uPortal IRC Logs-2007-05-09

[00:29:36 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal
[08:58:35 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal
[09:57:32 EDT(-0400)] * jayshao (n=jayshao@jshao.rutgers.edu) has joined ##uportal
[10:12:02 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:20:48 EDT(-0400)] * esm_ (n=esm@asdf.dkc.jhu.edu) has joined ##uportal
[10:26:26 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:46:51 EDT(-0400)] * pberry (n=pberry@waldorf.CSUChico.EDU) has joined ##uportal
[12:41:08 EDT(-0400)] <EricDalquist> grrr
[12:41:20 EDT(-0400)] <EricDalquist> the maven ant tasks have a nasty little bug
[12:41:53 EDT(-0400)] <EricDalquist> the artifact:dependencies task fails to include -SNAPSHOT dependencies in the generated path
[13:32:38 EDT(-0400)] <esm_> interesting. thats a very old bug
[13:32:44 EDT(-0400)] <EricDalquist> yeah
[13:32:46 EDT(-0400)] <EricDalquist> and still not fixed
[13:32:51 EDT(-0400)] <EricDalquist> well as far as I can tell
[13:33:08 EDT(-0400)] <EricDalquist> I'm trying to build the 2.0.6 ant plugin from svn but not having any luck getting the built jar to work
[13:34:00 EDT(-0400)] <esm_> jason moved it from MNG-1408, and a bunch of field values got dropped from the issue
[13:34:09 EDT(-0400)] <esm_> including the fix version
[13:34:16 EDT(-0400)] <esm_> it was set to fix version 2.0.x
[13:34:38 EDT(-0400)] <esm_> there are no svn commits for it either
[13:35:01 EDT(-0400)] <EricDalquist> yeah :/
[13:35:06 EDT(-0400)] <EricDalquist> and a lot of votes
[13:35:07 EDT(-0400)] <esm_> of course the svn plugin may not be smart enough to index the issue's old issue ids.
[13:38:36 EDT(-0400)] <esm_> at least those two artifacts should be released as not snapshots
[13:38:42 EDT(-0400)] <esm_> sooner rather than later
[13:38:53 EDT(-0400)] <EricDalquist> yeah
[13:38:59 EDT(-0400)] <EricDalquist> that will be an eventual fix
[15:17:10 EDT(-0400)] <EricDalquist> hey esm_ I'm going to do a release of the BookmarksPortlet
[15:17:16 EDT(-0400)] <EricDalquist> any tips for using the release plugin?
[15:28:27 EDT(-0400)] <esm_> no, not really!
[15:28:38 EDT(-0400)] <EricDalquist> so I run release:prepare
[15:28:40 EDT(-0400)] <esm_> it should theoretically go smoothly
[15:28:44 EDT(-0400)] <EricDalquist> then release:perform
[15:28:56 EDT(-0400)] <esm_> make sure you don'thave dependencies on snapshots, and that you have the correct tagBase defined
[15:29:44 EDT(-0400)] <esm_> yes, just run 'release:prepare' followed by 'release:perform'
[15:29:52 EDT(-0400)] <EricDalquist> ok
[15:30:29 EDT(-0400)] <EricDalquist> here goes prepare
[15:30:48 EDT(-0400)] * esm_ crosses fingers
[15:33:11 EDT(-0400)] <EricDalquist> looks like that worked ok
[15:33:16 EDT(-0400)] <EricDalquist> http://developer.ja-sig.org/source/browse/jasigsvn/portlets/BookmarksPortlet
[15:33:26 EDT(-0400)] <EricDalquist> now to do the perform
[15:33:50 EDT(-0400)] <esm_> sweet.
[15:34:04 EDT(-0400)] <EricDalquist> although I realized I named this rel tag differently
[15:34:12 EDT(-0400)] <EricDalquist> used . instead of - in the version #
[15:34:15 EDT(-0400)] <EricDalquist> oh well
[15:35:10 EDT(-0400)] <esm_> yeah theroetically you can redo it
[15:35:20 EDT(-0400)] <EricDalquist> eh
[15:35:37 EDT(-0400)] <esm_> exactly
[15:35:40 EDT(-0400)] <EricDalquist> I say we just stop using - in version nubmers
[15:35:50 EDT(-0400)] <EricDalquist> since we don't have to deal with that cvs limitation any more
[15:36:47 EDT(-0400)] <EricDalquist> hrm
[15:36:55 EDT(-0400)] <EricDalquist> the site build is failing
[15:37:02 EDT(-0400)] <EricDalquist> looks like there may be a bug in my new release
[15:37:06 EDT(-0400)] <EricDalquist> oh well
[15:37:37 EDT(-0400)] <esm_> yeah. that is one thing with the release plugin is that it pushes out the site
[15:37:41 EDT(-0400)] <EricDalquist> yeah
[15:37:43 EDT(-0400)] <EricDalquist> that is fine
[15:37:44 EDT(-0400)] <esm_> which is sometimes not what you want
[15:37:56 EDT(-0400)] <esm_> e.g. if you are building a release candidate
[15:38:00 EDT(-0400)] <EricDalquist> but the fact that it fails I'm just going to leave for now
[15:38:01 EDT(-0400)] <EricDalquist> ah
[15:38:02 EDT(-0400)] <EricDalquist> yeah
[15:38:21 EDT(-0400)] <esm_> yeah i'd leave it alone for now.
[15:38:28 EDT(-0400)] <esm_> if people complain, thats a good thing
[15:38:36 EDT(-0400)] <esm_> it means they've checked out the code and tried to build the site
[15:52:27 EDT(-0400)] <EricDalquist> I want to add two options to the up3 ear deployer and I'm wondering if I should add them in the generic code or in the tomcat specific code
[15:52:35 EDT(-0400)] <EricDalquist> they are 1. deploy exploded
[15:52:46 EDT(-0400)] <EricDalquist> and 2. delete existing first
[15:52:59 EDT(-0400)] <EricDalquist> any thoughts either way?
[15:53:30 EDT(-0400)] <esm_> is there an interface in the generic code?
[15:53:49 EDT(-0400)] * esm_ goes to eclipse
[15:53:51 EDT(-0400)] <EricDalquist> abstract class
[15:54:03 EDT(-0400)] <EricDalquist> I could/should move some of this to interfaces
[15:56:57 EDT(-0400)] <esm_> at first blush i'd add deploy exploded and delete existing first as abstract methods
[15:57:14 EDT(-0400)] <esm_> in AbsEarDeployer and put the impls in the TomcatEarDeployer
[15:58:39 EDT(-0400)] <esm_> adding a flag to deploy war may be enough
[15:59:16 EDT(-0400)] <esm_> boolean exploded
[15:59:25 EDT(-0400)] <EricDalquist> yeah
[15:59:32 EDT(-0400)] <EricDalquist> I think that is what I'll probably do
[15:59:40 EDT(-0400)] <EricDalquist> and add another method for the remove
[15:59:48 EDT(-0400)] <esm_> yeah sounds good
[16:00:25 EDT(-0400)] <esm_> and keeping it abstract seems fine with me
[16:30:14 EDT(-0400)] <esm_> EricDalquist: quick q
[16:30:20 EDT(-0400)] <EricDalquist> k
[16:30:34 EDT(-0400)] <esm_> does mvn -Dtest=LegacyLayoutSourceTest clean test work for you when executed from uportal3-impl?
[16:32:33 EDT(-0400)] * colinclar2 (n=atrcwrk2@142.150.154.155) has joined ##uportal
[16:32:48 EDT(-0400)] <EricDalquist> yes it does
[16:34:05 EDT(-0400)] <esm_> hrm ok.
[16:37:43 EDT(-0400)] * colinclar1 (n=atrcwrk2@142.150.154.155) has joined ##uportal
[16:52:54 EDT(-0400)] * colinclar1 (n=atrcwrk2@142.150.154.155) has left ##uportal
[16:57:33 EDT(-0400)] * jayshao (n=jayshao@jshao.rutgers.edu) has joined ##uportal
[17:03:38 EDT(-0400)] <EricDalquist> I think I enjoy the Java IO apis a little too much
[17:04:17 EDT(-0400)] <esm_> do i want to know why you think that?
