Topics: 1) style api 2) meeting time 3) Jalopy format
Apparently the meeting time is not set up for most peoples happiness!

• move Forward to be more europe-firendly?
• move Backward to be more asia-firendly?

Here is the chart: worldclock

jgarnett Hi Justin, want to have a meeting?
jgarnett Right now the build is down and I have spent an hour on it w/ no success.
chorner ... just peeking at GridCoverageRenderer... why is the CRS null in the test case? (is this a result of the attribute juggling?)
jgarnett not sure ... at all.
jgarnett So due to attendence - do we want to have a meeting?
jdeolive is there anything to meet about?
jdeolive i guess we could have
jdeolive 1) style api changes update
jgarnett Right now I am not that impressed, I show up at meetings and there is not enough people to vote on things. I send email and there is not enough response to proceed.
jgarnett That is a good one ...
jdeolive yeah i know, and then people complain after the fact
jgarnett James is in a window talking about switching the meeting time to either Asia happy (Martin and David) or Europe happy (Italians?)
jgarnett I think that amounts to -4 hours or +5 hours
jdeolive -4 would be 7:30
jdeolive thats early
jgarnett Well if there are not enough people in north america to float the project we gotta go where the people are ...
jgarnett 2) Meeting time ...
jgarnett The style api changes are here
jgarnett http://docs.codehaus.org/display/GEOTOOLS/Events+for+Styles
jgarnett Although we could add a section on the adding of Defaults that you were doing Justin.
jdeolive sure, however its not complete, i kind of just did it on a "as i needed" basis
jmacgill Hi all
jgarnett Understood, that page just outlines the kind of thing we were doing...
jgarnett Hi James
jdeolive also related to styles, i have finished the gtxml bindings for sld
jmacgill sorry I'm late, this is not a good time of day for me
jgarnett 1) Style api changes update 2) meeting time
jmacgill jdeolive that is excelent news
jgarnett 2 - is because this is not a good time for you
jgarnett ANd I am stuck on filter - but that is being worked on. Tuesday.
jgarnett ALso made FIlter 1.1 SortyBy, and it looks like I can do Filter 1.0, Filter 1.0.20 and Filter 1.1 in GTXML.
jgarnett 3) Jalopy format policy - at least one module wants to opt out...
jgarnett Any thing else on Style api changes? I have sent lots of email, so I hope everyone is in the loop...
jmacgill I'm a little behind on email, but everything seems to be going in the right direction
jgarnett Well there is still one trouble, the policy of adding methods to the interfaces is going to hurt the toolkit long term - every new spec would break client code ... we can damage control with a starter Abstract class. But still ...
jgarnett I am trying the TextSymolizer TextSymbolizer2 approach -
jgarnett we will see how it goes.
jgarnett Justin any thoughts on the direction?
jdeolive yeah interface2 and absract class is probably the best bet, a bit confusing though, it would be nice not to be constrained
jgarnett I can't figure out a way to do better - until Java 5.
jgarnett Note writing code for others is always a constraint
jdeolive how will java 5 help with out api change policy?
jmacgill whats teh xxx2 naming mean?
jgarnett And standards even more so -
jgarnett All weird non standard hacks go in the xxx2
jgarnett stanards in xxxx
jmacgill 2 is a very bad name
jgarnett so you have to cast to xxx2 when you are doing some mad hack thing
jgarnett It is Java tradtion ...
jmacgill says who?
jgarnett Um LayoutManager2
jmacgill Ive been coding Java for 10 years and would not recognise taht as a tradition
jdeolive eclipse also does it alot in there code base
jmacgill oh eclipse does it...
jgarnett http://java.sun.com/j2se/1.4.2/docs/api/java/awt/LayoutManager2.html
jmacgill give me another example
jgarnett normal java does this, when they want to add methods without breaking stuff
jgarnett Graphics to Graphics2d is similar use.
jmacgill its just athwe had SOOO much pain with LiteRenderer and LiteRender2 (where the second was just a n alternative impl)
