[01:13:33 EST(-0500)] * esm (n=esm@207-53-193-105.dynamic-dsl.qis.net) has joined ##uportal
[09:33:31 EST(-0500)] * esm (n=esm@asdf.dkc.jhu.edu) has joined ##uportal
[09:55:27 EST(-0500)] * clown (n=clown@guiseppi.atrc.utoronto.ca) has joined ##uportal
[10:00:25 EST(-0500)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:58:31 EST(-0500)] * athena7 (n=athena7@lumina.its.yale.edu) has joined ##uportal
[11:29:11 EST(-0500)] * awills (n=awills@uni1.unicon.net) has joined ##uportal
[11:39:29 EST(-0500)] <EricDalquist> hey there athena7, up3 m5 working for you?
[11:39:40 EST(-0500)] <athena7> yes!
[11:39:42 EST(-0500)] <athena7> looks great
[11:39:45 EST(-0500)] <EricDalquist> cool
[11:39:47 EST(-0500)] <athena7> got portlets working in it
[11:39:56 EST(-0500)] <athena7> i've also got the feedback portlet almost finished
[11:39:58 EST(-0500)] <EricDalquist> if you grab the trunk I just removed the random tag from URLs last night
[11:39:59 EST(-0500)] <EricDalquist> sweet
[11:40:08 EST(-0500)] <EricDalquist> yeah the portlet support is much much more solid
[11:40:18 EST(-0500)] <athena7> yeah everything seems good so far
[11:40:19 EST(-0500)] <EricDalquist> no need to restart the portal to pick up new portlets any more 
[11:40:27 EST(-0500)] <athena7> oh nice!
[11:40:35 EST(-0500)] <athena7> by the way, i could really use some maven development advice
[11:40:48 EST(-0500)] <athena7> i don't know if there's some documentation already available you might recommend
[11:40:52 EST(-0500)] <EricDalquist> feel free to ask questions in here
[11:40:59 EST(-0500)] <athena7> but i'm not really clear on what the best way to actually develop is
[11:41:08 EST(-0500)] <EricDalquist> did you run through the bit of build/deploy docs on the wiki for up3?
[11:41:12 EST(-0500)] <athena7> so if i'm making changes to the bookmarks portlet, for example
[11:41:20 EST(-0500)] <EricDalquist> ah
[11:41:25 EST(-0500)] <athena7> what's the most efficient way to do development?
[11:41:37 EST(-0500)] <athena7> do i have to run a maven build and a portlet deploy every time i make a change?
[11:42:13 EST(-0500)] <EricDalquist> so you're working on the bookmarks portlet in a separate project right?
[11:42:16 EST(-0500)] <athena7> yes
[11:42:48 EST(-0500)] <EricDalquist> once you package it into a .war doing a 'ant deployPortletApp' from the uPortal directory will deploy it
[11:43:19 EST(-0500)] <EricDalquist> to package the bookmarks portlet using maven it is 'mvn package'
[11:43:41 EST(-0500)] <esm> question: if there are threads waiting on a lock, somehwere in the thread dump you woudl expect to see which thread is holding the lock, right?
[11:43:45 EST(-0500)] <athena7> right, all that works
[11:43:50 EST(-0500)] <EricDalquist> esm: yes
[11:44:01 EST(-0500)] <esm> EricDalquist: ok then I have a treat for you
[11:44:10 EST(-0500)] <esm> or i'm just being an idiot
[11:44:15 EST(-0500)] <EricDalquist>
[11:44:26 EST(-0500)] <athena7> i was just wondering if there's a way to do development that doesn't involve packaging, running the ant deployPortletApp task, and restarting - i miss being able to just make changes and have eclipse magically pick them up
[11:44:45 EST(-0500)] <esm> http://uportal.pastebin.com/m50fb2342
[11:44:54 EST(-0500)] <esm> that lock happens to be held by Finalizer.
[11:45:30 EST(-0500)] <EricDalquist> athena7: hrm ... yeah this is a problem with maven ... there isn't an easy way to skip the packaging part of the process
[11:45:31 EST(-0500)] <esm> i had my JVM hang. So fun. Objects coudln't be added to the finalizer queue.
