1: feature model changes for 2.4
2: More Filter updaters volunteers required
jdeolive: 1: feature model changes for 2.4
jdeolive: i am ready to commit my changes to the feature model api
jdeolive: which is the last blocker for me against 2.4
jdeolive: i have the chagnes local and have run tests, and geoserver cite tests
jdeolive: all is good
jdeolive: minus one patch i am currently waiting and discussing with gabriel
jdeolive: so... as per discussion on the list
groldan: the one on DataUtilities?
jgarnett: nice work; Andrea and myself had a look at Jira - it should now reflect a bit better what is needed for 2.4.
jdeolive: yes, and FilterAttributeExreactor
jdeolive: so this will break geoserver and udig
jdeolive: well maybe not udig, i dont know if it directly subclasses Feature or FeatureCollection
groldan: just go for it, I got just confused by the lack of evidence on the methods intent
jdeolive: but geoserver does and it broke
jdeolive: ok cool... i will commit a unit test for it as well
jdeolive: chorner, does udig implmeent Feature or FeatureCollection directly?
jdeolive: ok, i get it lurk mode... will send email
jdeolive: ok, thats it for me on 1 then
jgarnett: There is at least one Feature implementation (that adapts to layer, georesource etc...) - pretty sure it is a pure wrapper though (since FeatureFactory is not injectable yet)
groldan: wondering, do you have the community-schema modules loaded on your eclipse workspace?
jdeolive: yes
jdeolive: have changed the GTAdapters and stuff
groldan: cool
jdeolive: although i never did the geoserver part of it
jdeolive: i can do that thouhg, not many compile errors
groldan: so what's the next step? I'm quite ignorant on what your end goals are (sure its just me not looking for in the right place, but you know how it is when there's too much stuff to come up with)
jdeolive: yup
jdeolive: well most of the info is here
jdeolive: http://jira.codehaus.org/browse/GEOT-1191
jdeolive: but basically i am just making the api changes i need to in order to allow the old model to implement the new
jdeolive: and making the changes before 2.4 goes out so we can have a clean cycle of deprecattion
jdeolive: so basically this task here
jdeolive: http://jira.codehaus.org/browse/GEOT-1372
groldan: thats cool (even if I'm still hesitant on the approach)
but I wonder if theres a plan to actually use complex features in the mid termgroldan: or do we just want to stick with SimpleFeature?
jdeolive: not really... my goal is to just switch client code to SimpleFeature
jdeolive: complex features is the next major milestone after that
groldan: k, I'll try to keep an eye on the topic, though not sure I could do that much over the next month as I'm flying back home
jgarnett: justin I made a picture of what was coming up; can you check it over and make sure it agrees with what you just said?
jdeolive: sure...
jdeolive: link?
jgarnett: http://docs.codehaus.org/display/GEOTOOLS/2007+Q2
jdeolive: ahh, this picture again
jdeolive: anyways, gabriel, we can move onto your issue
jdeolive: sorry, agenda item
jgarnett: (Also note the home page, and the 2.3.x and 2.4.x page now have a slightly better list of bugs)
jdeolive: 2: More Filter updaters volunteers required
groldan: ok
groldan: as said on the ml, I got stuck for a good while on that bug that raised on IndexedShapeFileDataStore so didn't advanced far beyond updating main to geoapi filters
groldan: that said, most of the work shall be quite mechanical
jdeolive: yeha, its no easy task
jdeolive: i have an idea
groldan: but there might be a couple conflictive pieces, of the one I can remember off the top of my head...
jdeolive: for the grunt feature model i plan to utilize the foss4g code sprint
jdeolive: we could possibly include switch to filter in that mix
groldan: somewhere we need to know the functions arg count up front...
jdeolive: yeah i had this issue
jdeolive: and argued with jgarnett about it for quite some time
groldan: complementary idea: out of library (ie, in plugins, unsupported, etc) we can use CQL as a test dependency to simplify unit tests (lots of them build filters)
jdeolive: its a major issue blocking full acceptance by geoserver
jdeolive: thats a good idea
jgarnett: justin are you going to organize the geotools code sprint then?
groldan: if andrea were here he would argument CQL still don't support Id filters... yet it could be useful and time saving for any non Id filter building
groldan: ha, talking about rome... hi andrea!
aaime: hi
jdeolive: jgarnett i plan to yes
groldan: so, if filter gets worked out at the code sprint, we still have the issue of migrating to SimpleFeatureBuilder. And the code spring being next month, it doesn't get ready for this month's release?
groldan: (code sprint I mean)
jdeolive: yeah... we agreed to move back the requirement to move over to SimpleFeatureBuilder
jdeolive: for 2.4
jdeolive: and push it off to 2.5
groldan: ok, sad thouhg, guilty me
jdeolive: so is that it, anything else fo rthe meeting?
groldan: guess not
jdeolive: ok, lets call it then
jdeolive: thanks guys
iant_: bye
Labels