1) JGrass on trunk
2) docs!
3) breakout hack session
4) native code example

(10:01) <chorner> 1) JGrass on trunk
(10:01) <jgarnett> oh can we do one more...
(10:01) <moovida> ok, did you see my mails on the list?
(10:02) <chorner> yes
(10:02) <jgarnett> I did; it made a large amount of sense ...
(10:02) <moovida> is this a migration issue yet?
(10:02) <moovida> or am I doing something wrong?
(10:02) <jgarnett> (and I would like to udate my silly picture to reflect that plan - if we all think it is a good idea?)
(10:03) <moovida> yes, right
(10:03) <moovida> do we think it's a good idea?
(10:03) <moovida>
(10:03) <chorner> pages of interest for those on trunk:
(10:03) <chorner> http://docs.codehaus.org/display/GEOTOOLS/Upgrade+to+2.4
(10:03) <chorner> http://udig.refractions.net/confluence/display/UDIG/GeoTools+2.4+for+uDig+Developers
(10:04) <chorner> c'mon confluence... load that page – you can do it
(10:04) <moovida>
(10:04) <moovida> very nice page (the second is loading)
(10:05) <jgarnett> yeah people are reading them
(10:05) <moovida> I believe it it the bible at the moment, I guess
(10:06) <moovida> but... are the issues I have also in there (in the page that doesn't load)
(10:06) <moovida> ?
(10:07) <chorner> website seems to be down
(10:08) <chorner> some are not there
(10:08) <chorner> we will investigate them tonight in our hack session
(10:08) <moovida> alright
(10:09) * moovida wondering what the hack session is
(10:09) <chorner> unpaid hacking on udig after work
(10:09) <moovida> ahhh, the dream of every developer... I like those
(10:10) <moovida> alright, back to the problems
(10:10) <chorner> your upgrade to trunk might be a little bumpy
(10:10) <moovida> why that
(10:10) <chorner> i believe everything should compile though
(10:10) <moovida> I miss the FactoryFinder
(10:11) <jgarnett> which one?
(10:11) <chorner> not many people are developing against it, since we've been focusing on cleaning up 1.1.x
(10:11) <moovida> Severity and Description Path Resource Location Creation Time Id
(10:11) <moovida> The import org.geotools.referencing.FactoryFinder cannot be resolved net.refractions.udig.project/src/net/refractions/udig/project/internal/impl AbstractContextImpl.java line 37 1182403518933 13546
(10:11) <moovida>
(10:11) <jgarnett> part of the problem is we "skipped" a release (ie 2.3) so many things that we should of had lots of warning (ie deprecations) to change
(10:12) <jgarnett> are hitting us harder (since they are now gone, along with the exact advice on how to switch in the form of the deprecated message)
(10:12) <moovida> I have really few problems that require adapting
(10:12) <jgarnett> FactoryFinder ... thinking.
(10:12) <chorner> FactoryFinder should be covered by that geotools upgrade page
(10:12) <jgarnett> ReferencingFactoryFinder ?
(10:12) <moovida> I just have those three errors in the udig code that I can't solve
(10:12) <moovida> org.geotools.referencing.FactoryFinder
(10:12) <jgarnett> I have not put ReferencingFactoryFinder on their yet
(10:13) <jgarnett> (but if people are using it I can ...)
(10:13) <moovida> hold on a second, I think I miss something
(10:14) <moovida> I checked the code out and did whatever i usually do to set up the source environment in eclipse
(10:14) <moovida> those few errors should appear to everyone, right?
(10:14) <moovida> and udig should not be able to start from source
(10:14) <moovida> is that true or am I going slowly mad?
(10:15) <moovida> (just to be sure we are talking about the same thing)
(10:15) <jgarnett> thinking
(10:15) <jgarnett> I wonder if we need to deploy geotools? chorner?
(10:16) <chorner> richard did it yesterday
(10:16) <jgarnett> did he deploy trunk?
(10:16) <moovida> nothing changed from yesterday night
(10:16) <moovida> had the same problems this morning
(10:16) <jgarnett> let me back up ... moovida are you having this problem on trunk? or on 1.1.x ?
(10:16) <jgarnett> (problem = missing factory finder)
(10:17) <moovida> also this urlToFile is not there
(10:17) <moovida> I'm talking about trunk
(10:17) <moovida> 1.1.x is ok for me
(10:17) <chorner> i see gt2-main-2.4-SNAPSHOT.jar was updated last night at 10 PM
(10:17) <chorner> so richard did deploy
(10:18) <chorner> to get the new jars, you'll have to clean the udig libs plugin
(10:18) <moovida> urlToFile breaks in the geotools, where it can't find it in the DataUtilities
(10:18) <chorner> udig website is back
(10:18) <moovida> I did it at least 6 times
(10:19) <chorner> ok... something may have broken
(10:19) <chorner> it might be best to wait for us to have a look tonight, and in the meantime use 1.1.x
(10:19) <rgould> i deployed 2.2.x and 2.4.x yesterday
(10:19) <moovida> I do clean projects
(10:19) <moovida> refresh libs
(10:19) <moovida> and it downloads the gt2
(10:20) <moovida> after refresh and update classpath
(10:20) <jgarnett> Ah so you did deploy 2.4.x yesterday; I think udig was still using the FactoryFinder as moovida described - we need to change it to ReferencingFactoryFinder (I was not going to deploy geotools until I had uDig sorted out)
(10:20) <moovida> the problem is still there
(10:20) <jgarnett> So you are not insane moovida - just faster than me.
(10:20) <moovida> thanks jgarnett
(10:21) <moovida> but what about the urlToFile?
(10:21) <jgarnett> Back to richard on that
(10:21) <moovida> that is missing in the geotools part, so the libs seem to be out of sync
(10:21) <rgould> aha, so apparently udig was not ready for new geotools jars
(10:22) <rgould> i will set up a trunk workspace quickly and look at the problem
(10:22) <moovida> that would be great, great rgould
(10:22) <rgould> shouldn't take long
(10:22) <rgould> i'm just breaking everything with these fixes
(10:23) <moovida> what do you mean?
(10:23) <rgould> i broke geotools too
(10:23) <moovida>
(10:24) <jgarnett> go richard!
(10:24) <jgarnett> (I won't say where ...)
(10:24) <rgould> haha
(10:24) <jgarnett> okay so we are stuck on some details here
(10:24) <jgarnett> jgrass on trunk is a good idea
(10:24) <moovida> glad to hear that
(10:25) <jgarnett> I want the style editor fixed on trunk (that is what I was waiting for before making the recomendation - and it is the subject of tonights hacking session).
(10:25) <jgarnett> We may ask you guys for a code review tomorrow morning? If that is okay ...
(10:25) <moovida> whatever that means, yes, sure
(10:26) <jgarnett> it means update trunk, and laugh at us on email if we forgot to do our javadocs.
(10:26) <jgarnett> Running the style editor would also be nice.
(10:27) <moovida> yes, absolutely
(10:27) <jgarnett> I think if we can get this; then JGrass is good to go for trunk. We may have a few more bugs (uDig is almost two years behind geotools trunk right now), but we also have much nicer performance.
(10:27) <jgarnett> 2) docs!
(10:27) <jgarnett> This is just quick ..
(10:27) <jgarnett> acuster has been beating me up (in a cheerful manner) over the state of documentation ...
(10:28) <jgarnett> http://docs.codehaus.org/display/GEOTDOC/02+API
(10:28) <jgarnett> I am making a little bit of progress ... you can click on the arrows under the picture and go back and forth between the different geotools "layers"
(10:28) <chorner> yay
(10:28) <jgarnett> including the "JTS and GeoAPI" layer
(10:28) <chorner> hehehe... "you are here"
(10:28) <jgarnett> http://docs.codehaus.org/display/GEOTDOC/01+GeoAPI+and+JTS
(10:29) <moovida> this is very nice
(10:29) <jgarnett> Well I don't know how else to explain it
(10:29) <jgarnett> Well pass thanks on to acuser as well - he has been very helpful.
(10:29) <moovida> thanks Adrian
(10:32) <moovida> anything you wanted to say about that jgarnett?
(10:33) <moovida> anyway also the migration doc is great, thanks
(10:33) <jgarnett> thanks for the reminder, I will make a note on the "proposal" page to update the migration doc.
(10:34) <jgarnett> um this may or may not be old new?
(10:34) <jgarnett> GeoTools has switched to a formal process for chaning the api of anything
(10:34) <jgarnett> so a proposal page is created
(10:34) <jgarnett> that also has the BEFORE / AFTER section
(10:34) <jgarnett> Once the proposal page is accepted it is moved under the release being hacked on
(10:34) <jgarnett> http://docs.codehaus.org/display/GEOTOOLS/2.4.x
(10:35) <jgarnett> You can see a bunch of pages there (some of which were proposals)
(10:35) <jgarnett> example: http://docs.codehaus.org/display/GEOTOOLS/DataSource+Hint+for+EPSG
(10:35) <jgarnett> Has a good BEFORE / AFTER section for the API changes
(10:35) <jgarnett> (so in the future - since geotools is now more organized) the migration instructions will be easier to write.
(10:36) <jgarnett> end rant ...
(10:36) <jgarnett> 3) breakout hack session
(10:36) <jgarnett> we kind of already covered this
(10:36) <jgarnett> Some of us are hacking tonight, since we miss working on uDig, and there is so much cool stuff to get done.
(10:36) <jgarnett> If we can ask for a code review (or even email abuse) after it woudl be great.
(10:36) * moovida would like to add one after this (native code example)
(10:37) <jgarnett> that would be interesting
(10:37) <moovida> yes, but you will hack on what?
(10:37) <jgarnett> chorner anything to add? Basically we are trying to follow the migration guide to update the StyleEditor tonight.
(10:37) <moovida> truk or 1.1.x?
(10:37) <jgarnett> trunk
(10:37) <moovida> great then
(10:38) <chorner> that's it
(10:38) <moovida> alright, very short about native code if you mind
(10:38) * aaim1 has joined #udig
(10:38) <aaim1> Hi udiggers
(10:38) <moovida> I hoped acuster to be here...
(10:38) <moovida> hi aaim1
(10:39) * aaim1 is now known as aaime
(10:39) <chorner> 4) native code example
(10:39) <moovida> I made the example we talked in the last irc
(10:39) <moovida> a plugin that takes a feature from a layer
(10:39) <moovida> extracts its coordinates
(10:39) <moovida> and prints them in C and fortran
(10:39) <moovida> fortran?!?!?!
(10:40) <moovida> so a really awful example of JNI and C2fortran
(10:40) <moovida> it work well, but:
(10:41) <moovida> 1) I made makefiles and everything for linux, since I'm probably not able to do it for windows
(10:41) <jgarnett> wow that is fun!
(10:41) <moovida> 2) I did it with swig and am not able to understand how to pass multidimensional arrays...
(10:41) <moovida> any swig genious here?
(10:41) <moovida>
(10:42) <aaime> moovida, I suggest you ask on the swig ml, or on the gdal one (they use swig quite a bit)
(10:42) <moovida> I guess you are absolutely right,
(10:43) <moovida> but I'm afraid of talking to people that handle such incredible things
(10:43) <moovida> so I hoped in someone "nearer"
(10:43) <aaime> Ah, just be polite and thankful and do not fear anything
(10:43) <jgarnett> um try #osgeo channel
(10:43) <aaime> Oh, and say you read the manual, and did not understood it
(10:43) <jgarnett> they are spatial people, and friendly, and some of the gdal people hang out there
(10:43) <aaime> (so they cannot tell you RTFM)
(10:43) <jgarnett> aaime++
(10:44) <moovida> what is RTFM?
(10:44) <jgarnett> better do a web search on that one.
(10:44) <aaime> Read The Fucked Manual
(10:44) <moovida>
(10:44) <moovida> great!
(10:44) <moovida> i read it and that part says:
(10:44) <moovida> you better take you r C book out of the shelf again
(10:45) <moovida> if you want to pass multi-arrays
(10:45) <moovida> as pointers
(10:45) <aaime> Oh, a suggestion, read the manual of the lastest version, not the stable one
(10:45) <aaime> it has lots more info
(10:45) <moovida> ok, thanks, I'll check that out
(10:47) <moovida> alright, anyone anything else?
(10:47) <jgarnett> thanks for the meeting!