jgarnett Well this time we have an actual reason ... note we could also use GeoAPI interfaces for the standard
jgarnett and GeoTools ones for the hacks, extentions, non released versions
jgarnett This involves using GeoAPI stroke and so on - but need to fix the feature model first. And understand what happended to SLD 1.0.20?
jgarnett Did it never make it out the gate? Does geoapi need to roll back to 1.0.0 ?
markhamc http://help.eclipse.org/help31/topic/org.eclipse.platform.doc.isv/reference/api/org/eclipse/ui/package-summary.html - Notice IPerspectiveListener and 2 and 3
jmacgill 1.0.20 should have no imact on us at all, it does not officialy exist and any reference to it should die
jgarnett (I have been asking these scalability questions on the GeoAPI list as well)
jgarnett Well GeoAPI is based on it ...
jmacgill and it should not be
jgarnett ... I asked them to kill it but it has some kind of link thing they need.
jmacgill yeah ok, SLD 1.1 will fix that
jgarnett So polically I can probably not kill it until we match SLD 1.1 ?
jgarnett And SLD 1.1 exists ... sigh.
jmacgill anyway, don't let me distract things. If xxx2 makes sense to people then stick with it
jgarnett Q: Does xxx2 make sense to people?
jgarnett (Note this change is in response to the shield labels request of some weeks ago)
jdeolive i havent seen a better way of doing it so yes
jgarnett It makes sense to me +1
jdeolive +1
jmacgill +0
jgarnett And we are stuck ... hense why we should give up on meetings...
jgarnett Did what we do with defaults make sense James - we put them up at the interface level.
jgarnett Storke.STROKE and STROKE_NULL
jdeolive nope
jdeolive Stroke.DEFAULT, and Stroke.NULL
jgarnett Ack that is what I get for talking on my feet
jgarnett 2) Meeting time ...
jgarnett James what time would be good for you, we have little enough attendence that moving times should not impact many people.
jmacgill Lets do this my email, because we need good times for some of the people who are not here
jmacgill s/my/by/
jgarnett 3) Jalopy format policy
jgarnett I would like to leave it unchanged, even though the latest Jalopy seems to crash on me.
jgarnett But email seems to not want to have a format policy at all... if we do have one I want a tool to do it.
jgarnett I don't like fussing over such things on the email list...
chorner justin mentioned maven jalopy, is this an option?
jgarnett Is it - could we have a jalopyAll target?
jdeolive the maven jalopy stuff looks pretty easy, dont think it would be hard to throw in
jdeolive probably make the build even slower
jmacgill I would say that we should encorage formating standards and the use of jalopy. BUT! respect other peoples formating (even if it is wrong). Do not go on reformatting sprees
jgarnett We do have at least one module that wants to opt out - see email.
chorner perhaps we should have a policy of not reformatting AND making code changes? (so the diffs are useful)
jgarnett Something about headers, not being GeoTools generic but puliic domain instead...
jgarnett When I started you would change your code, and run Jalopy right before commiting
jgarnett (or else the module maintainer - usually chris - would yell at ya)
jgarnett (being chris he yelled nicely)
chorner that's fine, if jalopy was run previously
jgarnett So we are kind of stuck, I would like to see it run as a matter of course - would cut down on merge conflicts.
jgarnett Could we turn Jalopy on module by module?
jgarnett module/api would be a great candidate for example ...
jmacgill maven should allow that you can have module level config
jdeolive yup, you coudl easily turn off formatting for a module
chorner let's try it module by module, until the assimilation is complete
jgarnett Could we try Jalopy ON for module/api ?
jmacgill sorry all, gota run
jmacgill with regards time, earler would be better, or quite a bit later
jgarnett Okay - cool.
jgarnett Hi brent - I think you just missed the meeting.
jgarnett Unless you had something for the agenda?
bowens oh
bowens thought it started at 12:30
bowens I can just look at the logs, and address the list after