Thanks Richard,
Martin also noted that he could no longer work with the wiki - I sent
email out about this last week - please register with xircles etc...
(the home page has the new instructions).
Martin I am not sure if this new approach (ie xircles is actually
working ...). Can anyone help test/debug this? I am away for the next
two weeks so this is a time limited offer.
Cheers,
Jody
> So it appears I (and thus gtbot) no longer have access to post news items.
> Gtbot will be busted until this is remedied! In the mean time, here are the
> IRC logs:
>
> {excerpt}
> 1: release
> 2: Streaming Renderer
> 3: What about a meeting time one hour earlier?
> 4: introduce gdavis
> {excerpt}
> jgarnett: 1) release
> simboss: ciao guys!
> jgarnett: Not much to say here - I am trying to upload now.
> desruisseaux: Hello Simone
> jgarnett: Thanks to Martin for fixing bugs over the weekend
> desruisseaux: Does the release seems to work Jody?
> jgarnett: It seems to work (even javadocs)
> jgarnett: and stress testing ...
> jgarnett: I will ask someone to download the source release and try it out ...
> desruisseaux: Do you have any pain point that you would like to report?
> jgarnett: The instruction's pages seems to include a lot of duplicated effort
> jgarnett: but it seems much better then trying to do everything in maven
> aaime: What do you mean by duplicated effort in particular?
> aaime: (that is, is it something we can fix?)
> jgarnett: the stress tests are run a couple of times for example
> jgarnett: I think we could start the tag and so on and then start testing
> jgarnett: (maybe check the build box to make sure trunk is good...)
> jgarnett: We can wait to see what the next person thinks?
> jgarnett: Sorry guys - very busy and have not provided useful feedback :-(
> desruisseaux: This is true that the number of time we run the full test suite
> may be reduced (matbe).
> desruisseaux: typo: (maybe)
> aaime: I think the issue with double testing is that
> aaime: you may end up doing a tag and then noticing quite a bit of bugs
> aaime: that need to be backported from tag to trunk
> aaime: How much time does it take to do a super-tested build? (stress and
> whatnot?)
> jgarnett: It did not take that long (so I am not sure it was working?)
> jgarnett: I think it may only be a good test for Cory who has all the online
> tests configured on his box?
> aaime: hmmm... very few tests do separate a "stress" part
> jgarnett: I see; I was under the impression that the build box ran for 40
> mins on this stuff, my box ran for 10 mins .
> jgarnett: (a normal install takes longer)
> aaime: 40 minutes? Man!
> aaime: Hmmm... no, it should take more than a normal build anyways...
> jgarnett: As I said - I think the trick is to have the online tests working -
> so the library is actually tested.
> jgarnett: okay let's leave it - the release goes out today. I will create an
> email about the details so we can think about things for next time.
> jgarnett: can we go on?
> aaime: yup
> jgarnett: 2) streaming renderer
> Jesse_Eichar77: not much to say
> Jesse_Eichar77: just wanted to let everyone know that I've found
> Jesse_Eichar77: that there is a memory leak on trunk.
> Jesse_Eichar77: I'm going to try and track it down.
> aaime: Ouch...
> Jesse_Eichar77: But if anyone notices something or some hints before I get my
> profiler license let me know.
> aaime: Hmmm.... may it be one of the changes I made to get performance back
> to 2.3.x level?
> aaime: I did not stress test it
> Jesse_Eichar77: I haven't tested 2.3.x so I don't know if it exists there.
> Jesse_Eichar77: it occurs after some seriously heavy rendering for quite some
> time. I found it when running my performance tests.
> Jesse_Eichar77: I'll check out 2.3.x to make sure that is clean.
> Jesse_Eichar77: that's about it just thought people would want to know.
> desruisseaux: Item 3: What about meeting time one hour earlier?
> desruisseaux: Do we have IRC peoples in Australia or some region around there?