[17:04:36 EDT(-0400)] <esm_>
[17:04:42 EDT(-0400)] <EricDalquist> I'm writing code that will extract individual files from a WAR file that is being read from an EAR file
[17:05:08 EDT(-0400)] <EricDalquist> and only the EAR and the end individual files ever get written to disk and the largest chunk ever to be in memory is 4k
[17:06:04 EDT(-0400)] <esm_> cool
[17:09:27 EDT(-0400)] <jayshao> EricDalquist - I agree... I think you enjoy the Java IO apis a little too much...
[17:09:49 EDT(-0400)] <jayshao> esm_ what's the _ convention mean in IRC?
[17:10:18 EDT(-0400)] <esm_> jayshao: I just am logged in twice to the irc server, so my second client appends the underscore
[17:10:31 EDT(-0400)] <esm_> to avoid nickname collision
[17:12:11 EDT(-0400)] <jayshao> ahhh... logged in from 2 places?
[17:12:22 EDT(-0400)] <jayshao> I only see the one username in the room...
[17:12:36 EDT(-0400)] <esm_> yeah. well i think my other client has me connected to the server, just not in the room
[17:12:49 EDT(-0400)] <EricDalquist> jayshao: do /whois esm
[17:12:54 EDT(-0400)] <EricDalquist> you can see the other client
[17:13:03 EDT(-0400)] <esm_> nicks are in a "global" namespace not scoped to a particular room
[17:13:23 EDT(-0400)] <jayshao> interesting...
[17:13:56 EDT(-0400)] <esm_> castor 1.1.1 pom sucks
[17:14:01 EDT(-0400)] <jayshao> shrug I know next to nothing about IRC, I just hang out here to rub elbows with you guys
[17:14:08 EDT(-0400)] <esm_> ahahha
[17:15:24 EDT(-0400)] <EricDalquist> oi
[17:15:35 EDT(-0400)] <EricDalquist> esm_: your assembler tests really need to clean up after themselves
[17:15:49 EDT(-0400)] <esm_> EricDalquist: in pluto or in up
[17:15:51 EDT(-0400)] <esm_> i've tried.
[17:16:15 EDT(-0400)] <esm_> i'm not sure sure what is up with them. I have them set to delete the files it creates, unless I"ve missed something
[17:16:38 EDT(-0400)] <esm_> shoot me the test name I will dig into it again
[17:16:50 EDT(-0400)] <EricDalquist> don't know I just peeked in my temp dir
[17:16:55 EDT(-0400)] <EricDalquist> and there are a lot of assembly dirs
[17:16:56 EDT(-0400)] <EricDalquist>
[17:16:57 EDT(-0400)] <esm_> ok.
[17:16:59 EDT(-0400)] <esm_> yeah
[17:17:05 EDT(-0400)] <EricDalquist> no biggie
[17:17:14 EDT(-0400)] <esm_> well, yeah it is dirty
[17:17:16 EDT(-0400)] <EricDalquist> just something I can give you a hard time about
[17:17:37 EDT(-0400)] <esm_> ahahaha man, you'll find that there are plenty of things you can give me a hard time about
[17:17:46 EDT(-0400)] <esm_> just wait...
[17:21:52 EDT(-0400)]

[17:22:07 EDT(-0400)] <EricDalquist> that will delete the directory and all children
[17:22:22 EDT(-0400)] <esm_> ok cool I will try that
[17:22:39 EDT(-0400)] <esm_> I will clean up the tests, just as soon as I get the assembly stuff all together
[17:22:50 EDT(-0400)] <esm_> putting everything under one dir is a good idea
[17:23:08 EDT(-0400)] <EricDalquist> np
[17:23:14 EDT(-0400)] <EricDalquist> how goes the assembly stuff?