[11:45:46 EST(-0500)] <EricDalquist> esm: sounds like fun
[11:46:05 EST(-0500)] <esm> so why can't i find who holds 0xdfa12d90
[11:46:08 EST(-0500)] <EricDalquist> so previously with ant we would compile & deploy only the changed files directly to tomcat
[11:46:15 EST(-0500)] <athena7> ok, i just wondered if there was something slick that other people were using that i was missing
[11:46:36 EST(-0500)] <EricDalquist> maven will be smart and just compile what is needed when running 'package'
[11:46:46 EST(-0500)] <athena7> and in portlet applications, i didn't have to do any build at all, as long as they were structured right
[11:46:50 EST(-0500)] <esm> http://dspace.pastebin.com/d1616263e
[11:46:53 EST(-0500)] <esm> bbiab lunch
[11:46:59 EST(-0500)] <athena7> that pastebin tool is pretty neat
[11:47:03 EST(-0500)] <EricDalquist> but that packaging step adds extra time since you then need to deploy the webapp to the portal
[11:48:26 EST(-0500)] <athena7> so, once i mavenize the feedback portlet, i think it should be more or less ready for initial sharing
[11:48:43 EST(-0500)] <EricDalquist> athena7: yeah, that is still possible with uPortal 3 just not a portlet that is using maven
[11:48:48 EST(-0500)] <EricDalquist> cool
[11:48:51 EST(-0500)] <athena7> it's in yale's svn right now, but i'm assuming it would make sense to move it to something more shared (and public)
[11:48:56 EST(-0500)] <EricDalquist> yup
[11:49:08 EST(-0500)] <EricDalquist> we have both /portlets/ and /contrib/ in jasig svn
[11:49:21 EST(-0500)] <EricDalquist> so once you're done with it just ping uportal-dev and we can see where it should go
[11:49:28 EST(-0500)] <athena7> ok, that sounds great
[11:49:35 EST(-0500)] <EricDalquist> and if it is being built using Maven including it into a future up3 release is really easy
[11:49:52 EST(-0500)] <athena7> maybe once it's mostly done i'll try and put a war up so people can look at it before making decisions about where it should go
[11:50:21 EST(-0500)] <athena7> i've been updating the documentation in http://www.ja-sig.org/wiki/display/PLT/Unconference+Feedback+Portlet
[11:50:31 EST(-0500)] <EricDalquist> cool
[11:51:04 EST(-0500)] <athena7> we're also about to launch this: http://tp.its.yale.edu/confluence/display/YIP/My+Calendar+Portlet
[11:51:26 EST(-0500)] <athena7> i guess i'll send an email to the uportal list and see if other people are interested in having it distributed more publicly?
[11:51:29 EST(-0500)] <EricDalquist> neat!
[11:51:31 EST(-0500)] <EricDalquist> we could use that
[11:51:42 EST(-0500)] <athena7> yeah i'm excited about it
[11:51:46 EST(-0500)] <athena7> it's read-only
[11:52:08 EST(-0500)] <athena7> but we've got it reading ical feeds, casified ical feeds, and some other custom xml feeds
[11:52:25 EST(-0500)] <EricDalquist> nice
[11:52:26 EST(-0500)] <athena7> in theory you could easily get it to display information from a database, or anything else you wanted
[11:52:41 EST(-0500)] <EricDalquist> we have ical feeds of our campus calendaring system
[11:52:53 EST(-0500)] <athena7> i keep saying i want everyone else to share their local content, so i'm trying to do my part
[11:52:55 EST(-0500)] <athena7> oh neat!
[11:52:59 EST(-0500)] <EricDalquist> yeah
[11:53:06 EST(-0500)] <athena7> we don't really have much in the way of campus calendaring - what are you guys using?
[11:53:12 EST(-0500)] <EricDalquist> oracle calendar
[11:53:24 EST(-0500)] <EricDalquist> a guy here copied google calendars idea of 'big random token' ical urls
[11:53:24 EST(-0500)] <athena7> ah ok
[11:53:40 EST(-0500)] <athena7> right, that seems to be a pretty popular solution these days
[11:53:41 EST(-0500)] <EricDalquist> so I can go to this site and generate a calender url for me
[11:53:44 EST(-0500)] <EricDalquist> yeah
[11:53:48 EST(-0500)] <athena7> so that would probably work well in this portlet
[11:53:57 EST(-0500)] <EricDalquist> gets around the authn problem with web service type APIs
[11:54:04 EST(-0500)] <athena7> yeah
[11:54:14 EST(-0500)] <awills> athena7: are you adding the feedback portlet to the same portletApp as Bookmarks?
[11:54:16 EST(-0500)] <athena7> this will also be our first casified portlet
[11:54:27 EST(-0500)] <EricDalquist> cool
[11:54:34 EST(-0500)] <athena7> no, i messing around with the bookmarks portlet for other reasons
[11:54:49 EST(-0500)] <athena7> i think it would be really neat if the bookmarks portlet could add "remote" read-only folders
[11:54:59 EST(-0500)] <athena7> like rss feeds from del.icio.us or whatever
[11:55:01 EST(-0500)] <awills> ah, too bad – i'd love to see bookmarks evolve into a general-purpose portlet suite
[11:55:25 EST(-0500)] <athena7> then we could define things like "staff links" or "academic links" and have users subscribe to them
[11:55:36 EST(-0500)] <athena7> and whenever we updated those folders, they'd get the updates
[11:55:58 EST(-0500)] <EricDalquist> awills: they aren't really related apps
[11:55:59 EST(-0500)] <athena7> well i wouldn't be opposed to eventually combining some portlets
[11:56:17 EST(-0500)] <athena7> i do worry about what's going to happen to permgen someday, when we have the entire spring library in every single portlet
[11:56:21 EST(-0500)] <EricDalquist> seems to make more sense to only have portlets share a WAR if they actually need to share runtime code/data
[11:56:31 EST(-0500)] <EricDalquist>
[11:56:33 EST(-0500)] <EricDalquist> yeah
[11:56:39 EST(-0500)] <awills> EricDalquist only related in that they're both general-purpose tools that many portals might like to use
[11:56:50 EST(-0500)] <athena7> i think there's a better argument for someday packaging feedback and the impersonation portlet together, like you once suggested awills
[11:56:58 EST(-0500)] <EricDalquist> permgen can get a bit big
[11:57:01 EST(-0500)] <EricDalquist> but really
[11:57:09 EST(-0500)] <EricDalquist> we have servers with 50+ portlet apps in them
[11:57:12 EST(-0500)] <awills> yes, we hit a permgen problem at CO, w/ 12-15 portlet apps
[11:57:17 EST(-0500)] <athena7> then maybe we could do something cool like have a link from a feedback item to actually impersonate that user and see exactly what the user was looking that
[11:57:19 EST(-0500)] <athena7> that'd be pretty neat
[11:57:27 EST(-0500)] <EricDalquist> and we're just fine with 128M of permgen
[11:57:46 EST(-0500)] <athena7> are they all spring-y?
[11:57:55 EST(-0500)] <awills> but the bigger issue w/ putting nearly every portlet into its own WAR is this...
[11:58:07 EST(-0500)] * deuce (n=deuce@uni1.unicon.net) has joined ##uportal
[11:58:15 EST(-0500)] <awills> is increases the burdon of contributing a new portlet to JASIG
[11:58:34 EST(-0500)] <EricDalquist> how does it increase the burdon?
[11:58:55 EST(-0500)] <awills> you have to create a whole new project, provision space in tools like SVN, jira, and confludence, and possibly start a new mailing list
[11:59:11 EST(-0500)] <athena7> well, maybe we can work to clarify and streamline the process of how one might add a new portlet?
[11:59:16 EST(-0500)] <EricDalquist> that is all a symptom of how we currently handle portlet applications
[11:59:21 EST(-0500)] <awills> if there was a suite of smallish, general-purpose tools you could just contribute to that
[11:59:31 EST(-0500)] <EricDalquist> there is no reason we can't create a JA-SIG Portlets Jira project/Confluence space
[11:59:36 EST(-0500)] <EricDalquist> and just sub-divide that
[12:00:13 EST(-0500)] <athena7> actually we do already have a confluence space for jasig portlets
[12:00:18 EST(-0500)] <EricDalquist> yup
[12:00:18 EST(-0500)] <athena7> which is where i added the feedback portlet
[12:00:19 EST(-0500)] <awills> that could be another way... maybe 1 mvn module / WAR, but share mail list and issue tracking, etc
[12:00:24 EST(-0500)] <EricDalquist> yup
[12:00:34 EST(-0500)] <EricDalquist> I think doing a single portlet actually increases the work
[12:00:42 EST(-0500)] <EricDalquist> now you have to make sure the portlet.xml and web.xml are in sync
[12:00:47 EST(-0500)] <EricDalquist> that there aren;'t library versioning issues
[12:01:03 EST(-0500)] <awills> can we transition the bookmarks and the feedback to head towards this model? is that something we intend to do?
[12:01:11 EST(-0500)] <athena7> and for people with different development styles, it may be harder to add to a project - i think you could argue either way
[12:01:28 EST(-0500)] <EricDalquist>
[12:01:29 EST(-0500)] <athena7> sounds good to me
[12:01:30 EST(-0500)] <EricDalquist> yup
[12:01:50 EST(-0500)] <athena7> is there a portlets space in jira?
[12:01:52 EST(-0500)] <EricDalquist> send a note to uportal-dev to continue this talk and I'm sure the changes can be made
[12:01:58 EST(-0500)] <athena7> cool
[12:02:00 EST(-0500)] <EricDalquist> I have to run to a meeting
[12:02:01 EST(-0500)] <awills> i actually may have time for the impersonation portlet this week... so it's a good time for me to ask these sorts of things 
[12:02:09 EST(-0500)] <EricDalquist> oh awills
[12:02:18 EST(-0500)] <EricDalquist> we need to work on portlet entity preferences too
[12:02:20 EST(-0500)] <EricDalquist> ok
[12:02:22 EST(-0500)] <EricDalquist> now brb
[12:02:26 EST(-0500)] <athena7> bye!
[12:02:29 EST(-0500)] <awills> yeah, this week let's do it
[12:02:33 EST(-0500)] <awills> cya
[12:02:44 EST(-0500)] <athena7> awills: the impersonation portlet is on toro portlets?
[12:03:34 EST(-0500)] <EricDalquist> heh
[12:03:37 EST(-0500)] <EricDalquist> meeting was canceled
[12:03:49 EST(-0500)] <EricDalquist> awills: we actually have a idswapper portlet too
[12:04:03 EST(-0500)] <EricDalquist> all this stuff at UW that was never contributed back ... :/
[12:04:07 EST(-0500)] <EricDalquist> and an attribute swapper
[12:04:21 EST(-0500)] <awills> impersonation isn;t in anything... was developed by CO, who wish to contribute
[12:04:38 EST(-0500)] <deuce> eric, i'm working on getting dynamic channel titles to work. can we chat about that sometime today?
[12:04:44 EST(-0500)] <athena7> ah ok
[12:05:01 EST(-0500)] <athena7> ooh is this the dynamic titles for portlets?
[12:05:05 EST(-0500)] <deuce> si
[12:05:15 EST(-0500)] * athena7 cheers
[12:05:18 EST(-0500)] <EricDalquist> deuce: in what version?
[12:05:23 EST(-0500)] <EricDalquist> they work in up3 I believe
[12:05:36 EST(-0500)] <EricDalquist> for channels and portlets
[12:06:05 EST(-0500)] <EricDalquist> I'll be around until about 3:30pm central time today so feel free to chat with me in here until then
[12:06:28 EST(-0500)] <deuce> oh? i hadn't checked up3. I was looking at up26. I'll have to take a look to see how you got around the channel chrome caching issue
[12:06:37 EST(-0500)] <EricDalquist> oh, I probably didn't
[12:06:45 EST(-0500)] <deuce> but i guess the rendering is totally different
[12:07:16 EST(-0500)] <EricDalquist> I unfortunately am not that familiar with the caching mechanisms in the portal framework so I was a bit concerned that some of the changes I made for portlet rendering would break things
[12:07:23 EST(-0500)] <EricDalquist> and the rendering is very similar
[12:07:33 EST(-0500)] <deuce> hmm
[12:07:36 EST(-0500)] <deuce> i'll lookie
[12:07:55 EST(-0500)] <EricDalquist> the only big changes I think were moving all the parameter processing out to a chain before the rendering pipeline ever gets called
[12:08:03 EST(-0500)] <EricDalquist> awills: when would be a good time for you to talk about the entity prefs?
[12:13:29 EST(-0500)] <awills> 3 my time ok? would that be 4 for you?
[12:13:58 EST(-0500)] <EricDalquist> I'm out for the day at 3:30 central time today
[12:14:47 EST(-0500)] <awills> also EricDalquist: how did you get past the missing jaxen dependency (of dom4j) in uP3?
[12:15:04 EST(-0500)] <EricDalquist> ?
[12:15:11 EST(-0500)] <awills> i have to take my car in to service in like 30 min
[12:15:31 EST(-0500)] <awills> http://groups.google.com/group/cernunnos-discussion/browse_thread/thread/9a6be12bfc3a1cec
[12:15:53 EST(-0500)] <awills> i've also googled this issue – seems that it's common
[12:16:29 EST(-0500)] <awills> seems that mvn commonly doesn't pick up jaxen as a transitive dependency of dom4j
[12:16:37 EST(-0500)] <EricDalquist> well it looks like uP3 is pulling jaxen-1.1.1 from somewhere
[12:16:41 EST(-0500)] <EricDalquist> let me check where
[12:17:06 EST(-0500)] <awills> i figured you must have encountered it, and done something to correct it
[12:17:13 EST(-0500)] <EricDalquist> http://developer.ja-sig.org/projects/uportal/3.0.0-M5/uportal-impl/dependencies.html
[12:17:20 EST(-0500)] <EricDalquist> looks like xom depends on jaxen
[12:17:44 EST(-0500)] <EricDalquist> I have no idea what xom is either ... when we converted to maven we just copied over the dependencies that uPortal 2.6 had
[12:17:52 EST(-0500)] <EricDalquist> I'd love to go through and see what we actually need
[12:20:18 EST(-0500)] <awills> EricDalquist i'll contact you when i get back from dropping my car off to see if there's time to coordinate then
[12:20:22 EST(-0500)] <EricDalquist> ok
[12:34:51 EST(-0500)] <deuce> hey anyone know why CSpringPortletAdaptor doesn't get deployed in up3 trunk?
[12:35:22 EST(-0500)] <EricDalquist> it doesn't get deployed?
[12:35:52 EST(-0500)] <EricDalquist> does it not make it into the uportal-impl jar?
[12:36:46 EST(-0500)] <deuce> i got into the jar, but not out to the webapp. wait. does the jar get deployed un-exploded?
[12:37:03 EST(-0500)] <deuce> actually, i checked all jars under tomcat .. no CSpringPortletAdaptor
[12:37:39 EST(-0500)] <EricDalquist> yeah
[12:37:58 EST(-0500)] <EricDalquist> uPortal/WEB-INF/lib/uportal-impl-3.0.0-M5.jar
[12:38:15 EST(-0500)] <EricDalquist> or uportal-impl-3.0.0-RC2-SNAPSHOT.jar depending on which code you have checked out
[12:38:50 EST(-0500)] <deuce> ah the jar didn't get deployed .. but how is any page rendering?? hmm
[12:39:21 EST(-0500)] <EricDalquist> heh ... it should be there that is 100% of the uPortal code
[12:39:32 EST(-0500)] <deuce> hmmm.. it looks like it's exploded
[12:39:44 EST(-0500)] <EricDalquist> interesting ...
[12:39:50 EST(-0500)] <deuce> under WEB-INF/classes/org/jasig/portal
[12:39:56 EST(-0500)] <deuce> but not the CSpringPortletAdaptor class
[12:39:58 EST(-0500)] <deuce> fooey
[12:40:03 EST(-0500)] <EricDalquist> how did you deploy to tomcat?
[12:40:09 EST(-0500)] <deuce> ant initportal
[12:40:13 EST(-0500)] <EricDalquist> hrm
[12:40:26 EST(-0500)] <EricDalquist> oh ... random Java questions
[12:40:34 EST(-0500)] <deuce> hrmm indeed. let me double check my config
[12:40:57 EST(-0500)] <EricDalquist> if I serialize an object under JDK5 ... can I deserialize it under JDK4 if the serialVersionUID didn't change?
[12:41:33 EST(-0500)] <deuce> only if it doesn't contain any container objects
[12:41:46 EST(-0500)] <EricDalquist> hrm ok
[12:41:50 EST(-0500)] <deuce> those uids will most likely have changed
[12:42:10 EST(-0500)] <EricDalquist> I'm pretty sure they didn't in this case but I'll double check
[12:42:31 EST(-0500)] <EricDalquist> we're dealing with a place that we have to provide a serialized java.security.PublicKey to
[12:42:44 EST(-0500)] <EricDalquist> and they require JDK1.4 ... but we're using 1.5 for everything
[12:42:46 EST(-0500)] <EricDalquist> annoying ...
[12:42:54 EST(-0500)] <deuce> well .. any of the properties of the class actually. not just containers
[13:03:47 EST(-0500)] <EricDalquist> well I can deserialize a JDK14 serialized version in JDK15
[13:03:50 EST(-0500)] <EricDalquist> so that will work
[13:13:47 EST(-0500)] <esm> so why can't i find which thread is holding this phantom lock.
[13:13:50 EST(-0500)] <esm> very puzzling to me
[13:14:01 EST(-0500)] <EricDalquist> does the dump include the actual finalizer threads?
[13:14:07 EST(-0500)] <esm> yeah
Page Comparison
General
Content
Integrations