> jgarnett: aside: uDig on trunk is working and the raster stuff looks very cool
> aaime: Nice
> aaime: +1 btw
> aaime: I would like to meet one hour earlier
> aaime: never asked because of the "Andrea standard time" meeting stuff :-p
> aaime: (that is, I cannot monopolize the meeting time, at least not on
> Geotools)
> jgarnett: I like Andrea standard time
> jgarnett: I am fine with either (Andrea you have to stay up late, if there is
> something we can do to make things easier for those in europe please speak up)
> jgarnett: +0
> aaime: Yep, it's 22.30 here
> desruisseaux: Same for me
> desruisseaux: So is there any objection to a meeting time one hour earlier?
> gdavis: how often do you guys meet like this?
> aaime: Once a week
> gdavis: every Mon?
> aaime: Same channel, same day
> desruisseaux: Yes
> aaime: yeah
> gdavis: okay
> Jesse_Eichar77: fine by me
> desruisseaux: So I will send an email on geotools devel list, if and nobody
> object I will update the wiki before next meeting. Is that okay?
> aaime: Ok
> aaime: next topic?
> gdavis: sure (I think Jody's PC froze on him)
> jgarnett: back
> jgarnett: (fun build process)
> jgarnett: 4) intro gdavis
> gdavis: hey, just wanted to introduce myself. I'm Graham Davis and I work
> with Jody and Jesse and Richard in the same office. I will be working on the
> unsupported geometry stuff in GeoTools for the next couple weeks (or until
> whenever it is fixed and ready). So I'll probably be breaking geotools on
> and off for a bit with my commits ;). Anyways, just thought I'd say hello so
> you know who to blame for the upcoming broken builds.
> aaime: Welcome
> jgarnett: um - let's leave out the breaking the geotools part - you do not
> want to follow the Jody and Cory example that close
> gdavis: hehe
> gdavis: thanks
> aaime: well, everybody breaks the build
> aaime: just make sure not to committ 5 minutes before leaving
> aaime: and if you see the build is broken try to fix it right away :)
> gdavis: that's a good rule, I'll be sure to follow it
> jgarnett: Currently we are looking at how to untangle the
> unsupported/geometry stuff - it needed a review earlier when it was being
> created (a lot of hacks are around because they could not create a CRS, and
> because there are so many factories). Wish us luck
> jgarnett: That is kind of it for meeting topics
> aaime: Good luck :-)
> jgarnett: General discussion? Or let those from Europe go to sleep ...
> aaime: everybody doing what he feels like :)
> jgarnett: bah - README.txt generation is broken
> jgarnett: (sad)
> jgarnett: Well let's chat, and I can turn off gtbot in 20 mins
> jgarnett: ...
> jgarnett: So andrea about your FeatureSource cache thing?
> jgarnett: You know we have a SoC student for that right?
> aaime: Yes, I tried to contact Ian T.
> jgarnett: (doing all their work for them .... suppose it will save you email)
> aaime: He said me, I'll look at the wiki page and let you know
> aaime: and haven't heard back since
> aaime: (was only a few days ago)
> jgarnett: Well we don't have to do anything w/ the students until May 28th ...
> jgarnett: I had some ideas about rendering though ... I added them to the
> wiki page for rendering
> aaime: Anyways, that cache is just a band aid
> jgarnett: (basically I would like to make sure MapContext is used for the 3D
> guys)
> aaime: I mean, it's another stupid "read it all in memory" kind of cache
> aaime: if the dataset is too big, it'll just break
> aaime: I have a better design on that wiki page
> aaime: but implementation is far more expensive
> jgarnett: Understood; I have the feeling that we should do some more stupid
> things like that to make the library more popular
> aaime: Well, the thing is
> aaime: do the stupid thing if you don't have time for the smart one
> aaime: at least do something for the common case
> jgarnett: understood
> jgarnett: (and I tried to make the point that you could be paid - TOPP does
> do some consulting)
> aaime: Hum... yeah, but that's the kind of thing that I don't expect to be
> paid for :) (not the stupid one at least)
> jgarnett: often it is easier to get paid for the stupid one (because someone
> needs it *now*...) and the cost is less
> aaime: Now, if only one of those smart guys that paied 40000$ a full fledged
> Oracle install shelled out some 5000 to fix the datastore, now that would be
> useful!