[17:23:27 EDT(-0400)] <EricDalquist> I'm going to be starting on domain object/hibernate re-working tomorrow
[17:24:36 EDT(-0400)] <esm_> good. the tests that are failing that i asked you about earlier have something to do with dependency changes in pluto
[17:24:45 EDT(-0400)] <esm_> i've been ignoring them
[17:24:48 EDT(-0400)] <esm_> but now i need to fix them
[17:24:56 EDT(-0400)] <EricDalquist> ok
[17:25:15 EDT(-0400)] <esm_> i think it has to do with castor, and including an older xml api impl on the test classpath
[17:25:30 EDT(-0400)] <esm_> i get abstractmethoderrors when calling Element.setIdAttribute(string, boolean)
[17:25:53 EDT(-0400)] <esm_> in a couple of the layout manager tests
[17:26:25 EDT(-0400)] <EricDalquist> uhg
[17:26:29 EDT(-0400)] <EricDalquist> caster == evil
[17:26:32 EDT(-0400)] <EricDalquist> castor*
[17:26:47 EDT(-0400)] <esm_> yeah. i put my assembly code under yourkit
[17:26:54 EDT(-0400)] <esm_> it is insane how much garbage it generates
[17:27:10 EDT(-0400)] <EricDalquist> it will be good once we're on the jaxb stuff
[17:27:11 EDT(-0400)] <esm_> it is not cheap to instantiate the pluto descriptor api services
[17:27:24 EDT(-0400)] <esm_> sigh.
[17:27:30 EDT(-0400)] <esm_> i have my doubts about jaxb
[17:27:56 EDT(-0400)] <esm_> i made the descriptor impls nicer, but I took the optimizations back out.
[17:28:01 EDT(-0400)] <esm_> i'm not sure its worth the complexity
[17:29:59 EDT(-0400)] <esm_> one of the problems with jaxb is, well, xml schema
[17:30:15 EDT(-0400)] <esm_> and your build process complicates b/c every change to a schema requires a re-compile
[17:30:39 EDT(-0400)] <esm_> but i think the biggest issue is that you can be tempted to tightly couple
[17:30:46 EDT(-0400)] <EricDalquist> well hopefully the portlet schema doesn't change that often
[17:30:53 EDT(-0400)] <EricDalquist> ?
[17:31:11 EDT(-0400)] <esm_> e.g. using the jaxb compiled classes to enforce schema restrictions
[17:31:31 EDT(-0400)] <esm_> but anyway
[17:32:18 EDT(-0400)] <esm_> relax ng is so much nicer, syntactially anyway
[17:32:19 EDT(-0400)] <EricDalquist> ah
[17:32:58 EDT(-0400)] <esm_> that's why i like xstream, it is very pojo-ish
[17:33:59 EDT(-0400)] <esm_> but xstream is not a binding framework, where jaxb is
[17:34:14 EDT(-0400)] <esm_> so i guess its applying the right tool to the right problem
[17:37:27 EDT(-0400)] <EricDalquist> do you think I could get similar options (extractWar, removeExistingDirectory) to the pluto assembler?
[17:37:34 EDT(-0400)] <EricDalquist> or is that too much scope creep?
[17:38:37 EDT(-0400)] <esm_> hrm. I don't know...
[17:39:05 EDT(-0400)] <esm_> i don't have a real opinion on that
[17:39:10 EDT(-0400)] <EricDalquist>
[17:39:16 EDT(-0400)] <EricDalquist> ok I'll email the pluto list about it
[17:39:18 EDT(-0400)] <esm_> we have tomcat-specific impls behind interfaces already in pluto
[17:39:27 EDT(-0400)] <esm_> eg for the maven install mojo
[17:40:17 EDT(-0400)] <esm_> i was thinking when adding in ear support that asssembly should almost be its own thing
[17:40:20 EDT(-0400)] <esm_> own subproject
[17:41:08 EDT(-0400)] <esm_> i guess it already is
[17:41:14 EDT(-0400)] <esm_> there's not much other stuff in pluto-util
[17:53:44 EDT(-0400)] <esm_> talk to you all tomorrow, i should be on later finishing up assembly
[17:53:55 EDT(-0400)] <EricDalquist> cool
[17:53:58 EDT(-0400)] <EricDalquist> catch ya later esm_
[17:54:10 EDT(-0400)] <esm_> later
[17:54:13 EDT(-0400)] <esm_>
[19:20:34 EDT(-0400)] * esm_ (n=esm@esm.qis.net) has joined ##uportal
[19:58:44 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal
[20:08:26 EDT(-0400)] * esm_ (n=esm@esm.qis.net) has joined ##uportal
[22:04:49 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal