GeoTools : 5 November 28th

jgarnett Well we can start to set up an agenda, apparently James is trying to attend, this will give us something to do while we wait.
rschulz I could mention referencing progress (Martin is likely still sleeping)
jgarnett That would be nice, I notived EPSG was broken (so parsing has obviously changed)
jgarnett 1) CRS progress
jgarnett 2) Metadata unprogress
jgarnett Hi James
jgarnett we are setting up an agenda (while we waited for you)
jgarnett 1) CRS progress
jgarnett 2) Metadata unprogress
jmacgill Hi all, sorry I;m late - firewall troubles again
jgarnett 3) svn.geotools.org
jgarnett 4) 2.0 status update
jgarnett A quiet - that means we should start (smile)
jgarnett Or do you guys want to hang for a bit?
jyutzler I see no reason to wait
jmacgill lets kick off
jgarnett Alright rschulz that means you are up 1) CRS progress
rschulz ok, the new work is in org.geotools.referencing
rschulz this will replace the classes in org.geotools.cs/ct/pt
cholmes 5) consider moving irc time to a bit later, so that Martin could attend, as Andrea is on a PMC sabbatical.
rschulz martin finished the parser and I will commit finnished map projection transforms tonight
rschulz oh, and see geot-26 for progress here
cholmes awesome.
jmacgill indeed it is
rschulz the only major items left are crs authority factories (me) and the CoordinateOperationFactory (Martin)
rschulz at the current rate, this will hopefully be done by the end of December
rschulz then the task is to update geotools to use the new interfaces
rschulz just wanted to give you guys a heads up about this
rschulz it will probably be a bit of work to track down all the code using the old classes
jmacgill refactorIT -> where used?
rschulz the positive side is that we should be able to parse esri wkt now (projections names can have aliases)
rschulz Ok, that's it
jgarnett Right now CRSService exists for code to be independent of old or new CRS implementation.
cholmes Well, since none of the datastores have really gotten around to adding referencing information it shouldn't be too bad, we'll just use the new code...
jgarnett We can hunt down references early and convert them over to CRSService.
jgarnett Actually chris a fair number have started using CRSService.
jgarnett Hi David
rschulz good, also the old classes are being depreciated as we port them
polio heya
jgarnett One troubling thing is that EPSG has broken on me - with a null point execption.
jmacgill I added suport to Shapefile not long ago... I'll update it to CRSService if you can give me some pointers
jgarnett CRSService is pretty simple - give it a EPSG:# or an AUTO:#,lat,long and it will give you a CRS object.
jgarnett (right now the old implmenetation, although we can cut over to the new at any time)
rschulz I can try looking at the epsg module tonight if I have time
cholmes Cool.
jgarnett Do you need more pointers than that James?
jgarnett I am looking into EPSG right now
jgarnett but the code has not changed to my knowledge (or permission as module maintainer).
rschulz James, you may need to wait until we are done - then you can use the new wkt parser to create CRS objects
jmacgill yeah thats what I need
jgarnett Note the old objects support the CRS interface (which is why CRSService can work as a middle ground).
rschulz jody, martin made some changes to the older cs classes
jmacgill which reminds me - Jody, do you do anyting with the crs info in the shp.xml doucment?
jgarnett No
jgarnett They are using the *.prj file as far as I know
jmacgill I'll rephrase that, sometimes when there is no prj file the xml file could be used instead. How easy would it be to extract that info using the code that you have writen
jgarnett Fairly easy
jgarnett I am using JDOM to do the parsing though - pretty lame but it was quick to write.
rschulz is the crs info in xml?
jgarnett Some of it is
jgarnett Let me find the link ...
jgarnett hrm don't see it rightnow
jgarnett Probably need a more complete example
jgarnett (I was focused on lat/long bounding box at the time.
jgarnett So James, if we can find the information in shp.xml it will be easy to extract.
rschulz a xml to crs object parser can be added to referencing
jgarnett I think it may be easier than that, WKT or a EPSG number.
rschulz good
jgarnett Thank for the update. That serves as a good transition to agenda item 2) Metadata unprogress
rschulz ok, goodbye
jgarnett Darn he was quick :-P was it something I said?
jgarnett Or more likly everyone is sick of metadata.
jgarnett I will make it quick.
polio (smile)
jgarnett Chris has offered to help us "Solve" metadata access in the udig code base, we will get back to the group mid january with a solution.
jgarnett I can't get much quicker than that.
jmacgill (smile)
jgarnett Although with everyones permission I would not mind maintaining some of the email discussion on the geotools list (since this code is intended for geotools).
jgarnett Would that be okay?
cholmes That's great for me.
jgarnett And thanks for stepping forward Chris
cholmes (though there are times where I would like more updates on the uDig list as to what you are doing there, since I never have any idea)
jgarnett (smile)
jgarnett Apparently I don't either (just back from holidays).
polio some of the updates ... or lack there of have been my fault
cholmes You may consider putting more of your discussions up there, so people get the idea that work is being done, and that it is as open as you say.
jgarnett On the bright side we have started updating the http://docs.codehaus.org/display/UDIG/Home page with news items again.
jgarnett (And I want to get those entries CCed to the email list).
jgarnett Confluence skills permitting.
jgarnett But point noted; more email required.
jgarnett (I will send one about Java 5 vs. Mac support this afternoon).
jgarnett 3) svn.geotools.org
jgarnett James an update?
jmacgill easyspace the complany which I have used for 4+ years to manage the geotools.org domain has gone though a major face lift (new web interface for managment)
jmacgill the shift seems to have killed a lot of settings and data
jmacgill including our zone table
jmacgill easyspace have 500,000+ users and I think a large number of them are having trouble
jmacgill basicaly the geotools.org domain could be dead for some time as I am having a hard time communicating with them
jmacgill I'm now penning an email to their parent company in the hope of getting a responce
jmacgill as many of you may have seen jody has put out an email with a workaround for svn
jgarnett James I am going to post the workarounds to the front page of the wiki.
jmacgill and whilst www.geotools.org is down docs.codehaus.org/... will work
jyutzler Sorry, what is a zone table?
polio jody, i'll give you the loc for the linux hosts file
jmacgill we own *.geotools.org
jgarnett Does everyone understand the workarounds (ie did it work for you?)
jmacgill the zone table tells DNS servers where the host machines for different values of * are located
jmacgill e.g. svn.geotools.org modules.geotools.org www.geotools.org
jmacgill as, in fact, each of these is on a different machine
jmacgill without a zone table none of the above addresses will resolve properly
jmacgill so, for now, please use the workarounds and feel free to email the devel list if you are having touble getting themt o work
jgarnett Okay main page is ugly but updated
jgarnett Hi James while we have you on the spot - 4) 2.0 status update?
jgarnett Hmm - he must be busy
jmacgill most of my spare time over the last week has been taken up with shouting at easyspace. 2.0 is almost good to go.
jmacgill now that I have a way into svn I'll setup the tag v.soon
polio I would wait for easyspace to release perhaps?
jmacgill It was thanksgiving here so I spent most of the week at home with wife and doughter
polio thinking of how potentail users would get at the fresh version
jmacgill polio yeah, esp as www.geotools.org is down
jgarnett So to back up what is the "almost" in ready part of geotools 2.0?
jgarnett I understand it had trouble building for some people?
jmacgill that is fixed
jmacgill the only things that are left are some of the more obscure maven targets
jmacgill like maven site
jmacgill that past users have tried and point out that they fail
jgarnett We can remove those for the release.
jgarnett ...can't we?
jmacgill Yes, and thats my plan if a fix takes more than 5mins
jmacgill there are some references to geotools-src left in the build script, thats all.
jgarnett So James is there anything more we can help you with w/ respect to 2.0? Just try out the next sanity check.
jmacgill no there is nothing else that I need
jgarnett Well that was the agenda
jgarnett may be a record for a quick meeting (smile)
jgarnett Is there anything else before we get back to it?
jmacgill Feature and GeoAPI?
jgarnett Ah
jgarnett darn - some content.
jgarnett Should we ask for another GeoAPI meeting?
cholmes I added a 5) of considering moving IRC to a later time, so that Martin might be able to show up.
jgarnett Oh right
jgarnett I am all for it
cholmes As Andrea isn't doing PMC for a bit.
jgarnett what do people think?
cholmes I could do an hour or two later...
jgarnett Or to rephrase what time could we move it to that Martin could attend? I recall he often shows up around 2:30 local time.
cholmes It's unclear how much later we would need it for martin to show up.
jyutzler For me, later is better. How late is too late for the Europeans among us?
jmacgill an hour later is easy for me
jmacgill two hours starts to confict with heading home time.
jgarnett Lets email the list (and Martin) and see what people think?
cholmes Well, the current time is really sorta too late, the suggestion to move it later was because the european contingent isn't too strong with Andrea not showing up for a bit.
cholmes And Martin is doing a bunch of work these days, so it'd be nice to talk to him directly
jgarnett I am also infavour of breakout IRCs, something we did to communicate to Australia.
jgarnett Chris would you like to write up an email? May want to talk to martin first to see what time would actually work for him first.
cholmes Yeah, we could almost designate monday 'irc day', everyone tries to stay on the whole day in their time zone, have main meeting at some point, but let discussion flow throughout the day...
jgarnett Owe
cholmes It's getting late for me, I was going to cut out right after the meeting, but I could write it tonight if no one else has a chance.
jgarnett I can write the email
cholmes cool.
cholmes thanks
jgarnett A whole day would not work at all for me (it would prevent me working on anything very well)
jgarnett I will ask martin first.
polio one idea that came to mind though it to publish a list of im contact info for module maintainers though
jgarnett Sent email to martin, will send email to list when he gets back to me.
jgarnett Now that is a smart idea.
jmacgill Yeah, I like that
jgarnett May be too much love for some module maintainers- IM can be interuptive.
jgarnett But yeah - can we add that to the maven information somewhere?
cholmes I'm for im info for mod. maintainers.
jgarnett James is this possible w/ maven
jmacgill should be
jmacgill let me check what is possible
jgarnett Anything more guys?
jgarnett I am going to go prep for a job review.
jmacgill developers have a URL slot in the project.xml file
jmacgill that might be something we could abuse for this purpose
polio we could just put it in as xml comments too ... by convention