> jgarnett: aaime++
> simboss: guys
> aaime: yup
> simboss: just in case someone is wondering about where we are
> simboss: we are still alive :-)
> aaime: I had no doubt about it
> simboss: we should get back working on geotools next week
> aaime: Now that's good news
> simboss: actually daniele should start the work for gsoc
> aaime: Hmmm.... basically same timing as Martin working on nd coverages?
> simboss: we would also start talking about revised DataAccess for coverages
> aaime: Same topic, same time: coordination or disaster???
> jgarnett: simboss I may be away the next two weeks
> gdavis: I'm heading out to lunch. bbl, nice meeting you all
> simboss: I would be hoping in some coordination :-)
> jgarnett: but you may have more fun without me
> jgarnett: and the time is actuall -1 hours
> simboss: let's see
> simboss: if he wants some help
> simboss: :-) desruisseaux ping
> simboss: :-)
> desruisseaux: Hello Simboss
> simboss: ciao martin!
> simboss: when are you going to start with the work on ND coverages?
> simboss: we have not talked in the last two weeks
> desruisseaux: This week. But it will take a few days before I have a better
> idea of the kind of work to do. Anyway, I will do everything on Geotools SVN
> and will post email on the geotools mailing list when I will start to commit.
> simboss: any chance we can share some load here?
> simboss: on friday I am going to have a meeting with people who has money
> simboss: they could decide to fund part of the work
> simboss: and daniele is going to work on imageio plugins for new formats
> anyway
> desruisseaux: It is very hard for me to evaluate the work to do before I come
> back to the coverage module (I have been away of this module for a while...)
> simboss: would it be helpful
> simboss: if I start sharing ideas/draft of design/experience/whatsoever?
> simboss: to fire up some discussion
> simboss: (not too much fire though)
> simboss: hello?
> jgarnett: aside: gdavis ping - can you download the src from here
> http://docs.codehaus.org/display/GEOTOOLS/2.4-M1 (and tell me if it builds?)
> desruisseaux: It often help. But really I have been away of coverage for such
> a time that I don't know what is its current state anymore...
> simboss: when you say coverage
> simboss: you do not mean the I/O part
> simboss: right?
> desruisseaux: I'm thinking about both modules/library/coverage and the I/O
> part.
> simboss: k
> simboss: the modules/library part
> simboss: did not change much
> simboss: in general
> simboss: and especially lately
> simboss: my interests
> simboss: are mainly in more formats for I/O
> simboss: so that part is where we would like to work more
> simboss: I am pretty sure
> simboss: first thing will be more HDF
> desruisseaux: This is the place where I expect to spend more time. But
> because the design of this part is new for me, it is hard for me to give
> advice before I looked at the related class starting from the root.
> simboss: np
> simboss: We will start working
> simboss: only inside ImageIO
> simboss: we are seeking to extedn the work for HDF
> simboss: and hopefully other formats
> jgarnett: My interest is in making something to replace GCE
> simboss: same here
> simboss: if we get some money to do it of course :-)
> jgarnett: (because I do not want to document something we are going to
> remove, and parametergroup is hard to use :-) )
> simboss: (we should get it :-) )
> jgarnett: cool
> jgarnett: if you need a letter of support from the PMC then ask (if it would
> make your hunting go better)
> simboss: I like parametergroup :-|
> desruisseaux: I promise to send email to geotools mailing list as soon as I
> will start to have some proposal idea for GCE replacement.
> simboss: I am a JAI guy :-)
> jgarnett: also refraction's may be able to provide a letter
> simboss: letter?
> jgarnett: I like parameter group for some things; but building a (dynamic)
> user interface with it is hard
> jgarnett: I like it for connecting via source code
> jgarnett: (although it really hurt updating udig because of parameter group
> breakage ...)
> desruisseaux: Jody, could you give us an example of pain point with
> ParameterGroup? We could try to fix it...
> jgarnett: Sometimes if you are going after a contract; client's like to see a
> letter of support.
> jgarnett: martin I have to run off to another meeting; for an example of the
> difficulity the WMS bridge code (from WMS to GCE) was hard to write.
> jgarnett: It is easier to make up parameter group entries for known things in
> file format header files, and harder to do it "on the fly" based on a web
> service ... or anything else dynamic ...
> jgarnett: (bah ... I best go ... mind stopping gtbot when you are done?)
> aaime: np
> jgarnett: cheers!
> simboss: anyway, let's try to cooperate as much as possible
> simboss: so that we try to achieve more
> CIA-5: 03desruisseaux * r25389 10geotools/gt/modules/ (12 files in 8
> dirs): GEOT-1160: Make IdentifiedObjectFinder public, so we can keep
> AbstractAuthorityFactory a little bit simplier (less methods there).
> Jesse_Eichar77: Hey simboss.
> Jesse_Eichar77: Thanks for your work on the GC stuff.
> simboss: ciao jesse
> Jesse_Eichar77: Got it going again with uDig... And I'm impressed again.
> simboss: ?
> Jesse_Eichar77: What an improvement over 2.2.x
> simboss: I thought you were making fun of me :-)
> Jesse_Eichar77: (remember I am from the dinosaur age)
> simboss: ah ah
> desruisseaux: I will go. Bye all
> Jesse_Eichar77: I'm going to make a MileStone Release for uDig trunk, just
> for fun, so people can play with it.
> simboss: ciao martin
> Jesse_Eichar77: ciao martin.
> Jesse_Eichar77: too slow.
> simboss: I will try to play quickly with it
> simboss: I have a bunch of new plugins to try
> Jesse_Eichar77: I'll let you know. I was especially happy though because I
> was able to render Tiffs on my Mac.
> Jesse_Eichar77: I've never been able to do GC stuff until now on my mac
> simboss: that's cool
> Jesse_Eichar77: Just thought you'd like to know that I'm playing with it
> again.
> simboss: I have soon to buy a mac myself
> Jesse_Eichar77: And more seriously this time.
> simboss: to test the work we are doing with gdal
> simboss: we test on linux and win
> simboss: but never on mac
> Jesse_Eichar77: Well there are JAI problems with Macs since the binaries
> aren't made for Macs.
> Jesse_Eichar77: So it is a little tricky to do what you've been doing.
> simboss: on jai
> simboss: you should go pure java
> simboss: right?
> Jesse_Eichar77: I'm trying... But All the obvious downloads have binaries.
> For the dev version there aren't classpath installs that I found.
> simboss: mmmhh
> simboss: I would suggest two paths
> simboss: 1
> simboss: look under the binary builds
> simboss: for the pure java zip
> simboss: 2 ask the mailing list
> simboss: I have seen a discussion topic JAI on MAC
> simboss: lately
> simboss: I did not follow it
> simboss: though
> simboss: as I don't have a MAC
> simboss: :-)
> Jesse_Eichar77: I have too. Geotools didn't build on a mac until recently.
> Jesse_Eichar77: but things are looking a lot better.
> simboss: cool
> simboss: next round, if ever make it to the end, should be much much better
> simboss: I have been compression 2gb geotiff in jp2 and ecw lately
> simboss: without any problems
> simboss: it takes 15 minutes though
> simboss: :-)
> simboss: gotta run for a while
> simboss: be back later
> simboss: ciao guys
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Geotools-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/geotools-devel
>
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel
