadable pdf and an online html (on our servers)
17:04:21 bmmpxf: So, aside from search and PDF...
17:04:31 bmmpxf: aaime: That's totally the plan
17:04:41 bmmpxf: Do we want to move the current location of the online HTML?
17:05:01 bmmpxf: Right now it's at gridlock...will that be confusing for users?
17:05:21 aaime: it may be
17:05:31 aaime: do we want to have live, daily docs published
17:05:41 aaime: or to publish the docs of the latest releases
17:05:51 dwins: the latter seems preferable to me
17:05:51 aaime: and link to snapshots of the specific per-version docs?
17:06:01 jdeolive: gridlock was just a temp place to buld the docs nightly
17:06:11 jdeolive: i think the plan was to put them up eventually at or something
17:06:11 bmmpxf: jdeolive: Of course
17:06:21 bmmpxf: jdeolive: ahhh
17:06:41 bmmpxf: jdeolive: I propose we start that move then.
17:06:41 dwins: georj: log about
17:06:41 aaime: So, how does one access the docs relative to version x.y.z? PDF download only?
17:06:41 sigq: Title: #geoserver log for all dates (docs) (at
17:07:01 bmmpxf: aaime: Oh good question.
17:07:11 bmmpxf: Can they be tagged with the release?
17:07:21 dwins:
17:07:21 sigq: Title: #geoserver log for 2008-12-16 (at
17:07:31 dwins: bmmpxf, all: previous convo we had about this
17:07:31 aaime: bmmpxf, imho they will
17:07:41 jdeolive: well we will be publishing doc artifacts for each release no?
17:07:51 bmmpxf: jdeolive: The artifact will hopefully be a PDF....but there should also be a hosted HTML version.
17:08:01 bmmpxf: dwins: thanks for the link.
17:08:01 aaime: jdoelive, the artifacts -> docs in both pdf and zipped html?
17:08:11 jdeolive: i was thinking the doc artifacts would be both
17:08:11 aaime: we can have both no?
17:08:21 bmmpxf: Sure
17:08:31 bmmpxf: Do we want the tagged docs to be accessibel online?
17:08:41 bmmpxf: (as well as on the DL page)
17:08:41 bmmpxf: thinks we would
17:08:41 jdeolive: well they will be tagged like the source tree is
17:08:51 jdeolive: i guess you mean built
17:09:01 bmmpxf: yup
17:09:01 jdeolive: i guess we could yeah
17:09:11 aaime: having them in the downloads and online would be extra work for the release thought
17:09:31 jdeolive: i don't see a huge win to doing that across minor release version changes... i mean you pretty much always want the latest
17:09:41 jdeolive: major versions yes
17:10:11 bmmpxf: jdeolive: You don't always want the latest docs. Case in point: The KML reflector syntax changed in 1.7.1, and if you had the latest docs, you might get confused...
17:10:31 jdeolive: fair enough
17:11:01 bmmpxf: I think although it's extra work, it's a big win for us to have tagged built online HTML as well as packaged. Should be a way to automate it, no?
17:11:11 jdeolive: yeah, i don't think it is that much work
17:11:31 jdeolive: i mean the release artifacts should be baked right into the release process
17:11:41 jdeolive: might be one more step to build the docs
17:12:11 jdeolive: and then "Tagging" them online would be just unzipping them on the "doc server"
17:12:11 bmmpxf: jdeolive: And it's a big win for users.
17:12:21 bmmpxf: jdeolive: So we could set up links from gridlock to etc...
17:12:21 jdeolive: so yeah, i am fine with that
17:13:01 bmmpxf: jdeolive: Can I ask you to be the point person on redirects?
17:13:01 jdeolive: sure, or just deploy them on the doc server directly , under a "latest" tag or something
17:13:21 bmmpxf: sure
17:13:21 jdeolive: not sure what you mean?
17:14:11 bmmpxf: jdeolive: The docs are on gridlock, and we talked about putting them on I'd feel more comfortable if you or someone else took care of the switch/change/link/redirect.
17:14:31 bmmpxf: I'd like to leave it to you how you would like to architect it, that's all.
17:14:41 jdeolive: bmmpxf: gotcha, sure i can handle that
17:14:51 jdeolive: is already set up?
17:14:51 bmmpxf: jdeolive: Thanks!
17:15:11 bmmpxf: It doesn't appear to be, no
17:15:21 bmmpxf: at least not in the place where I expected it to be...
17:15:31 jdeolive: ok... could you file a ticket for arne to do that, and once done i will be happy to migrate
17:15:41 bmmpxf: jdeolive: Got it. Will do.
17:16:01 bmmpxf: Also, I propose that we start updating the docs on Sphinx as part of our SOP. I'll send out a message to the lists to promote this. Is this acceptable to everyone?
17:16:41 bmmpxf: hears crickets.
17:16:41 jdeolive: SOP = standard operating procedure ?
17:16:51 jdeolive:
17:16:51 bmmpxf: jdeolive: Sorry, yeah
17:17:01 jdeolive: yeah, start a mailing list thread about it
17:17:11 bmmpxf: right, that makes sense
17:17:51 jdeolive: cool
17:18:01 bmmpxf: anything else that people want to bring up with this? I'm out of items.
17:18:11 bmmpxf: I think we can move to the mailing list thread.
17:18:31 bmmpxf: thinks he killed the room.
17:18:51 dwins: well it's 20 minutes past the scheduled end of the meeting
*17:20:21 dwins: would it be inappropriate to go on to 4) ?*
17:20:51 bmmpxf: dwins: I guess that depends on who else is still around...
17:21:01 jdeolive: dwins: i think there are still some folks lingering so sure
17:21:11 jdeolive: people that left can just read the logs
17:21:11 dwins:
17:21:31 simboss: I am alive and kicking
17:21:31 simboss:
17:21:41 dwins: my basic question is how to move forward with this extension
17:22:11 simboss: I should stop kicking though, I just kicked my workstation out of the office
17:22:31 jdeolive: dwins: so you think it is ready for prime time?
17:22:41 aaime: dwins, I'm still around ifyou need me
17:23:01 jdeolive: have you started on a gsip?
17:23:01 dwins: i know it's sufficient for my needs, but it might make more sense as an extension for a while so people won't expect backward compatibility if it turns out things need to change
17:23:11 dwins: i do not have a gsip
17:23:41 jdeolive: has anyone reviewed the code formally?
17:23:41 jdeolive: aaime: ?
17:23:41 dwins: is all the documentation i have written
17:23:41 aaime: dwins, I still owe you are review
17:23:41 sigq: Title: WMS Decorations - GeoServer (at
17:23:51 jdeolive: that answers it
17:24:51 jdeolive: while I will try to review it, a proposal is good because it forces people to review or they do not get a say
17:25:21 simboss: (nice work dwins)
17:27:11 dwins: okay. i will write up a GSIP. does anyone feel strongly about extension vs. core on this issue?
17:27:31 simboss: I'd say core
17:27:41 simboss: as long as it does not kick in by default
17:27:51 dwins: i would prefer to bring it into core as that will let me avoid the 'png-decorated' pseudo-format
17:28:01 aaime: core here too
17:28:01 simboss: many people would like to see this in place
17:28:11 simboss: (I am one of them )
17:28:11 jdeolive: interface in core, and perhaps some of the well known stuff
17:28:11 dwins: okay. I'll write up a GSIP and mail the list
17:28:21 jdeolive: with any specialized decorations in extension?
17:28:31 dwins: makes sense
17:28:31 simboss: +1
17:28:41 aaime: what are the "specialized extensions"?
17:28:51 dwins: although the three decorations described there are the only three I am aware of
17:28:51 aaime: (like, do we have any today?)
17:29:01 aaime: imho title, legend, static image, scale and north arrow should be core
17:29:21 aaime: if you need anything else then yeah, probably good idea to put it as an extension?
17:29:31 simboss: I agree
17:29:31 dwins: sounds good to me.
17:29:41 jdeolive: aaime: i did not have any in mind, just stating that if there were specialized ones they should be extensions
17:30:01 dwins: okay, I think this is covered well enough that we can finish it up on the ml
17:30:01 aaime: I feared you considered north arrow or scale the "specialized ones"
17:30:11 jdeolive: no
17:30:11 jdeolive: of course not
17:30:21 dwins: thanks for the input all (and staying late)
17:30:41 aaime: so ok, that's it?
17:30:51 simboss: eventually
17:30:51 dwins: i'm done

Posted at 31 Mar @ 2:48 PM by Mike Pumphrey | 0 comments
Last changed: Mar 17, 2009 14:28 by Andrea Aime

0) whats up
1) 1.7.3a
2) meeting logging
3) road map update
4) GSIP 34

jdeolive: 0) what's up
***jdeolive is working on wicket ui stuff and trying to get nightly builds going in hudson
rpenate ha abbandonato la stanza (quit: Client Quit).
***groldan is setting up dev env to keep fixing massgis bugs, old laptop finally went bye bye
***aaime hitting his head on the Wicket catalog pages
***bmmpxf is writing docs
ahocevar [n=andreas@] è entrato nella stanza.
jdeolive: ok, next topic?
jdeolive: 1) 1.7.3a
jdeolive: so with some testing from iwillig (thanks) we fixed a pretty big bug that came up with 1.7.3
sonicx ha abbandonato la stanza (quit: Read error: 104 (Connection reset by peer)).
CIA-16: jdeolive * r11712 /trunk/src/community/web2/ (2 files in 2 dirs): GEOS-2758, have WicketTestApplication support i18n resource lookups
bmmpxf: this was the can't edit the featuretype you've added bug.
jdeolive: so i patched the release, updated the artifacts, and announced the patched release on teh blog and mailing lists
jdeolive: bmmpxf: yup
bmmpxf: jdeolive: I think only the SRC and perhaps the changelog needs to be updated.
***bmmpxf assumes the extensions should stay the same
rpenate [n=scrollie@] è entrato nella stanza.
jdeolive: yeah, i did not both re-uploading the source, i probably should though
jdeolive: both = bother
bmmpxf: Yeah, that would be good. Thanks. So that's I guess all we need to do then.
bmmpxf: is there a really quick explanation on why that bug was there?
bmmpxf: (is just curious)
jdeolive: a recent chagne i made to support the schema.xml functionality from 1.6.x
jdeolive: that and the the fact that the struts ui is unmaintainable
bmmpxf: nice
jdeolive: shall we move on?
bmmpxf: ok
jdeolive: 2) road map update
bmmpxf: ahem
bmmpxf: 2) meeting logging
jdeolive: oops
jdeolive: sorry
jdeolive: 2) meeting logging
bmmpxf: this can be quick...
bmmpxf: I don't mind posting logs, but I'd like to find a way to automate the process as much as possible...
bmmpxf: Right now, my client logs with a bunch of styling characters, which I have to manually take out, which takes a while.
bmmpxf: Then there's Confluence wiki markup, which we all know is fun.
bmmpxf: dwins, jdeolive: You started to mention ideas?
jdeolive: yeah, i think getting georj to do it is probably the best option
jdeolive: i looked at phenny a while ago, adding modules it pretty easy
jdeolive: just a matter of getting someone to do it... (in spare time)
CIA-16: aaime * r11713 /trunk/src/community/web2/core/src/main/java/org/geoserver/web/acegi/ Removed unused imports
bmmpxf: Would be neat if it could embed the confluence tags so I could just create new news, then hit save. (If we can't get georj to post it him/herself!)
cabraham [n=cabraham@] è entrato nella stanza.
bmmpxf: Who is the georj expert around here?
dwins: the confluence tags are definitely doable; would have to do some research wrt the auto-posting
jdeolive: tschuab
bmmpxf: autoposting is less important
dwins: tschaub set up georj, but phenny's not that tough to hack on
bmmpxf: Well, what do you all suggest? (Buzzword: Deliverables)
***bmmpxf doesn't know much about the specific methods, so can't suggest much offhand
dwins: pick a person to make the extension
aaime: I have the impression the idea is nice, but the resources to carry it on are not there?
bmmpxf: okay, so howzabout this. I'll email tschaub, and ask if georj has this capability without much fuss
cabraham ha abbandonato la stanza (quit: Read error: 104 (Connection reset by peer)).
cabraham [n=cabraham@] è entrato nella stanza.
bmmpxf: If yes, then I'll see if he and I can get something going. If not, we'll let the matter drop for now, unless someone else volunteers.
CIA-16: jdeolive * r11714 /trunk/src/community/web2/app/src/test/java/org/geoserver/wicket/test/ cleaned up imports
dwins: we would have to write code for this to happen (or find a different irc bot)
SandGorgon ha abbandonato la stanza (quit: Read error: 110 (Connection timed out)).
bmmpxf: dwins: Okay, makes sense
bmmpxf: or we could see what other orgs do in IRC meetings.
bmmpxf: I can look around
***bmmpxf bangs gavel
jdeolive: 3) road map update
sigq: Title: Roadmap - GeoServer (at
jdeolive: i have still not gone through 1.7.4 and moved features back to 2.x
***jdeolive slaps his own wrist
jdeolive: but assuming i will do that today there is probably not much to report on 1.7.4?
jdeolive: we will basically handle bugs as they come up?
aaime: I guess so, yeah
aaime: we should maybe cheer up the release with some new extension module? :-p
jdeolive: how about validation ?
aaime: lol
jdeolive: haha
***dwins supposes it's okay if not every release is world-shattering
aaime: I was thinking xslt, if I want to show up at Bolsena face up
cabraham ha abbandonato la stanza (quit: ).
bmmpxf: How about Phase 1 plus a call to arms about our new doc system?
jdeolive: aaime: i actually experimented with an xslt extension
***aaime 's dark passenger is screaming "new features, new features!"
jdeolive: it was actually quite trivial
jdeolive: the transformation that is
jdeolive: how to hook it up... another matter
jdeolive: the code i experimented with was put directly in GML3OutputFormat
aaime: yeah, where to hook up, how to configrue, which params, that's the hard part
jdeolive: yup
aaime: a user contributed a complete patch one year ago afaik
aaime: pity to let it rot
jdeolive: i could never find it...
jdeolive: however, we digress
jdeolive: for 2.0-beta
aaime: (here:
groldan ha abbandonato la stanza (quit: Read error: 113 (No route to host)).
aaime: please go on, did not want to interrupt
jdeolive: no problem, thanks for the link
jdeolive: do we want to report on progress moving toward 2.0-beta, which is really just
sigq: Title: GEOS-2134 Bring the new UI to feature parity with the old UI - (at
jdeolive ha scelto come argomento: 0) whats up 1) 1.7.3a 2) meeting logging 3) road map update 4) GSIP 34
aaime: Hum... sure... replacign the data tree with table based stuff is almost done, but the UI is still not really working that well
aaime: quite easy to make it throw exceptions ...
aaime: after that, we have to fix the layer confgiruation pages, there is quite a bit of work to be done there
aaime: that jira has a tree of stuff to be fixed underneath
jdeolive: ok... i guess we are not really at a point to put a date on 2.0-beta?
aaime: hmmm... given I'm not seeing the light at the end of the tunnel with this, and we lack also the persistence
aaime: probably not?
jdeolive: ok... persistence brings us to 4)
grobda24-ghost [] è entrato nella stanza.
jdeolive: ok to move on?
aaime: one thing
aaime: wouldn't it be better to name it alpha-something?
aaime: we had huge changes recently
jdeolive: yeah, i just had the same though
jdeolive: we should maybe release alpha2
aaime: jdeolive++
jdeolive: and could probably release sooner that way
jdeolive: so yeah, i like that
aaime: k
jdeolive: cool
jdeolive: 4) GSIP 34
jdeolive: so the feedback thread has sort of ground to halt
jdeolive: i updated the proposal as per feedback from aaime
aaime: I guess if you just changed the persistence to save the layers in a default map
aaime: most other negative feedbacks would be accounted for?
jdeolive: yeah... but i am quite hesitant to do that at this point
jdeolive: i don't see much point in increasing scope at this point when the upgrade path is clean
aaime: the main drawback I see is, what if we don't do map the way we're thinking about today?
aaime: that is, the change request assumes we're going to follow that plan
aaime: what if on deeper analysis we decide on something different?
jdeolive: well there are lots of hypothetical arguments that could be made
aaime: What I'm trying to say is that the change request is based on an assumption that might be not hold true
jdeolive: but incroporating maps into the picture now means we wont have peristence on trunk for a while longer
aaime: a transparent switcher once the maps are in the picture works for me
jdeolive: unless there is a concrete argument against the current idea of how to implement arguments i think we should discount that argument
jdeolive: i mean... i could bring up that tommorrow we may not choose to use xstream
jdeolive: which means the proposal is completely invalid
cabraham [n=cabraham@] è entrato nella stanza.
***aaime was actually trying to argument against introducing maps in the persistence now
jdeolive: oh
aaime: (since we don't know what maps will really look like)
jdeolive: ahh... ok, so we are making teh same argument
jdeolive: i should also add that the way the changes are implemented they are easy to turn off
jdeolive: so we could revert to what is currently on trunk if things go south
aaime: ok
aaime: btw, what kind of coverter do we have now?
aaime: from old data dir to the new one?
aaime: asking since devs have a mixed bag of data dirs
aaime: like, if we have tried to save with the current xstram support, we should just ditch the data dir, right?
jdeolive: well the converter is sort of baked directly into GeoServerLoader
jdeolive: not sure i understand
jdeolive: sorry
aaime: I have data dirs around in which I have tried to save with the current trunk persister
aaime: which does not work
aaime: I can just say bye bye to them or what?
aaime: well, never mind
jdeolive: ahh, yeah, the current trunk persister does not work
aaime: the important thing is that the user can transparently upgrade from 1.7.x-like to gsip 34 up
jdeolive: if i commit my changes the current trunk persister will go bye bye
jdeolive: as there is no real way to "persist at a button"
cabraham ha abbandonato la stanza (quit: Read error: 104 (Connection reset by peer)).
jdeolive: changes will be made automatically as pages are submitted
juliatorti ha abbandonato la stanza (quit: ).
aaime: ok
cabraham [n=cabraham@] è entrato nella stanza.
jdeolive: it is too bad groldan did not stick around since he had an opinion on this
aaime: Well, in the end it's a vote count no?
aaime: Two positives I guess are not enough?
jdeolive: i just meant that since we are discussing the issues it would be nice to have him around
jdeolive: but just so i am clear
jdeolive: you and i are in agreement at this point?
aaime: yeah
jdeolive: cool
aaime: I'm just saying you should call a vote for the gsip as is
aaime: and see if any -1 pops up
jdeolive: cool
jdeolive: so yeah, i will call for a vote on the mailing list

Posted at 17 Mar @ 2:23 PM by Andrea Aime | 0 comments
Last changed: Mar 10, 2009 15:41 by Mike Pumphrey

0) What's up
1) Roadmap update

[12:05] <bmmpxf> Dare I say...meeting?
[12:09] <aaime> Uh, right?
[12:10] <aaime> that's a meet-o-day for me
[12:10] * bmmpxf is wishing that groldan had haded off info about SDE rasters before he want awol...
[12:11] <aaime> Okey, so, topics?
[12:12] <bmmpxf> Looks like it's an empty room today
[12:12] * aaime changes topic to '0) What's up 1) Roadmap update#'
[12:12] <aaime> well, we have aaime, jdeolive, bmmpxf, cholmes, simboss
[12:12] <aaime> (apologies if I forgot anybody)
[12:14] <simboss> anybody ping (just in case we forgot him)
[12:14] <aaime> omg
[12:14] <aaime> Shall we get started?
[12:14] <aaime> 0) What's up?
[12:15] * aaime is looking into Oracle bugs, wicket ui
[12:15] <aaime> also, aaime.brain("scala" -> (scala.syntax() ::: scala.library())
[12:15] * bmmpxf is writing a post announcing 1.7.3, and feverishly rewriting the GS user docs as he converts them to Sphinx
[12:15] <aaime>
[12:15] <bmmpxf> aaime: I so wish I understood that syntax
[12:16] <aaime> in more human terms, I'm trying to learn some SCALA programming
[12:16] <jdeolive> howdy
[12:17] * jdeolive is trying to organize wicket ui stuff after putting out the 1.7.3 release
[12:17] <simboss> simboss: writing proposals
[12:17] <aaime> as in, GSIPs?
[12:19] <aaime> ok, so let's move forward?
[12:19] <aaime> 1) Roadmap update
[12:19] <simboss> as in proposals-to-get-money-to-write-GSIPs
[12:19] <jdeolive>
[12:19] <sigq> Title: Roadmap - GeoServer (at
[12:20] <jdeolive> most of those features for 1.7.4 should be moved back to 2.x...
[12:20] <aaime> so, we have a number of new features there... and we agreed to only have bug fixes from now onwards?
[12:20] <jdeolive> if we are indeed following the only bug fix policy
[12:20] <aaime> in that case, shouldn't we change the filters to show which bugfixes we're going to do?
[12:20] <jdeolive> although, GEOS-1162, it sounds like simboss is just tightening up
[12:20] <aaime> or else, we don't need planning?
[12:20] <jdeolive> aaime: +1
[12:21] <jdeolive> lets move feautres that are not In Progress back
[12:21] <jdeolive> and change the filter to only high priority bug fixes
[12:21] <jdeolive> sound good?
[12:21] <aaime> ok... which is 2 out of 103
[12:21] <aaime>
[12:21] <sigq> Title: Issue Navigator - (at
[12:21] <simboss> jdeolive
[12:21] <simboss> actually
[12:22] <simboss> that's one is more or less closed
[12:22] <simboss> I will wait for more input from mike
[12:22] <simboss> but I am looking at this one
[12:22] <simboss>
[12:22] * dany_r ( has joined #geoserver
[12:22] <sigq> Title: GEOS-2520 Bring Imageio-ext extension to supported status - (at
[12:23] <simboss> daniele what's the status of this one
[12:23] <simboss>
[12:23] <simboss> ?
[12:23] <sigq> Title: GEOS-2520 Bring Imageio-ext extension to supported status - (at
[12:23] <aaime> actually, if we filter out just "bugs" we get only 36 items:
[12:23] <sigq> Title: Issue Navigator - (at
[12:23] <bmmpxf> simboss: Sorry, input about which? (The raster labels work?)
[12:23] <simboss>
[12:23] <dany_r> The description requires several binaries and installers. we have built binaries for windows 32, and linux 32.
[12:23] <dany_r> we can close the jira
[12:24] <jdeolive> cool
[12:24] <dany_r> aaime ping
[12:24] <bmmpxf> simboss, I pushed back on that since it looked like the fixes weren't part of the 1.7.3 release, so it could wait, but yes I will provide feedback.
[12:24] <dany_r> did you have any chance to test the linux 64 bindings?
[12:24] * fizzi (n=francesc@ has joined #geoserver
[12:25] <simboss> bmmpxf, cool, np, just looking for external feedback
[12:25] * cholmes ( Quit (No route to host#)
[12:25] <aaime> dany_r, were the instructions pages updated?
[12:26] <dany_r> aaime yes
[12:26] <aaime> excellent!
[12:26] <dany_r>
[12:26] <sigq> Title: ImageIO-ext GDAL extensions - GeoServer (at
[12:27] <dany_r> jdeolive, I guess that I can close the Jira adding a comment saying that windows 64 are future plan and linux 64 need to be better tested
[12:27] <dany_r> what do you think about it?
[12:27] <jdeolive> sounds good
[12:27] <aaime> doh, we suddendly have ecw support in geoserver it seems
[12:27] <jdeolive> i would close and link to any relevant future tasks
[12:27] <simboss> and mac as well
[12:27] <simboss>
[12:27] <aaime> I don't see that:
[12:27] <sigq> Title: imageio-ext: Documents & files: 1.0.1 (at
[12:28] <dany_r> aaime... what do you mean?
[12:28] <aaime> I guess you still have to upload it?
[12:28] <aaime> I don't see mac support simboss mentions
[12:28] <simboss> it's on the roadmap
[12:28] <simboss> future plans
[12:29] <aaime> aah
[12:29] <simboss> well, more a wish, unless someone pays for it
[12:29] <dany_r> aaime the binaries are uploaded... you just need to expand the release link
[12:29] <aaime> dany_r, when I saw simboss comment I thought I could find Mac binaries as well, that's all.
[12:30] <aaime> a misunderstanding
[12:30] <dany_r> ah.. ok-
[12:30] <dany_r> anyway I will add also the linux64 bindings.
[12:30] <dany_r> about ecw, I have built a separate windows installer
[12:30] <aaime> My Ubuntu 64bit notebook will appreciate greatly
[12:30] <dany_r> which doesn't provide ECW SDK.
[12:30] <dany_r> the user should manually download it.
[12:31] <dany_r> therefore geoserver isn't distributing ecw SDK
[12:31] <dany_r> using that installer.
[12:31] <aaime> well, the thing is, we're not distributing it
[12:31] <aaime> imageio-ext is
[12:31] * Bojan ( has left #geoserver
[12:32] <aaime> and it's the user that's combining them into a server side solution, so I guess that lives us clean
[12:32] <dany_r> if you download the complete installer you also get the ECW SDK but geoserver users should install the reduced one
[12:32] <dany_r> about linux, we haven't found time yet to enable ecw as a plugin since this will require porting some code from lasts gdal versions whilst we are linked to 1.4.4
[12:33] <aaime> dany_r, I would put a warning in the "descripton" field of the windows installer as well
[12:33] <aaime> "not meant to be used freely in server side apps, consult the ecw license for details"
[12:33] <dany_r> I can close the ECW jira stating that only Windows is actually available for geoserver
[12:34] <dany_r> ok aaime. I'll update the description
[12:34] <aaime> dany_r, whatever, just pay attention to always cite the license and not
[12:35] <aaime> make people thinkg there is any legal way to run geoserver + ecw without paying a server side licence to erdas
[12:35] <dany_r> anyway, the installer will show the license
[12:35] <dany_r> which you NEED to accept if you want to proceed using the ECW
[12:35] <dany_r> using --> installing
[12:35] <aaime> ah, click thru license, good one
[12:36] <dany_r> otherwise, you should disable ECW support and you can proceed with the installation (without ecw)
[12:37] <dany_r> ok. I have updated the description.
[12:37] <dany_r> I'm going to edit/close JIRA... ok?
[12:37] <aaime> ok
[12:37] <aaime> jdeolive, what else?
[12:38] * jdeolive is catching up
[12:38] <jdeolive> what else for road map stuff?
[12:38] <aaime> yep
[12:38] <aaime>
[12:38] <sigq> Title: Issue Navigator - (at
[12:38] <jdeolive> not sure there is much else interesting...
[12:38] <jdeolive> for 1.7.4 that is
[12:38] <aaime> there are 100+ issues scheudled for 1.7.4.but apparently only 30+ are marked as bugs
[12:39] <aaime> uh, well, make that 45
[12:39] <aaime> hmm... something is going on with those filters...
[12:40] <aaime> here:
[12:40] <sigq> Title: Issue Navigator - (at
[12:41] <dany_r> jdeolive... I have fixed that JIRA for 1.7.3
[12:41] <jdeolive> dany_r: cool, thanks
[12:42] <aaime> jdeolive, given there are only 2 critical bugfixes scheduled, what would you suggest? We look for all critical and outstanding issues and move them to 1.7.3, reschedule the existing ones and pump up the priorities or some the existing ones?
[12:42] <aaime> I guess my point is, only 3 issues for 1.7.4 seems too little?
[12:42] <jdeolive> well given that 1.7.3 is not even out yet i am sure that will chagne
[12:42] <jdeolive> i dont think we will have to actively go chasing bug reports
[12:43] <aaime> so, so "wait and see" policy
[12:43] <jdeolive> i would say so... i think with bug fix releases less planning is required
[12:43] <jdeolive> but that is just my opinion
[12:43] <aaime> "let the bugreports come to me" policy then (already heard a similar one, can't remember where thought ... )
[12:44] <jdeolive> not sure what that means but ok
[12:45] <aaime> it's a citation from the holy bible, Jesus Christ once said "let the children come to me"
[12:45] <jdeolive> ahhh...
[12:45] * jdeolive did not expect that from an athiest
[12:46] <simboss> looking at the roadmap
[12:46] <simboss> I might want to add
[12:46] <simboss> the capability
[12:46] <simboss> to build mosaics
[12:46] <simboss> transparentrly
[12:46] <simboss> I have the code ready
[12:47] <simboss> I just need to do some clean up + I need some help with testing
[12:47] <aaime> that would be against the "no new features" policy, but then again, we can't stop geotools lib from improving, can't we?
[12:48] <simboss> well, infact
[12:48] * vhamer ( Quit
[12:48] <simboss> I was thinking to bring that in for 2.0
[12:48] <aaime> works for me
[12:48] <simboss> it is a quite minor improvements codewise
[12:49] <aaime> a small step for Simone == a huge step for mosaic users
[12:49] <simboss> jdeolive
[12:49] <simboss> how can I add the jira to the roadmap
[12:49] <simboss> ?
[12:49] * aaime is in the mood for citations today
[12:49] <jdeolive> just add it as a new feature and it will show up
[12:50] <jdeolive> in jira
[12:50] <jdeolive> the road map is an active query of jira
[12:50] <jdeolive> it will alsoshow up if you add it as a task,improvment,or bug with high priority
[12:50] <jdeolive> (critical or blocker)
[12:52] <simboss> trying to link the relevant geotools bugs
[12:53] * pramsey_ (n=pramsey@ has joined #geoserver
[12:54] <grobda24> Hello. Can I get KML output from a Feature Type as a single file download ? Or would that be the same as a WMS request for -360, 360 to -90, 90 ?
[12:54] * pramsey (n=pramsey@ Quit (Remote closed the connection#)
[12:55] <rpenate> am i the only one that didn't know that GeoServer is also a Raytheon product?
[12:55] <simboss> added, cool -)
[13:00] <simboss> meeting's over?
[13:02] * vheurteaux ( Quit (Read error: 110 (Connection timed out)#)
[13:02] <aaime> simboss, I guess so?

Posted at 10 Mar @ 3:35 PM by Mike Pumphrey | 0 comments
Last changed: Mar 04, 2009 12:49 by Mike Pumphrey

0) whats up
1) road map update
2) 1.7.3 release

[12:05] <aaime> yep, meeting time
[12:05] <arneke> pong
[12:05] * jdeolive changes topic to '0) whats up 1) road map update 2) 1.7.3 release#'
[12:09] <jdeolive> shall we start
[12:09] <aaime> sure
[12:09] <bmmpxf> 0) what's up
[12:09] * bmmpxf is writing docs and trying to recall where geosearch was before he left the country
[12:10] * jdeolive has been hacking hard at the osgeo code sprint, now getting back to 1.7.3 release
[12:10] * aaime too
[12:11] <aaime> (aaime loves "me too")
[12:11] * groldan been hacking hard, having interesting meetings and getting back to real world where the build is broken
[12:11] <jdeolive> groldan?
[12:12] <jdeolive> cool
[12:12] * aaime feels odd chatting with jdoelive and groldan when he's sitting side by side with them
[12:12] <jdeolive> me too
[12:12] <jdeolive>
[12:12] <aaime> deh!
[12:12] <jdeolive> jokes
[12:13] <jdeolive> 1) road map update
[12:13] <groldan> well we just improved our communication channels for this week
[12:13] <jdeolive> so the road map has sort of missed an update last week due to most of the opengeo crew being "offline"
[12:13] <aaime>
[12:13] <sigq> Title: Roadmap - GeoServer (at
[12:13] <groldan> who goes first
[12:13] <jdeolive> thanks aaime
[12:14] <jdeolive> how about you groldan
[12:14] <groldan> ok
[12:14] <groldan>
[12:14] <sigq> Title: GEOS-2567 Support new ArcSDE raster formats - (at
[12:14] <groldan> ready to ship
[12:14] <jdeolive> nice
[12:14] <jdeolive> is that issue resolvable?
[12:14] <groldan> I'll close the issue right away though
[12:15] <jdeolive> nice
[12:15] <jdeolive> only other open issues are from myself and dwins
[12:15] <jdeolive> both of which are ready to move ovcer
[12:15] <jdeolive> so should probably be closed today
[12:16] <jdeolive> other issues are from geosolutions
[12:16] <jdeolive> but looks like will get get pushed back regardless
[12:16] <aaime> simboss?
[12:16] <jdeolive> ahh, i did not see simone in the channel
[12:16] <aaime> his computer his... him, dunno
[12:17] <aaime> maybe not?
[12:17] <jdeolive> ok, no problem
[12:17] <simboss> ciao guys
[12:17] <simboss> what's up
[12:17] <jdeolive> ciao simboss
[12:17] <simboss> ciao
[12:17] <jdeolive> we are just going through teh road map update
[12:17] <jdeolive> a couple of issues have geosolutions name on them and are unresolved
[12:18] <jdeolive> so hoping for a quick update
[12:18] <jdeolive> namely
[12:18] <jdeolive>
[12:18] <jdeolive> and
[12:18] <sigq> Title: GEOS-2520 Bring Imageio-ext extension to supported status - (at
[12:18] <jdeolive>
[12:18] <sigq> Title: GEOS-1162 Support custom raster legends via GetLegendGraphic - (at
[12:18] <simboss> let's grab meeting daniele for the first one
[12:18] <jdeolive> cool
[12:18] <simboss> it should be done though
[12:18] <simboss> about geos-1162
[12:18] <simboss> things have been implemented, but I would like to add some test + 1 more config options
[12:18] * dany_r ( has joined #geoserver
[12:19] <simboss> therefore I am not sure I would close it
[12:19] <simboss> dany_r what about
[12:19] <simboss>
[12:19] <sigq> Title: GEOS-2520 Bring Imageio-ext extension to supported status - (at
[12:19] <simboss> ?
[12:19] <dany_r> gt-imageio-ext-gdal has been moved to supported status. once everything is ok with license and the gt release
[12:20] <dany_r> the extension can be adopted.
[12:20] <simboss> we just need to add the license txt file to the release
[12:20] * aaime thinks we're getting old... only 56 issues solved for 1.7.3?
[12:20] <simboss> so that everyone is happy
[12:20] <sigq> Title: Browse Project - (at
[12:21] * doktoreas ( Quit (Read error: 145 (Connection timed out)#)
[12:21] <dany_r> I already have fixed geoserver extensions poms to leverage on gt-imageio-ext-gdal
[12:21] <dany_r> I only need to committ them
[12:21] <aaime> simboss, so the binary builds are good for at least win and lin?
[12:21] <aaime> (simboss or dany_r, whoever)
[12:22] <dany_r> Before the last release, jesse encountered some issues with the linux libs. I have redeployed them and now seems ok (no more dependencies from Netcdf)
[12:23] <dany_r> therefore they should be ok.
[12:23] <simboss> yep, right now everything should be fine
[12:23] <simboss> for all of them
[12:23] <simboss> next round we'll do window 64 bits and probably MAC
[12:23] <dany_r> the uDig guys did some testing on the libs.
[12:24] * aaime looks at jdeolive's black macbook and smiles
[12:24] <jdeolive> ok... so this is not something included in teh release tommorrow or next day
[12:24] <jdeolive> so push back?
[12:25] <simboss> no, wait, what do you mean?
[12:25] <jdeolive> ok, sorry, mis understanding
[12:25] <jdeolive> so simboss, for GEOS-1162, is this code committed?
[12:25] <simboss> it is ready, I just need to give you a license txt file
[12:25] <simboss> that lists all the licenses in geotools
[12:25] <simboss> yeah it is
[12:25] <simboss> we are already using it
[12:26] <simboss> but I would like to add a page for it and some basic tests
[12:26] <jdeolive> ok, so for that one i say we resolve it, and can we create a seperate issue for any other tasks you would like o perform
[12:26] <jdeolive> like more tests, etc...
[12:26] <simboss> the problem is how to test that a legend is actually the way I want it
[12:26] <simboss> as you wish
[12:26] <simboss> I am fine with that
[12:26] <jdeolive> ok cool
[12:26] <jdeolive> and for GEOS-2520, do you know how long it will take to build the license file?
[12:27] <dany_r> jdeolive will the imageioext extensions included in the release as standard or as external extension?
[12:27] <aaime> question, what about this page:
[12:27] <sigq> Title: ImageIO-ext GDAL extensions - GeoServer (at
[12:27] <jdeolive> dany_r: external extension
[12:27] <aaime> it links to binary builds of the imageioext native libraries
[12:27] <jdeolive> but we will distribute it along with the other extensions
[12:28] <aaime> that is not part of the distributed extension, I guess we shoudl update it with new links/binaries for 1.7.3+?
[12:28] <aaime> (the ddl are big and platform dependent, that's why we don't include them in the standard download)
[12:28] <dany_r> k. so, I guess the actual poms/extensions folder in geoserver are ok. Right?
[12:28] <aaime> (the dll, or so files, that is)
[12:28] <dany_r> my last question is for jdeolive
[12:28] <jdeolive> i think so... i guess we just need to copy the LICENSE into release/extensions/imageio...
[12:29] <jdeolive> and that should be it?
[12:29] <dany_r> aaime we can give links to imageio-ext binaries release page
[12:30] <jdeolive> dany_r: replace release/extensions/imageio with release/extensions/gdal
[12:30] <aaime> works for me. No need to logins in order to download from the imageio-exgt page right?
[12:30] <dany_r> no need... I confirm it
[12:30] <aaime> and we should keep the old links for people using 1.7.0 -> 1.7.2
[12:31] <dany_r> jdeolive not sure I have understood...
[12:31] <dany_r> should I rename the gdal extension to imageio-ext-gdal?
[12:31] <dany_r> (on geoserver/release/extensions)
[12:32] <aaime> no, I think he just meant that the extension is actually named gdal and not imageio-ext-gdal
[12:33] <jdeolive> dany_r: yeah, what aaime said
[12:33] <jdeolive> i just wanted to point you to the place where you need to copy the license to
[12:33] <dany_r> ah... ok
[12:33] <CIA-16> arneke# * r599# #10#/trunk/src/gwc/src/test/java/org/geoserver/gwc/ Adding debug output to test
[12:33] <dany_r> ahh ok.. I didn't read your previous message... sorry
[12:34] <jdeolive> no problem
[12:34] <aaime> so I guess that should be it for the roadmap update? anything else?
[12:34] <jdeolive> so, shall we move onto 1.7.4?
[12:34] <jdeolive> or that can wait until next week if people wish
[12:34] <aaime> Ah, roadmap for 1.7.4? Sure
[12:34] <jdeolive> should be short
[12:35] <dany_r> when do you need the license file?
[12:35] <jdeolive> dany_r: in the next few days would be nice since we will be releasing soon
[12:35] <jdeolive> but if not we can always update the artifact after the fact
[12:35] <jdeolive> or not upload to sf until hte license is ready
[12:36] <jdeolive> i can go first for 1.7.4
[12:36] <jdeolive>
[12:36] <sigq> Title: GEOS-2498 Integrate GeoExt based styler - (at
[12:36] <jdeolive> not sure about this one
[12:36] <jdeolive> this may fall off... and will probably just be a 2.0 thing
[12:36] <jdeolive> actually
[12:36] <jdeolive> wait
[12:36] <jdeolive> since we agreed 1.7.4 would be a bug fix release
[12:37] <aaime> Right, was just about to ask
[12:37] <jdeolive> there should probably be no new features in the list...
[12:37] <jdeolive> so yeah... most of those should probably be moved to 2.x
[12:37] <jdeolive> any objections to that?
[12:37] * aaime feels pain for not working on
[12:38] <aaime> but then again, I can handle it
[12:38] <jdeolive>
[12:39] <aaime> simboss, what about the geosolutions point of view?
[12:40] <aaime> is sticking to bug fixes only on 1.7.x cool with you?
[12:40] * vheurteaux ( has joined #geoserver
[12:40] <simboss> yeah
[12:40] <simboss> for the things we have been already talking about
[12:40] <simboss> i believe I will be targeting trunk
[12:40] <simboss> I have a few patches for mosaic and pyramid
[12:40] <simboss> that I want to to commit
[12:41] <simboss> for the next release
[12:41] <simboss> I only miss one thing
[12:41] <simboss> to achieve that
[12:41] <simboss> time
[12:43] <jdeolive> ok... so if there are no objections then... (aaime?) then we shoudl restructure the road map
[12:43] <jdeolive> to reflect this
[12:43] <aaime> Ok, I can limit myself to trunk
[12:43] <groldan> I think this is the best approach. We have a cleaner trunk now
[12:43] <groldan> so merges will be painful though
[12:44] <groldan> I for want want an even cleaner trunk, meaning there's lot of cleanup to do in the wms front
[12:44] <groldan> that I'd love to see on 2.0
[12:44] <aaime> yeah, working on 1.7.x and forward porting just became harder
[12:45] <jdeolive> ok, sounds good then
[12:45] <jdeolive> so that it for road map update
[12:45] <jdeolive> ?
[12:45] <aaime> Ok, just one thing
[12:45] <aaime> what about the packages for linux?
[12:46] <groldan> I don't know.....
[12:46] <aaime> that does not really involve coding in the gs codebase
[12:46] <groldan> who knows how to do that, which distros
[12:46] <jdeolive> i woudl say stuff like that should still be ok...
[12:47] <aaime> ok, so we have some issue moving to do
[12:47] <aaime> what else?
[12:47] * cbriancon ( Quit ("Leaving."#)
[12:49] <jdeolive> 2) 1.7.3 release
[12:49] <jdeolive> not much, just that it will be released in teh next few days
[12:50] <bmmpxf> jdeolive: For 1.7.2, you started a blog post about iit, and I spruced it up...want to do the same for 1.7.3?
[12:50] <dany_r> (I'm updating the geos-imageio-ext-gdal wiki page)
[12:50] <jdeolive> yeah, sounds good
[12:50] * arneke1 ( has joined #geoserver
[12:51] <bmmpxf> I'm still feeling a bit out of the loop, but are there any new features that still need to be documented?
[12:51] * arneke ( Quit (Read error: 113 (No route to host)#)
[12:52] <aaime> mumble mumble... dunno?
[12:53] <aaime> let's look again at the 1.7.3 new featuers
[12:53] <aaime> I guess new SDE raster formats might need some docos? groldan?
[12:53] <aaime> not sure if the namespace filter has been documented
[12:54] <groldan> aaime: indeed
[12:54] <aaime> same goes for the new integrated directory data store
[12:54] <aaime> and the stroke control thing
[12:54] <groldan> so what do we do, schedule some time to update docs before release?
[12:54] <bmmpxf> hmm, wow
[12:54] <aaime> and yeah, we probably need some docs for the legends related to raster symbolizers
[12:55] <simboss> yeah 8-)
[12:55] <bmmpxf> Although I know I'm suggesting drinking from the firehose, but technically new features should be documented before they are technically "done".
[12:55] <groldan> bmmpxf: would it be ok if I send you the sde information by email and you clean it up for the wiki?
[12:56] <bmmpxf> groldan: Yes, perfect
[12:56] <aaime> so mike, how do we handle this? we send mails to you, or udpate the docs and ping you for cleanup?
[12:56] <bmmpxf> hmm
[12:56] <groldan> bmmpxf: you're right, unfortunatelly my doc'ing time was eaten by esri api bugs and schedule overrun
[12:57] <groldan> but yeah, I can surely send you that by email
[12:57] <groldan> with my usual grammar mistakes so you have some fun
[12:57] <bmmpxf> I think in the interests of time, you can either start with a mail or a confluence page and write some notes or whatever you feel like, and I can work them into finished docs. Since I'm not very familiar with the new stuff, any help you can provide would make my work go faster...
[12:57] <bmmpxf> Perhaps a JIRA instead of a mail?
[12:57] <bmmpxf> (two birds...)
[12:58] <groldan> seems good to me
[12:59] <aaime> bmmpxf, a jira asking you to review?
[12:59] <groldan> bmmpxf: you mean like this:
[12:59] <sigq> Title: GEOS-2671 Document support for new arcsde raster formats - (at
[13:00] <simboss> I might have to do the same for legends I guess
[13:00] <simboss> right?
[13:01] <aaime> bmmpxf, that works for me
[13:01] <bmmpxf> Well, I was thinking put the info that needs to be documented (rough notes) into the description of the JIRA, so as to alleviate having a JIRA and an email.
[13:01] <aaime> I can work on the docos and then pass the issue to you for docs beautification
[13:02] * vhamer ( has joined #geoserver
[13:02] <aaime> humm... we done?
[13:02] <aaime> lunch is here
[13:03] <bmmpxf> so docs to be doc'd are: 1) ArcSDE raster formats, 2) namespace filter, 3) integrated directory datastore, 4) stroke control ?
[13:03] <aaime> yeah
[13:03] <simboss> 5) raster legends
[13:03] <bmmpxf> ah right
[13:04] <bmmpxf> so has someone volutneered to send me information about each one of these numbers, or if not, who should I bother?
[13:05] <aaime> bmmpxf, for 2, 3 and 4, me
[13:05] <bmmpxf> 1 is groldan
[13:05] <bmmpxf> 5, simboss?
[13:06] <simboss> yeah
[13:06] <bmmpxf> this works for me. thanks all
[13:06] <aaime> cool
[13:06] <aaime> end of meeting?

Posted at 03 Mar @ 12:22 PM by Mike Pumphrey | 0 comments
Last changed: Feb 12, 2009 11:35 by David Winslow

0) What's up?

1) Road Map

2) OpenGeo Meeting

3) Google Summer of Code Project Ideas [12:03:36]  <jdeolive>    aaime, arneke, bmmpxf, dwins, groldan_ iwillig: meeting time?
[ <aaime>    indeed
[12:03:42] Quit    ahocevar1 has left this server (Read error: 104 (Connection reset by peer)).
[12:03:46] <iwillig>    yeah
[12:03:51] Join    ahocevar has joined this channel (
[12:03:53] <dwins>    hi all
[12:03:57]      * jdeolive is sorry if he missed anyone
[12:03:57] <groldan_>    hi all
[12:04:36] <arneke>    pong
[12:04:41] Topic    jdeolive sets the channel topic to "0) whats up 1) road map".
[12:04:45] <jdeolive>    anyone have any agenda items?
[12:05:13] <aaime>    roadmap update? Oh no, you beated me on that one
[12:06:05] <aaime>    OpenGeo tribe meeting?
[12:06:16] Topic    jdeolive sets the channel topic to "0) whats up 1) road map 2) opengeo meeting".
[12:06:18] <groldan_>    call for remote participation on opengeo sprint?
[12:06:22] <groldan_>    yeah that one aaime
[12:06:42] <aaime>    if cholmes_ is around it would be nice to know more about contributor agreement status?
[12:06:55] Nick    groldan_ is now known as groldan.
[12:07:30] <jdeolive>    cool, lets add that if cholmes_ shows up
[12:07:35] <jdeolive>    shall we start?
[12:07:42] <aaime>    sure
[12:07:55]      * jdeolive has been working on rest config stuff
[12:08:05]      * arneke is working on jdbc backend
[12:08:12]      * bmmpxf is in India at Map World Forum being a walking OpenGeo commercial + suffering from massive culture shock
[12:08:33]      * dwins has been playing with GeoExt, putting together talk for NC GIS
[12:08:55]      * aaime is playing with ogr, csv and excel output formats and doing random rendering improvements
[12:08:57]      * groldan has been working on supporting new arcsde raster formats, its quite stable on trunk, need to backport to 1.7.x
[12:09:22] <groldan>    and /me is making sure test suite is up to an acceptable level in order to backport to the stable branch
[12:09:30]      * iwillig is working on OSM demo and docs when he is done with OSM
[12:10:14] <jdeolive>    cool
[12:10:18] <jdeolive>    1) road map
[12:10:31] <jdeolive>    aaime: sorry, i have not reorganized it yet as you wanted
[12:10:42] <jdeolive>    let's sound off by person
[12:10:54] <jdeolive>    leaving aaime for last since he has the most issues
[12:10:56] <jdeolive>    arneke
[12:11:04] <jdeolive>    any changes?
[12:11:19] <jdeolive>    sorry, here is the link
[12:11:20] <arneke>    Not yet,
[12:11:21] <jdeolive>
[12:11:21] <sigq>    Title: [#GEOS-2605] GetCapabilities response for gwc - (at
[12:11:25] <sigq>    Title: Roadmap - GeoServer (at
[12:11:42] <arneke>    there was a check-in today, but I havent looked it over yet
[12:11:55] <arneke>    by Per
[12:12:09] <arneke> , not done any work on it
[12:12:11] <sigq>    Title: [#GEOS-2513] Add gwc links to the main welcome page - (at
[12:12:32] <jdeolive>    arneke: ok, shall we leave them as is  or push back?
[12:12:56] <arneke>    leave as is, may push the first one back
[12:13:01] <jdeolive>    cool
[12:13:05] <jdeolive>    thanks arneke
[12:13:14] <jdeolive>    groldan
[12:13:49] <jdeolive>    looks like that is quite well on track yes?
[12:13:52] <groldan>    this week I'm wrapping up the arcsde gce stuff, so I should be good as for the release time
[12:13:55] Join    ingenieroariel_ has joined this channel (n=ingenier@
[12:14:05] <jdeolive>    nice
[12:14:16] <jdeolive>    i will go next
[12:14:22] <jdeolive>
[12:14:23] <jdeolive>    moving along
[12:14:24] <sigq>    Title: [#GEOS-2497] move restconfig to core module - (at
[12:14:33] <jdeolive>    i committed my big massive commit
[12:14:40] <jdeolive>    i think dwins is hoping to look over it soon
[12:14:53] <jdeolive>    i am also almost ready to move rest to core... as most votes are in
[12:15:22] <jdeolive>    that is it for me
[12:15:31] <jdeolive>    dwins
[12:15:32] <dwins>    i see a lot of test failures in restconfig right now. have you committed this morning?
[12:15:43] <jdeolive>    dwins: nope...
[12:15:50] <jdeolive>    perhaps i forgot to commit something
[12:15:55] <dwins>    hm. guess there's some debugging to do then
[12:16:00] <jdeolive>    i will build on the server to see if i can replicate
[12:16:02] <dwins>    anyway, we can pick it up after the meeting
[12:16:48] <jdeolive>    dwins: nothing changed with i suppose
[12:16:50] <sigq>    Title: [#GEOS-2496] Move geosearch module to core distribution - (at
[12:17:05] <dwins>    right, no news there
[12:17:23] <jdeolive>    cool
[12:17:29] <dwins>    I guess I need to revive that email thread; I only remember receiving one vote
[12:17:30] <jdeolive>    so... last but not least, aaime!!
[12:17:35] <aaime>    ok
[12:17:42] <aaime>    well, citing just the issues that have some changes
[12:17:55] <aaime> seems well on its road to completion
[12:17:57] <sigq>    Title: [#GEOS-2615] Bring CSV/Excel output formats to extension status - (at
[12:18:21] <aaime> is done and hopefully will allow for nice road maps with embedded one way symbols. Users are testing it now
[12:18:23] <sigq>    Title: [#GEOS-2601] Ability to turn on/off TextSymbolizer "left-to-right" aligning feature - (at
[12:18:38] Quit    ingenieroariel has left this server (Operation timed out).
[12:18:39] <aaime> has been expanded to wfs and wcs too
[12:18:40] <sigq>    Title: [#GEOS-2471] Support namespace filter for WMS/WFS/WCS getCapabilities - (at
[12:19:03] <aaime> I did not get any negative feedback about, maybe time to push directory data store to supported land in gt2
[12:19:05] <sigq>    Title: [#GEOS-2288] Add directory data store to the list of standard GeoServer data stores - (at
[12:19:20] <aaime>    Finally,
[12:19:21] <sigq>    Title: [#GEOS-2584] Allow choice of color reduction algorithm for png8 images - (at
[12:19:38] <aaime>    in which I realized the change was not as easy as I thought, I need to expand the algorithm to 4 dimensions (RGB and A)
[12:19:48] <aaime>    so that will take some chewing
[12:20:08] <aaime>    hopefully I will manage to get it done by 1.7.3, if not, we'll move it to 1.7.4
[12:20:08] <jdeolive>    ok... leave as is or push back?
[12:20:32] <aaime>    if I'm not done one week before the release let's push back, otherwise keep it?
[12:20:54] <jdeolive>    sounds good
[12:21:03] Quit    ingenieroariel_ has left this server (Read error: 104 (Connection reset by peer)).
[12:21:11] <aaime>    On that topic, I think it would be nice if most new features were ready a bit before the releases
[12:21:17] Join    ingenieroariel_ has joined this channel (n=ingenier@
[12:21:25] <aaime>    so that we can ask users involved in the jira request to test out a little
[12:21:59] <jdeolive>    yeah... that would be nice... but sometimes hard to acheive
[12:22:55] <CIA-16>    03aaime * r11301 10/branches/1.7.x/src/community/ (5 files in 3 dirs): Adding feature id column in the csv/excel outputs
[12:23:02] <aaime>    I'm done
[12:23:24] <jdeolive>    cool, thanks aaime
[12:23:49] <jdeolive>    2) opengeo meeting
[12:23:53] <jdeolive>    aaime
[12:24:14] <aaime>    well, OpenGeo has a gathering in NY later this month
[12:24:30] <aaime>    the week of the 21, thought I guess many people will be around the week after it as well
[12:24:42] <aaime>    the main point of the gathering is to talk of course but
[12:24:53] <aaime>    we'll have a geoserver sprint as well
[12:25:06] <aaime>    we don't have a solid agenda so far, but the topics candidate for the sprint are
[12:25:10] Part    SEWilco2 has left this channel.
[12:25:18] <aaime>    catalog cleanup (get rid of all the old catalog and config classes)
[12:25:36] <aaime>    general code cleanup (get rid of all the lingering classes that stayed there, javadocs improvements)
[12:25:51] <aaime>    and bringing the wicket ui to a day to day usable state
[12:26:35] <aaime>    As usual with a sprint an avalanche of commits is to be expected
[12:26:51] <aaime>    not sure if anybody from remote (outside of opengeo) is interested in participating
[12:27:18] <aaime>    but I guess we should tell the above to Ben and Rob since they are working on trunk and during those days trunk will be hit pretty hard
[12:27:18] <dwins>    presumably we should make an announcement on the mailing list. I think most of the people here for the meeting are OG employees
[12:27:34] <aaime>    yep
[12:27:46] <aaime>    anyways, any update of the meeting agenda?
[12:28:41] <dwins>    I don't have any information about it that you don't
[12:29:08] <aaime>    well, maybe someone wants to propose something else, I don't know
[12:29:31] <dwins>    tbh, I think those three make for a pretty full plate
[12:29:45] <aaime>    oh for me too
[12:30:18] <dwins>    the jteam was hoping to do some quick iterations on the Ext preview application we mentioned on the opengeo mailing list as well
[12:30:48] <aaime>    yeah, one thing that I would like to see in this meeting, compared to the Bolsena one, is more talking less coding
[12:31:01] Quit    cbriancon has left this server ("Leaving.").
[12:31:02] <dwins>    ++
[12:31:03] <aaime>    that is, unless we have very detailed planning for coding in anger
[12:31:22] <bmmpxf>    aaime: why do you say "more talking less coding"?
[12:31:33] <aaime>    Imho coding we do fine even when at home, it's talking that we don't do very well over chat
[12:31:53] <aaime>    bmmpxf, because the Bolsena sprint last year disappointed me somewhat
[12:32:02] <aaime>    as we rushed into coding something that was not very well defined
[12:32:15] <aaime>    because the objective was to get the UI "done" during the meeting
[12:32:34] <aaime>    but in the end I have the feeling we've done something, but that something is not as useful as it could be
[12:32:48] <aaime>    a case of coding against of the wrong requirements, or something like that
[12:32:50] <bmmpxf>    was that a failure of planning or a failure of what is realistic to accomplish?
[12:33:17] <aaime>    Also it would be nice to take it a little more relaxed, we were coding each day up to 11PM
[12:33:31] <aaime>    under a not better specified sense of urgency
[12:34:01] <aaime>    bmmpxf, probably lack of preparation
[12:34:23] <aaime>    the code spring in Victoria for switching from the old to the new feature model at foss4g 2007 was sure hard work
[12:34:32] <aaime>    but it was prepared in detail and we managed to accomplish a lot
[12:34:34] <bmmpxf>    aaime: I ask mainly because I've never encountered an official sprint before, that's all.
[12:34:41] Part    Bojan has left this channel.
[12:34:54] <aaime>    ah, not an expert here either, only done two so far
[12:35:11] <aaime>    (the FOSS4G 2008 one did not really work out, everybody was way too tired)
[12:36:09] <aaime>    both were setup in a "code a lot" way
[12:36:13] <aaime>    but only one really worked out imho
[12:36:40] <aaime>    but maybe it's just me
[12:36:46] <aaime>    dwins, jdeolive, you were there too
[12:36:47] <dwins>    on the other hand, the Victoria sprint involved a lot of superficial changes, at least the part of it that I was around for
[12:37:03] <aaime>    dwins, yeah, it was pretty mechanical
[12:37:23] <aaime>    whilst in Bolsena we were facing a new technology, we did not have much web IU experience at all
[12:37:25] <dwins>    that seems to not be really taking advantage of having everyone in the same room
[12:37:56] <aaime>    dwins, in that case it was more of an issue of using those days because once home we would not have done it otherwise
[12:38:04] <aaime>    as it was the kind of change that nobody really sponsors for
[12:38:36] <dwins>    no argument there.
[12:38:39] Join    gdavis_ has joined this channel (
[12:39:17] <bmmpxf>    aaime: Yeah, perhaps the nice thing about a sprint is that it forces people to do things they wouldn't do otherwise...
[12:39:28] <aaime>    maybe
[12:39:41] <bmmpxf>    you know, progress through co-worker guilt. 
[12:40:02] <aaime>    well, shall we move forward?
[12:40:06] <aaime>    or else... close the meeting?
[12:40:13] <aaime>    any other topic?
[12:41:02] <arneke>    what's left ? copuright input ?
[12:41:06] <dwins>    maybe a call for Summer of Code project ideas
[12:41:38] <aaime>    dwins, good idea
[12:41:50] <aaime>    it's always hard to get the right balance for those things
[12:41:57] <dwins>
[12:41:59] <sigq>    Title: Google Summer of Code 2009 Ideas - OSGeo Wiki (at
[12:42:03] <aaime>    like not so hard that the students run away of get stuck midway
[12:42:12] <aaime>    but no so easy to make everybody bored either
[12:43:17] <aaime>    Hum, maybe we coudl have a stud make a full restful wfs
[12:43:25] <aaime>    that should strike a nice balance?
[12:43:43] <aaime>    (stud-ent)
[12:43:51] <arneke>    do we think anyone would use it ?
[12:43:59] <arneke>    make life easier for the JS guys ?
[12:44:23] <aaime>    maybe, everybody hates WFS-T
[12:45:02] <aaime>    also 3d thematic mapping in kml ala could be interesting
[12:45:23] <aaime>    we already have most of what is needed there, shoudl not be that hard? (or maybe it's too easy?)
[12:45:38] <dwins>    a Cascadenik clone for GeoServer would be cool too
[12:46:00] <aaime>    right, parse css and turn that into a Style object in memory?
[12:46:34] <aaime>    I'm only worried about the double maintenance that we'd have to deal with if that gets any traction
[12:47:36] <dwins>    like the maintenance for all the datastores and output formats we have?
[12:47:37] <aaime>    it may also be nice to have templates for pdf output
[12:47:56] <aaime>    like the mapfish one, but more visual
[12:48:11] <aaime>    like defining the template in SVG using Inkscape
[12:48:15] <arneke>    I think that would be handy for a lot of organizations
[12:48:25] <aaime>    and taggig certain areas as map/legend/scale
[12:48:32] Join    simboss has joined this channel (
[12:48:47] <aaime>    with mapfish you have options, but I've found getting a good output somewhat painful
[12:48:58] <arneke>    plotting templates are always painful
[12:48:59] <aaime>    a lot of try and fix and try and fix and curse
[12:49:22] <aaime>    yeah, the downside of the mapfish one beign that it's defined in a text file for everything
[12:49:33] <arneke>    1200 dpi  A0 plots -> fun
[12:49:50] <aaime>    it's also slwo because the servlet asks the map server for a SVG output (to have resolution indep vectors in the pdf)
[12:49:56] <aaime>    and then has to parse it back
[12:50:19] <simboss>    (hi all, sorry I am late)
[12:50:22] <arneke>    i like the SVG idea
[12:50:39] <arneke>    it offloads a lot of the problems i've dealt with in the past
[12:51:00] <aaime>    yeah, it's a good one, just not optimal from a perf pov
[12:51:02] <arneke>    like logos being resized and stuff
[12:51:18] <aaime>    not sure you noticed, but the native pdf output of geoserver is fully vector too
[12:51:36] <aaime>    (well, almost, there is a bug I need to fix there in case multiple feature type styles are in use)
[12:52:08] <arneke>    no, havent tried it, cool
[12:52:28] Quit    bmmpxf has left this server (Read error: 54 (Connection reset by peer)).
[12:53:57] <aaime>    any other suggestion for the Soc?
[12:55:41] <dwins>    i'm done for now

Posted at 10 Feb @ 1:12 PM by David Winslow | 0 comments
Last changed: Feb 03, 2009 16:31 by Mike Pumphrey

0) what's up?
1) road map update

[12:00] <dwins> meeting time?
[12:01] <aaime> yep
[12:01] <aaime> arneke, bmmpxf, dwins, groldan, jdeolive
[12:01] <arneke> poing
[12:01] <aaime> simboss
[12:01] <groldan> pong
[12:02] <aaime> (did I forget anybody? boys, it's getting harder to find the devs amont 38 people connected...)
[12:02] <arneke> I think that's it
[12:02] <aaime> iwillig
[12:03] <aaime> topics for today's meeting?
[12:03] <iwillig> aaime: ?
[12:03] * jdeolive changes topic to '0) whats up 1) road map update'
[12:03] <aaime> geoserver weekly dev meeting
[12:03] <simboss> no relevant topics here
[12:03] <aaime> I have two closely related to the road map
[12:04] <simboss> go ahead4
[12:04] <simboss> I can just reort progress on it
[12:04] <aaime> So let's go and we'll see if they need their own topic, or not
[12:04] <aaime> maybe jdeolive leave me as the last one for the road map update
[12:05] <jdeolive> aaime: sure
[12:05] * tschaub ( has joined #geoserver
[12:06] <arneke> 0 ?
[12:06] <aaime> sure
[12:06] <aaime> 0) What's up
[12:06] * jdeolive has been working on rest config
[12:06] * dwins has been playing with browser scripting
[12:06] * aaime is recoverig from flu, hammering jdbc-ng, testing out neuquant palette generation algorithm
[12:07] * groldan has been working on new arcsde raster formats support
[12:07] * arneke is writing up jdbc vs <other options and problems> , looking for a dataset to test new geosearch module on, + lots of minor tasks... welcoming new contributors
[12:07] <simboss> simboss: raster legends
[12:09] <jdeolive> cool
[12:09] <jdeolive> 1) road map update
[12:09] <jdeolive> lets go by person, leaving aaime for last
[12:09] <jdeolive> i will start
[12:09] <jdeolive>
[12:09] <sigq> Title: [#GEOS-2497] move restconfig to core module - (at
[12:09] <jdeolive> is still on track
[12:09] <jdeolive> groldan
[12:10] <jdeolive>
[12:10] <sigq> Title: [#GEOS-2567] Support new ArcSDE raster formats - (at
[12:10] <jdeolive> i wonder if that needs to be broken up into smaller chunks?
[12:10] * aaime notices the roadmap is still not linked from the home page?
[12:11] * jdeolive slaps himself
[12:11] <jdeolive>
[12:11] <sigq> Title: Roadmap Draft - GeoServer (at
[12:11] * aaime provides a trout
[12:11] <jdeolive> ok... this time i promise to do that today
[12:11] <aaime> lol
[12:12] * mlundblad1 (n=marcus@ Quit ("Leaving.")
[12:12] <aaime> groldan?
[12:13] * afabiani ( has joined #geoserver
[12:13] <aaime> jdeolive, maybe let's move next and get back to groldan when he's with us again
[12:13] <jdeolive> cool
[12:13] <jdeolive> arneke
[12:13] <arneke> I'll jump in while we wait,,, ... haven't started, should be easy. The getcapabilities response implies that we officially take over what jgarnett wrote
[12:13] <sigq> Title: [#GEOS-2513] Add gwc links to the main welcome page - (at
[12:13] * groldan_ ( has joined #geoserver
[12:14] * groldan ( Quit (Nick collision from services.)
[12:14] * groldan_ is now known as groldan
[12:14] <arneke> or my new swedish friends, pelle__ and pereng, are also working on a new getcapabilities response, so maybe that can be used
[12:14] <arneke> but the basic linking to the layers is easy, obviously
[12:15] <arneke> would it be feasible to create a getcapabilities hook during the sprint, and insert the gwc tilesets there ?
[12:16] <aaime> sounds good, how long are you staying there?
[12:16] <arneke> my swedish friends in the channel ?
[12:16] <aaime> no, you in NY
[12:16] <arneke> oh, 21st to 3rd
[12:16] <arneke>
[12:16] <aaime> I guess we have plenty of time thne
[12:17] <arneke> cool, so i'll add the links to the openlayers demos soon, and we can wait with the getcapabilities until NYC and decide then
[12:17] * alpiv (i=8f739f35@gateway/web/ajax/ has joined #geoserver
[12:18] <aaime> next?
[12:18] <jdeolive> sounds good, so arneke do you want to break that part into a separate issue?
[12:18] <arneke> yeah, will do that
[12:18] <jdeolive> moving along... dwins ?
[12:18] <jdeolive>
[12:18] <sigq> Title: [#GEOS-2496] Move geosearch module to core distribution - (at
[12:18] <jdeolive> i guess this is sort of blocking on the rest improvements
[12:19] <dwins> it is
[12:20] <dwins> but the rest changes should hit svn soon right?
[12:20] <jdeolive> yup, i should actually do that today... i have been just working on docs... not much more code to write
[12:20] <jdeolive> dwins: since i have the changes local do you want me to patch up the depending modules?
[12:21] <dwins> sure, I don't have a particular need to do the fixes myself
[12:22] <jdeolive> ok, i will see how much effort that will be
[12:22] <jdeolive> moving along
[12:22] <jdeolive> simboss
[12:22] <jdeolive>
[12:22] <sigq> Title: [#GEOS-1162] Support custom raster legends via GetLegendGraphic - (at
[12:24] <aaime> hmmm... slow moving meeting today
[12:24] <shadow_phoenix> Just a curiosity for a future project of mine. Does geoserver support real time changing maps? Like weather maps and the like
[12:24] <aaime> now now shadow_phoenix, we're doing the developer meeting
[12:24] <groldan> on a related note, I wanted to know if what Saul also proposes here ( is already supported with the histogram tag in an SLD
[12:24] <sigq> Title: [#GEOT-1391] Support Histogram Contrast Stretching in GeoTools Rasters - (at
[12:24] <shadow_phoenix> oops, sorry, I'll ask tomorrow
[12:24] * arneke tried to edit the roadmap draft to add his new issue.. oh my... that's fancy
[12:25] <jdeolive> arneke: yeah, its done in jira
[12:25] <aaime> arneke, you don't edit it
[12:25] <aaime> you just assign a "new feature" issue to 1.7.2
[12:25] <jdeolive> groldan: i think you dropped off during your issue
[12:25] <aaime> 1.7.3 sorry
[12:25] <jdeolive>
[12:25] <sigq> Title: [#GEOS-2567] Support new ArcSDE raster formats - (at
[12:25] <arneke> aaime: yeah, I realized
[12:25] <jdeolive> should we break it out into separate issues?
[12:26] <groldan> as to meet the release you mean?
[12:26] <groldan> (yeah I dropped off, btw)
[12:27] <groldan> jdeolive: I'm not getting the point of why to separate GEOS-2567 into different issues
[12:27] <jdeolive> groldan: well.. just that support new arcsde formats is sort of broad... i mean some of hte work is done now right?
[12:27] <jdeolive> breaking it up might be a bit nicer than continually pushing the issue back as more progress is made in various areas
[12:27] <groldan> ah ok, but no, none of them can be taken as finalized
[12:27] <jdeolive> ahh, ok
[12:28] <groldan> there are issues like tiles going blank or pyramid levels not rendering at all
[12:28] <groldan> so I guess we'll have to push it back altogether
[12:29] <groldan> if that weren't the case I would be ok with splitting it up
[12:29] <jdeolive> ok.. lets push back i guess.. and maybe as parts of completed, break them out into sub issues?
[12:30] <groldan> hum, they're already dependent issues, GEOS-2567 is just a grouping one?
[12:30] <groldan> ie, new feature->task, task, task
[12:31] <groldan> does that make sense?
[12:31] <jdeolive> yup
[12:32] <jdeolive> ok... moving on
[12:32] <jdeolive> aaime: i think you are up
[12:32] <aaime> Ah ok
[12:32] <aaime> well, I have a bunch of closed items
[12:32] <aaime> two merit some note
[12:32] <aaime> the first one is the directory data store inclusion
[12:32] <aaime> and release data switch to use it
[12:33] <aaime> I've done so to ensure some exposure
[12:33] <aaime> but it would be nice if you could throw it at a random folder filled of shapefiles on your system
[12:33] <aaime> and see how it reacts
[12:33] <aaime> maybe I should also ask on the users ml for eager people to try out the nightlies
[12:33] <jdeolive> yeah that might be nice
[12:34] <arneke> will do (test it)
[12:34] <aaime> Will do then
[12:34] <jdeolive> i can do some testing though when i wrap up the rest stuff
[12:34] <aaime> The second closed issue that needs some attention is the wms caps filtering by namespace
[12:34] <aaime> works fine, it's kind of crying murder the fact that wfs and wcs do not have the same
[12:34] <aaime> the patch is like 20 loc
[12:35] <aaime> so I was considering reopening and porting over to the two other services as well
[12:35] * pramsey is now known as pramsey_dentist
[12:35] <aaime> opinions?
[12:35] <arneke> consistency is nice
[12:35] <aaime> indeed, plus a client using both wms and wfs is getting common
[12:36] <aaime> (wcs is still dead in the water with no clients but hey, what can I do )
[12:36] <jdeolive> +1 on applying the patch to wfs and wcs
[12:36] <aaime> the last bit I wanted to talk about is the png palette reduction
[12:36] <aaime> the neuquant algorithm
[12:37] <aaime> I tested it yesterday a bit and the results seem very promising quality wise
[12:37] <aaime> it's just slow, but if we combine it with a tile cache, I guess it's not going to be such an issue
[12:37] <aaime> also, the current code first extracts a palette and then performs a conversion
[12:38] <aaime> I wanted to experiment using Simone palette inverter to see if it's any faster
[12:38] <arneke> what are we using today?
[12:38] * SEWilco2 (n=sewilco@ has left #geoserver
[12:38] <aaime> soo... expect good looking fully translucent images in an Internet Explorer 6 next to you for 1.7.3
[12:38] <arneke> median cut? just looking this stuff up for the first time
[12:39] <aaime> simboss boiled down a super-optimized algorithm
[12:39] <aaime> out of some books
[12:39] <aaime> darn fast and works fine when the image is opaque
[12:39] <simboss> aaime newquant should extract the palett first anyway, right?
[12:39] <aaime> but does not deal well with the way antialiasing is made on transparent image
[12:39] <aaime> images
[12:39] <arneke> ah
[12:40] <aaime> simboss, yep, that's the first stage of the algorithm
[12:40] <aaime> so I was considering extracting the palett, compressing it (it has no way to tell there are only 8 colors in the image, for example)
[12:40] <aaime> and then treating it like it was user provided
[12:40] <arneke> so what sampling factor will we be using ?
[12:41] <aaime> I'm thinking of making it configurable
[12:41] <simboss> +1 aaime
[12:41] <simboss> we have benn talking about that
[12:41] <simboss> since 2007
[12:41] <aaime> it's always slow, but it gets acceptable if you sample 1 out of 8 pixels
[12:41] <aaime> or 1 out of 4
[12:41] * bmmpxf ( Quit (No route to host)
[12:42] <simboss> are you planning on using it by default?
[12:42] <aaime> my idea was to use yours by default on opaque images
[12:42] <aaime> and neuquant by default on transparent
[12:42] <aaime> ones
[12:44] <aaime> as the current one is fast and works fine on opaque ones
[12:44] <aaime> but its output is no good on transparent ones
[12:44] <aaime> see here:
[12:44] <simboss> can you adapt some of the code of the current one?
[12:44] <aaime> png, transparent:,40.72812176530762,-73.97755615633545,40.75677019914551&transparent=true
[12:44] <aaime> png8, transparent:,40.72812176530762,-73.97755615633545,40.75677019914551&transparent=true
[12:44] <simboss> I know why it is that way
[12:45] <simboss> but fixing it
[12:45] <simboss> would require some changes to the algorithm
[12:45] <aaime> fixing it is beyond my abilities and the time I have
[12:45] <aaime> right, which would require me to study it first
[12:45] <simboss> right now I am searching in 3d space
[12:45] <simboss> I would need to search in 4d space
[12:45] <simboss> (rgba)
[12:45] <aaime> whilst with neuquant I just have to pick, adapt and include
[12:45] <simboss> the transparency right now is an hacl
[12:45] <simboss> yeah that's cool
[12:46] <aaime> if you think you can fix the original one and have the same quality and better performance
[12:46] <simboss> hck
[12:46] <simboss> hack
[12:46] <simboss>
[12:47] <aaime> then we can open another jira and you can do the change when you have time
[12:47] <simboss> it is not something I can do overnight anyway
[12:47] <simboss> but sooner or later it would be nice to
[12:47] <simboss> fix both
[12:47] <aaime> right, and you are the one that wrote the algorithm
[12:47] <simboss> feel free to do that
[12:47] <simboss> someone may even pay for that
[12:48] <aaime> for faster png8 fully antialiased output?
[12:48] <aaime> Hum, may be
[12:50] * pereng (n=pereng@ Quit ("Computer went to sleep")
[12:50] <aaime> is that it?
[12:50] <aaime> anything else, anybody?
[12:51] <arneke> nope
[12:51] <simboss> not me

Posted at 03 Feb @ 4:22 PM by Mike Pumphrey | 0 comments
Last changed: Feb 03, 2009 13:50 by Andrea Aime

0) whats up
1) road map update

aaime: 0) What's up
***jdeolive has been working on rest config
***dwins has been playing with browser scripting
***aaime is recoverig from flu, hammering jdbc-ng, testing out neuquant palette generation algorithm
***groldan has been working on new arcsde raster formats support
***arneke is writing up jdbc vs <other options and problems> , looking for a dataset to test new geosearch module on, + lots of minor tasks... welcoming new contributors
simboss: simboss: raster legends
jdeolive: cool
jdeolive: 1) road map update
jdeolive: lets go by person, leaving aaime for last
jdeolive: i will start
sigq: Title: GEOS-2497 move restconfig to core module - (at
jdeolive: is still on track
jdeolive: groldan
sigq: Title: GEOS-2567 Support new ArcSDE raster formats - (at
jdeolive: i wonder if that needs to be broken up into smaller chunks?
***aaime notices the roadmap is still not linked from the home page?
***jdeolive slaps himself
sigq: Title: Roadmap Draft - GeoServer (at
***aaime provides a trout
jdeolive: ok... this time i promise to do that today
aaime: lol
mlundblad1 ha abbandonato la stanza (quit: "Leaving.").
aaime: groldan?
afabiani [] è entrato nella stanza.
aaime: jdeolive, maybe let's move next and get back to groldan when he's with us again
jdeolive: cool
jdeolive: arneke
arneke: I'll jump in while we wait,,, ... haven't started, should be easy. The getcapabilities response implies that we officially take over what jgarnett wrote
sigq: Title: GEOS-2513 Add gwc links to the main welcome page - (at
groldan_ [] è entrato nella stanza.
groldan ha abbandonato la stanza (quit: Nick collision from services.).
groldan_ è ora conosciuto come groldan
arneke: or my new swedish friends, pelle__ and pereng, are also working on a new getcapabilities response, so maybe that can be used
arneke: but the basic linking to the layers is easy, obviously
arneke: would it be feasible to create a getcapabilities hook during the sprint, and insert the gwc tilesets there ?
aaime: sounds good, how long are you staying there?
arneke: my swedish friends in the channel ?
aaime: no, you in NY
arneke: oh, 21st to 3rd
aaime: I guess we have plenty of time thne
arneke: cool, so i'll add the links to the openlayers demos soon, and we can wait with the getcapabilities until NYC and decide then
alpiv [i=8f739f35@gateway/web/ajax/] è entrato nella stanza.
aaime: next?
jdeolive: sounds good, so arneke do you want to break that part into a separate issue?
arneke: yeah, will do that
jdeolive: moving along... dwins ?
sigq: Title: GEOS-2496 Move geosearch module to core distribution - (at
jdeolive: i guess this is sort of blocking on the rest improvements
dwins: it is
dwins: but the rest changes should hit svn soon right?
jdeolive: yup, i should actually do that today... i have been just working on docs... not much more code to write
jdeolive: dwins: since i have the changes local do you want me to patch up the depending modules?
dwins: sure, I don't have a particular need to do the fixes myself
jdeolive: ok, i will see how much effort that will be
jdeolive: moving along
jdeolive: simboss
sigq: Title: GEOS-1162 Support custom raster legends via GetLegendGraphic - (at
aaime: hmmm... slow moving meeting today
bmmpxf|mtg [] è entrato nella stanza.
shadow_phoenix: Just a curiosity for a future project of mine. Does geoserver support real time changing maps? Like weather maps and the like
aaime: now now shadow_phoenix, we're doing the developer meeting
groldan: on a related note, I wanted to know if what Saul also proposes here ( is already supported with the histogram tag in an SLD
sigq: Title: GEOT-1391 Support Histogram Contrast Stretching in GeoTools Rasters - (at
shadow_phoenix: oops, sorry, I'll ask tomorrow
***arneke tried to edit the roadmap draft to add his new issue.. oh my... that's fancy
aaime: arneke, you don't edit it
jdeolive: arneke: yeah, its done in jira
aaime: you just assign a "new feature" issue to 1.7.2
jdeolive: groldan: i think you dropped off during your issue
aaime: 1.7.3 sorry
sigq: Title: GEOS-2567 Support new ArcSDE raster formats - (at
arneke: aaime: yeah, I realized
jdeolive: should we break it out into separate issues?
groldan: as to meet the release you mean?
groldan: (yeah I dropped off, btw)
groldan: jdeolive: I'm not getting the point of why to separate GEOS-2567 into different issues
jdeolive: groldan: well.. just that support new arcsde formats is sort of broad... i mean some of hte work is done now right?
jdeolive: breaking it up might be a bit nicer than continually pushing the issue back as more progress is made in various areas
groldan: ah ok, but no, none of them can be taken as finalized
jdeolive: ahh, ok
groldan: there are issues like tiles going blank or pyramid levels not rendering at all
groldan: so I guess we'll have to push it back altogether
groldan: if that weren't the case I would be ok with splitting it up
jdeolive: ok.. lets push back i guess.. and maybe as parts of completed, break them out into sub issues?
groldan: hum, they're already dependent issues, GEOS-2567 is just a grouping one?
groldan: ie, new feature->task, task, task
groldan: does that make sense?
jdeolive: yup
jdeolive: ok... moving on
jdeolive: aaime: i think you are up
aaime: Ah ok
aaime: well, I have a bunch of closed items
aaime: two merit some note
aaime: the first one is the directory data store inclusion
aaime: and release data switch to use it
aaime: I've done so to ensure some exposure
aaime: but it would be nice if you could throw it at a random folder filled of shapefiles on your system
aaime: and see how it reacts
aaime: maybe I should also ask on the users ml for eager people to try out the nightlies
jdeolive: yeah that might be nice
aaime: Will do then
arneke: will do (test it)
jdeolive: i can do some testing though when i wrap up the rest stuff
aaime: The second closed issue that needs some attention is the wms caps filtering by namespace
aaime: works fine, it's kind of crying murder the fact that wfs and wcs do not have the same
aaime: the patch is like 20 loc
aaime: so I was considering reopening and porting over to the two other services as well
aaime: opinions?
pramsey è ora conosciuto come pramsey_dentist
arneke: consistency is nice
aaime: indeed, plus a client using both wms and wfs is getting common
aaime: (wcs is still dead in the water with no clients but hey, what can I do )
jdeolive: +1 on applying the patch to wfs and wcs
aaime: the last bit I wanted to talk about is the png palette reduction
aaime: the neuquant algorithm
aaime: I tested it yesterday a bit and the results seem very promising quality wise
aaime: it's just slow, but if we combine it with a tile cache, I guess it's not going to be such an issue
aaime: also, the current code first extracts a palette and then performs a conversion
aaime: I wanted to experiment using Simone palette inverter to see if it's any faster
arneke: what are we using today?
SEWilco2 ha abbandonato la stanza.
aaime: soo... expect good looking fully translucent images in an Internet Explorer 6 next to you for 1.7.3
arneke: median cut? just looking this stuff up for the first time
aaime: simboss boiled down a super-optimized algorithm
aaime: out of some books
aaime: darn fast and works fine when the image is opaque
aaime: but does not deal well with the way antialiasing is made on transparent image
aaime: images
simboss: aaime newquant should extract the palett first anyway, right?
arneke: ah
aaime: simboss, yep, that's the first stage of the algorithm
aaime: so I was considering extracting the palett, compressing it (it has no way to tell there are only 8 colors in the image, for example)
aaime: and then treating it like it was user provided
arneke: so what sampling factor will we be using ?
aaime: I'm thinking of making it configurable
simboss: +1 aaime
simboss: we have benn talking about that
simboss: since 2007
aaime: it's always slow, but it gets acceptable if you sample 1 out of 8 pixels
aaime: or 1 out of 4
bmmpxf ha abbandonato la stanza (quit: No route to host).
simboss: are you planning on using it by default?
aaime: my idea was to use yours by default on opaque images
aaime: and neuquant by default on transparent
aaime: ones
aaime: as the current one is fast and works fine on opaque ones
aaime: but its output is no good on transparent ones
aaime: see here:
simboss: can you adapt some of the code of the current one?
aaime: png, transparent:,40.72812176530762,-73.97755615633545,40.75677019914551&transparent=true
aaime: png8, transparent:,40.72812176530762,-73.97755615633545,40.75677019914551&transparent=true
simboss: I know why it is that way
simboss: but fixing it
aaime: fixing it is beyond my abilities and the time I have
simboss: would require some changes to the algorithm
aaime: right, which would require me to study it first
simboss: right now I am searching in 3d space
simboss: I would need to search in 4d space
simboss: (rgba)
aaime: whilst with neuquant I just have to pick, adapt and include
simboss: the transparency right now is an hacl
simboss: yeah that's cool
aaime: if you think you can fix the original one and have the same quality and better performance
simboss: hck
simboss: hack
aaime: then we can open another jira and you can do the change when you have time
simboss: it is not something I can do overnight anyway
simboss: but sooner or later it would be nice to
simboss: fix both
aaime: right, and you are the one that wrote the algorithm
simboss: feel free to do that
simboss: someone may even pay for that
aaime: for faster png8 fully antialiased output?
aaime: Hum, may be
pereng ha abbandonato la stanza (quit: "Computer went to sleep").
aaime: is that it?
aaime: anything else, anybody?
arneke: nope
simboss: not me
pereng [n=pereng@] è entrato nella stanza.
jdeolive: so simboss about GEOS-1162
jdeolive: still on track for release?
simboss: I have started to committ
simboss: I am refactoring right now
simboss: (pretty much)
jdeolive: cool, sounds good
simboss: and then I will add more tests
simboss: aaime has already seen some images
simboss: he can judge
aaime: looks good from what I can tell
simboss: wait, no better not to, he is never happy ;-(
aaime: there's always room for improvement

Posted at 03 Feb @ 1:44 PM by Andrea Aime | 0 comments
Last changed: Jan 27, 2009 13:14 by Andrea Aime

0) what's up
1) road map update
2) directory ds
3) imagemosaic
4) setting up dev env

***jdeolive has been working on rest config, and doing geotools 2.5.3 release
***aaime has been fixing issues with geoserver/oracle and huge data sets, and working on directory data store on the remaining time
simboss: simboss: raster labels + autoconfig imagemosaic (so that aaime is happy )
***bmmpxf is writing and organizing the new sphinx docs
aaime: raster labels???
simboss: sorry legends
aaime: aaah
aaime: groldan does not seem to be among us
aaime: let's move on?
jdeolive: 1) road map update
sigq: Title: Roadmap Draft - GeoServer (at
jdeolive: instead of going through each issue, can anyone voice any issues they have?
jdeolive: anything that needs to be moved back? or anything that needs to be added?
jdeolive: or do people prefer to go through all the issues?
Bojan ha abbandonato la stanza.
aaime: I'm looking
aaime: jdeolive, what about going person by person
jdeolive: sure
jdeolive: i will go first
sigq: Title: GEOS-2497 move restconfig to core module - (at
jdeolive: on schedule
jdeolive: i am done
jdeolive: aaime?
aaime: sure
sigq: Title: GEOS-2584 Allow choice of color reduction algorithm for png8 images - (at
aaime: still haven't started, but confident I can pull it off
aaime: (we still have 3 weeks right?)
jdeolive: yup
sigq: Title: GEOS-2502 Labelling improvements - (at
aaime: actually, this one is done
jdeolive: closer to 4 actually
sigq: Title: GEOS-2471 Support namespace filter WMS getCapabilities - (at
aaime: will do this
sigq: Title: GEOS-2288 Add directory data store to the list of standard GeoServer data stores - (at
aaime: Almost there, want to discuss details later
sigq: Title: GEOS-1411 Allow for optional callback parameter in GET requests that return json formatted structures - (at
aaime: will do
sigq: Title: GEOS-2588 Configuration reader triggers bounds computation for that whose native bbox is not known - (at
aaime: actually already fixed too
aaime: there is one more that I already fixed that I do not see
aaime: probably mis-tagged
aaime: it's about being able to control line width under 1.5 pixels
jdeolive: right... we should tag it for 1.7.3 at the appropriate priority
aaime: this one:
sigq: Title: GEOS-2228 Stroke width control impossible below 1.5 pixels - (at
aaime: turned it into a feature
aaime: and with this one I'm done
jdeolive: cool, thanks aaime
jdeolive: groldan is not here
jdeolive: neither is daniele, dwins, or arneke
jdeolive: simboss: ?
simboss: let me ping daniele quickly
simboss: anyway
simboss: we have to solve some issues with geotools licensing
simboss: as far as our work on imageio-ext and gdal is concerned
simboss: everything is ready
aaime ha scelto come argomento: 0) what's up 1) road map update 2) directory ds 3) imagemosaic 4) setting up dev env
simboss: once we got the licensing straightened out
simboss: we should be ready to go
simboss: we have 4 weeks I hope it is enough
jdeolive: cool, thanks
jdeolive: what about
simboss: as far as legends are involved, the work is now stalled due to lack of time
sigq: Title: GEOS-1162 Support custom raster legends via GetLegendGraphic - (at
simboss: on my side
simboss: I am confident to be able to cut a first release for the 23rd
simboss: otherwise
simboss: a bit earlier I will update the roadmap by moving it forward
jdeolive: sounds good
simboss: (it's the beginning of the year, lot of time goes to talking to clients )
jdeolive: understandable
iwillig [] è entrato nella stanza.
jdeolive: ok... shall we move on?
aaime: sure
aaime: 2) directory data store
aaime: So, I believe I'm almost there with this one
aaime: however, testing it is crucial, especially since the usage of soft references might lead to some heisenbug
aaime: so I'd suggest the following
aaime: what if I change the 1.7.x release config to use it as soon as possible (tomorrow)
aaime: so that we have one full month or so of people using it before the release?
aaime: the idea is to have ny and tasmania layers use it
aaime: and eventually ask people on the users list
aaime: to try out the nightlies
aaime: with their own directories of shapefiles
jdeolive: aaime: sure, works for me. one thing, would it be possible to do in a way that is easy to move back and forth?
jdeolive: i guess svn revert
aaime: yeah, that was the plan
aaime: if it's not working, we just revert the config
jdeolive: ok, sounds good, +1 for me
aaime: well, if I don't get any -1, I'll try the switch tomorrow
aaime: unless I find some serious issues on my own
aaime: 3) imagemosaic
aaime: simboss?
simboss: quick question
simboss: how would you expect it to autocnfigure some layers?
simboss: I am doing some experiments
simboss: and feebdack is welcome
simboss: or better suggestions
aaime: Well, ideally
aaime: just point the thing to a directory filled with raster
aaime: eventually have the user specify the type of the rasters to be loaded
aaime: and have it build whatever it needs
aaime: even with hundreds of files building the shpaefile and the property file should not take that long?
aaime: As a less optimal option
aaime: have it generate just the property file
aaime: when the .shp is already available
aaime: in both cases, do the job only if the files are missing
aaime: no fancy directory contents tracking
aaime: (I mean, the latter could be useful, but I see that it would be hard as well)
simboss: my plan is to do this
simboss: the user con point us to a something like
simboss: path-to-dir(?name=xxxx&wildcard=xxxx)
simboss: (parenthesis mean optional params)
simboss: so that we build a mosaic with whatever is inside the path-to-dir dir that is suitable
simboss: and we name it like its own dir
simboss: we can add query params to override default behavior
simboss: feedback?
aaime: I guess that would be a good idea
aaime: maybe add a couple defaults
aaime: like name== name of the dir
simboss: y
aaime: and wildcard... dunno, tiff?
simboss: .
aaime: ah, yeah, that could work as well
aaime: simboss, anything else?
aaime: shall we move on to the next topic?
simboss: I am done
aaime: 4) setting up the dev env
aaime: well, it's not really news, but it's too hard for the average developer
aaime: even with instructions
aaime: wondering if we can do anything about that
jdeolive: i have an idea
jdeolive: we could upload some artifacts for each release
jdeolive: which are 1) a pre setup maven repository
jdeolive: 2) a pre eclipse:eclipse'd source tree
jdeolive: as long as they set up eclipse to point at maven, its just an import and then they are set
jdeolive: (if they use eclipse that is)
aaime: sure, that could help
aaime: seems kind of a manual process
aaime: also the failures we've seen on the ml
aaime: they seem kind of new to me
aaime: maybe he's building with java 1.6?
aaime: the fact that builds are checked only with 1.5 is kind of source of issues as well I guess
aaime: since a lot of people develop on 1.6
aaime: maybe also add some 1.6 builds on hudson that do run just once a day?
jdeolive: sounds like a good idea
aaime: Besides that, we need source cleanup and more extensive dev docs I guess
aaime: Maybe something we can tackle, at least partly, at the NY OpenGeo sprint
jdeolive: yeah, that was brought up as an idea, i am all for that
aaime: I'll open a jira issue for release artifacts and the 1.6 builders
jdeolive: cool

Posted at 27 Jan @ 1:09 PM by Andrea Aime | 0 comments