Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 91 Next »

[08:11:07 EDT(-0400)] * colinclark (n=colin@bas1-toronto09-1279543721.dsl.bell.ca) has joined ##uportal
[08:27:31 EDT(-0400)] * athena7 (n=athena7@c-24-2-189-230.hsd1.mn.comcast.net) has joined ##uportal
[08:37:24 EDT(-0400)] * colinclark (n=colin@bas1-toronto09-1279543721.dsl.bell.ca) has joined ##uportal
[09:12:41 EDT(-0400)] * anastasiac (n=team@142.150.154.105) has joined ##uportal
[09:30:40 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[09:41:18 EDT(-0400)] <athena7> anyone here tried out the duke facebook portlet recently?
[09:43:11 EDT(-0400)] <EricDalquist> nope
[09:43:16 EDT(-0400)] <athena7> ah ok
[09:43:20 EDT(-0400)] <athena7> trying to get it working
[09:43:21 EDT(-0400)] <athena7> we'll see
[09:48:36 EDT(-0400)] <athena7> i'm not really sure what's going on, but it looks like it's not even getting to the first line of the controller
[09:48:37 EDT(-0400)] <athena7> ugh
[09:53:34 EDT(-0400)] <Tuomaz> I'm trying all of the channels/portlets we are currently using in uPortal 2.5 in uPortal 3.0.0
[09:54:06 EDT(-0400)] <athena7> how's it been going so far?
[09:54:19 EDT(-0400)] <athena7> i've worked through a bunch of the CAnnouncements issues if you're using that one
[09:54:28 EDT(-0400)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[09:54:31 EDT(-0400)] <Tuomaz> we aren't using that one
[09:54:36 EDT(-0400)] <Tuomaz> but:
[09:54:44 EDT(-0400)] <Tuomaz> we are using CStockage (http://www.esup-portail.org/Projets/Fichedescriptive/?contentId=183)
[09:55:06 EDT(-0400)] <Tuomaz> CStockage needs commons-httpclient
[09:55:11 EDT(-0400)] <Tuomaz> of version 2.x
[09:55:35 EDT(-0400)] <Tuomaz> uPortal 3 ships with the 3.0 series of commons-httpclient
[09:55:39 EDT(-0400)] <EricDalquist> ah the joys of channels having to run in the same classloader (tongue)
[09:55:49 EDT(-0400)] <Tuomaz> yes, very bad!
[09:56:17 EDT(-0400)] <EricDalquist> unfortunatly Tuomaz I think your best bet is going to be to fix CStockage to be able to use the httpclient 3.0 codebase
[09:56:20 EDT(-0400)] <Tuomaz> the 3.0 -series is incompatible with esup's webdav-server (based om apache slide)
[09:56:33 EDT(-0400)] <EricDalquist> heh
[09:56:36 EDT(-0400)] <Tuomaz> EricDalquist: that was my first thought alsp
[09:56:37 EDT(-0400)] <Tuomaz> also
[09:57:49 EDT(-0400)] <Tuomaz> the correct way to do it should be to fix the server
[09:58:05 EDT(-0400)] <EricDalquist> heh yeah
[09:58:17 EDT(-0400)] <Tuomaz> but that feels overwhelming
[09:58:17 EDT(-0400)] <EricDalquist> so their webdav server doesn't play nice with httpclient 3.0?
[09:58:28 EDT(-0400)] <EricDalquist> then you can use httpclient 2.0
[09:58:57 EDT(-0400)] <Tuomaz> Their webdav server is based on slide with heavy modifications
[09:59:06 EDT(-0400)] <Tuomaz> and now slide is deprecated
[10:00:43 EDT(-0400)] <Tuomaz> my ugly ugly hack at the moment is to replace the commons-httpclient in up3 with the old version and extract one class from the new jar that is needed by uPortal
[10:00:46 EDT(-0400)] <Tuomaz> it works
[10:01:59 EDT(-0400)] <EricDalquist> heh
[10:02:08 EDT(-0400)] <Tuomaz> ConnectTimeoutException
[10:02:10 EDT(-0400)] <Tuomaz> is the class
[10:02:19 EDT(-0400)] <EricDalquist> that just sounds like a recipe for hurt down the road
[10:02:22 EDT(-0400)] <Tuomaz> yes
[10:03:44 EDT(-0400)] <Tuomaz> What are the alternatives? I have something about a briefcase portlet?
[10:03:52 EDT(-0400)] <Tuomaz> have heard
[10:04:21 EDT(-0400)] <EricDalquist> I don't know, we don't use any sort of web-dav in the portal feature
[10:05:43 EDT(-0400)] <athena7> i know there are a couple briefcase portlets out there
[10:05:50 EDT(-0400)] <athena7> there's some stuff listed on the jasig portlets page
[10:05:59 EDT(-0400)] <athena7> and i think there might be one in the toro project too?
[10:06:03 EDT(-0400)] <athena7> i haven't tried any of them
[10:27:14 EDT(-0400)] <Tuomaz> Seem like there is very low activity in toro-portlets project?
[10:30:29 EDT(-0400)] <EricDalquist> I just had fun explaining why 'solaris tar is bad' to a manager her
[10:38:26 EDT(-0400)] <athena7> i think some of the projects are things that have been donated to the open source world post-development
[10:46:09 EDT(-0400)] * michelled (n=team@142.150.154.197) has joined ##uportal
[10:54:44 EDT(-0400)] * esm (n=esm@207-53-193-157.dynamic-dsl.qis.net) has joined ##uportal
[11:54:11 EDT(-0400)] <athena7> eric is the webproxy portlet documentation on the jasig wiki likely to be up to date?
[11:54:23 EDT(-0400)] <EricDalquist> it is the best that is out there
[11:54:34 EDT(-0400)] <EricDalquist> we haven't really touched the code locally since the code dump was done
[11:54:43 EDT(-0400)] <EricDalquist> though I'll be re-visting it in a few months
[11:55:35 EDT(-0400)] * holdorph (n=holdorph@wsip-98-174-242-39.ph.ph.cox.net) has joined ##uportal
[11:55:42 EDT(-0400)] <athena7> ok (smile)
[11:55:43 EDT(-0400)] <athena7> thanks
[11:55:48 EDT(-0400)] <EricDalquist> yup
[12:01:28 EDT(-0400)] * MarkRogers (n=MarkRoge@addhcp129.cc.umanitoba.ca) has joined ##uportal
[12:21:34 EDT(-0400)] <EricDalquist> uhg: http://it.slashdot.org/article.pl?sid=08/05/13/1533212
[12:21:56 EDT(-0400)] <athena7> umm that doesn't sound good
[12:22:25 EDT(-0400)] <athena7> ubuntu too huh?
[12:22:54 EDT(-0400)] <EricDalquist> yup
[12:23:03 EDT(-0400)] <athena7> ugh.
[12:23:05 EDT(-0400)] <EricDalquist> time to go re-distribute my freaking key{color}
[12:23:11 EDT(-0400)] <athena7> that's going to affect a lot of people :/
[12:23:12 EDT(-0400)] <athena7> yeah
[12:23:16 EDT(-0400)] <athena7> that's a major pain in the ass
[12:23:24 EDT(-0400)] <holdorph> unfortunately the ubuntu update broken my package system
[12:23:24 EDT(-0400)] <athena7> i think i generated mine on the mac . . . not sure
[12:23:27 EDT(-0400)] <athena7> oh no
[12:23:31 EDT(-0400)] <holdorph> it's not updating now
[12:23:47 EDT(-0400)] <EricDalquist> holdorph: that's no good
[12:23:56 EDT(-0400)] <athena7> :/
[12:24:08 EDT(-0400)] <holdorph> Preconfiguring packages ...
[12:24:08 EDT(-0400)] <holdorph> openssh-server template parse error: Template #4 in /tmp/openssh-server.template.87580 has a duplicate field "template" with new value "ssh/vulnerable_host_keys". Probably two templates are not properly separated by a lone newline.
[12:24:14 EDT(-0400)] <holdorph> grr
[12:25:41 EDT(-0400)] <athena7> ugh
[12:26:20 EDT(-0400)] <EricDalquist> athena7: once you update they give you a ssh-vulnkey tool to check the keys installed on the machine
[12:26:31 EDT(-0400)] <athena7> ah
[12:26:43 EDT(-0400)] <athena7> is it available for other systems?
[12:26:51 EDT(-0400)] <athena7> i don't have an ubuntu machine anymore
[12:26:55 EDT(-0400)] <holdorph> COMPROMISED: 2048 b9:43:ba:01:4d:b1:43:bf:a3:4b:a4:40:22:19:02:b4 /etc/ssh/ssh_host_rsa_key
[12:26:55 EDT(-0400)] <holdorph> COMPROMISED: 1024 85:2a:54:f7:90:4c:52:8b:90:42:47:f9:3d:fc:7e:7c /etc/ssh/ssh_host_dsa_key
[12:27:00 EDT(-0400)] <athena7> WOPR got a brainwashing before i left
[12:27:23 EDT(-0400)] <EricDalquist> athena7: I don't know
[12:27:29 EDT(-0400)] <athena7> i'll have to look around
[12:29:27 EDT(-0400)] <EricDalquist> brb ... rebooting (tongue)
[12:32:20 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[12:33:00 EDT(-0400)] <athena7> rehi
[12:41:43 EDT(-0400)] <athena7> the parameters for the web proxy portlet are supposed to be channel parameters, yes?
[12:41:52 EDT(-0400)] <EricDalquist> yeah
[12:42:07 EDT(-0400)] <athena7> the names appear to be too long for the up_channel_param.chan_parm_nm field
[12:42:13 EDT(-0400)] <athena7> it's limited to 35 characters
[12:42:16 EDT(-0400)] <EricDalquist> so one issue that makes it hard to use in current uPortal incarnations is use of multi-valued preferences
[12:42:17 EDT(-0400)] <EricDalquist> yeah
[12:42:31 EDT(-0400)] <athena7> the names should be like "edu.wisc.my.webproxy.webproxy.general.config.sBaseUrl", right?
[12:42:36 EDT(-0400)] <EricDalquist> yup
[12:42:53 EDT(-0400)] <EricDalquist> and a real portlet preferences XML block into the .channel files
[12:43:05 EDT(-0400)] <EricDalquist> both do not really support portlet preferences like they should be
[12:43:42 EDT(-0400)] <athena7> yeah
[12:43:56 EDT(-0400)] <EricDalquist> we have some local hacks to work around this
[12:43:57 EDT(-0400)] <athena7> is up_channel_param where they're currently supposed to go though?
[12:44:01 EDT(-0400)] <EricDalquist> no
[12:44:04 EDT(-0400)] <athena7> oh ok
[12:44:09 EDT(-0400)] <EricDalquist> they are portlet definition preferences
[12:44:10 EDT(-0400)] <athena7> so they're supposed to get set as portlet prefs?
[12:44:12 EDT(-0400)] <athena7> ok, thanks
[12:44:13 EDT(-0400)] <EricDalquist> that is the channel manager hack
[12:44:17 EDT(-0400)] <athena7> that's what i was trying to figure out
[12:44:21 EDT(-0400)] <EricDalquist> portlets are not aware of channel parameters at all
[12:44:27 EDT(-0400)] <athena7> ohhh i see
[12:44:32 EDT(-0400)] <athena7> hm.
[12:44:47 EDT(-0400)] <EricDalquist> when portlet support was added the chan param screens were re-used
[12:44:50 EDT(-0400)] <athena7> so when you configure it in the channel manager it magically gets set as a portlet preference instead?
[12:44:55 EDT(-0400)] <EricDalquist> by pre-pending portlet preferences with PORTLET.
[12:45:04 EDT(-0400)] <EricDalquist> and the code magically makes those into definition prefs
[12:45:05 EDT(-0400)] <EricDalquist> yes
[12:45:14 EDT(-0400)] <athena7> so is there a way to set that in the .channel files?
[12:45:24 EDT(-0400)] <EricDalquist> prefix the names with PORTLET.
[12:45:25 EDT(-0400)] <athena7> can i just prepend those with PORTLET as well?
[12:45:28 EDT(-0400)] <athena7> great! thanks
[12:45:30 EDT(-0400)] <EricDalquist> yup
[12:45:37 EDT(-0400)] <EricDalquist> but you still can't do multi-valued
[12:45:45 EDT(-0400)] <athena7> gotcha
[12:45:47 EDT(-0400)] <EricDalquist> which limits some of the proxy portlet's features
[12:46:11 EDT(-0400)] <athena7> so uportal does suppose multi-valued portlet prefs from the user though
[12:46:16 EDT(-0400)] <athena7> just not from channel manager and such?
[12:46:29 EDT(-0400)] <EricDalquist> yeah
[12:46:40 EDT(-0400)] <EricDalquist> the portlet support in the framework supports multi-valued definition and entity prefs
[12:47:00 EDT(-0400)] <EricDalquist> just the current channel manager code (which also handles importing the .channel files) doesn't
[12:47:41 EDT(-0400)] <athena7> gotcha
[12:48:23 EDT(-0400)] <athena7> i need to make a fix for the ajax channel adding stuff
[12:48:34 EDT(-0400)] <athena7> it looks like it sort of fails awkwardly when the channel fails to be instantiated
[12:49:00 EDT(-0400)] <EricDalquist> my plan around that is to add a <preferences> section to the .channel file format like what portlet.xml supports and to just create a whole new Portlet Prefs screen for channel manager
[12:49:05 EDT(-0400)] <EricDalquist> that supports multi-valued prefs
[12:49:10 EDT(-0400)] <EricDalquist> without length limits
[12:49:37 EDT(-0400)] <athena7> you had a really nice example for the web proxy portlet for google at one point - is that still around somewhere?
[12:49:38 EDT(-0400)] <EricDalquist> brb agin
[12:50:19 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[12:50:23 EDT(-0400)] <athena7> welcome back
[12:53:00 EDT(-0400)] <athena7> so you had a really nice example for the web proxy portlet for google at one point - is that still around somewhere?
[12:53:22 EDT(-0400)] <EricDalquist> I think it is still up on the wiki
[12:54:09 EDT(-0400)] <athena7> hm, maybe it's just not a child page of the portlet project page
[12:55:29 EDT(-0400)] <EricDalquist> just remembered: http://developer.ja-sig.org/projects/webproxyportlet/configuration.html
[12:56:18 EDT(-0400)] <athena7> oh, thanks!
[12:56:49 EDT(-0400)] * jessm (n=Jess@142.150.154.101) has joined ##uportal
[13:16:06 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[13:19:58 EDT(-0400)] * athena7 (n=athena7@c-24-2-189-230.hsd1.mn.comcast.net) has joined ##uportal
[13:34:43 EDT(-0400)] * EiNZTEiN (n=einztein@205.241.143.4) has joined ##uportal
[13:56:00 EDT(-0400)] <athena7> any suggestions on what generally causes "The requested resource (/PlutoInvoker/socialnetportlet) is not available" type errors?
[13:57:39 EDT(-0400)] <EricDalquist> the portlet app either isn't there
[13:57:41 EDT(-0400)] <EricDalquist> or didn't start
[13:57:53 EDT(-0400)] <athena7> ok, that's what i thought
[13:57:54 EDT(-0400)] <athena7> hmm
[13:57:58 EDT(-0400)] <EricDalquist> and there is a bug in up3 that doesn't grab that output correctly
[13:58:07 EDT(-0400)] <EricDalquist> so it shows up in weird places on the portal page
[13:58:07 EDT(-0400)] <athena7> yeah
[13:58:19 EDT(-0400)] <EricDalquist> I filed a jira for that yesterday
[13:58:20 EDT(-0400)] <athena7> strange, because it looks like it's in the right context and the logs show it starting successfully
[13:58:23 EDT(-0400)] <athena7> oh good
[13:58:50 EDT(-0400)] <EricDalquist> does the web.xml for the portlet have the right stuff in it?
[13:58:57 EDT(-0400)] <athena7> looks like it
[14:00:16 EDT(-0400)] <EricDalquist> hrm yeah not sure
[14:00:51 EDT(-0400)] <EricDalquist> somehow Tomcat isn't finding the specified context or portlet name when the portlet container tries to initialize the portlet
[14:00:58 EDT(-0400)] <EricDalquist> double check the portlet publishing config?
[14:01:06 EDT(-0400)] <EricDalquist> if that is wrong it could cause a problem too
[14:02:34 EDT(-0400)] <athena7> hm, good point
[14:03:39 EDT(-0400)] <EricDalquist> another eventualy channel manager improvement is to give users a drop-down of available portlet contexts and portlets
[14:03:45 EDT(-0400)] <athena7> ah!!
[14:03:48 EDT(-0400)] <EricDalquist> since that info is now provided via pluto
[14:03:53 EDT(-0400)] <athena7> typo in the channel definition
[14:03:55 EDT(-0400)] <athena7> thank you
[14:03:57 EDT(-0400)] <EricDalquist> yup
[14:04:00 EDT(-0400)] <athena7> i knew i had to be missing something!
[14:04:02 EDT(-0400)] <athena7> ugh.
[14:04:50 EDT(-0400)] <athena7> yeah, that would be neat
[14:04:58 EDT(-0400)] <athena7> wasn't there something like that in the sandbox up3?
[14:05:05 EDT(-0400)] <EricDalquist> yup
[14:05:16 EDT(-0400)] <athena7> neat (smile)
[14:05:19 EDT(-0400)] <EricDalquist> we had a full (albiet not much better UI) portlet manager
[14:05:33 EDT(-0400)] <EricDalquist> yeah the way portlet app registration is managed with pluto 1.1 is SO much better
[14:05:38 EDT(-0400)] <EricDalquist> no file system scanning
[14:05:50 EDT(-0400)] <EricDalquist> all event based self reg/dereg by the portlet apps
[14:06:03 EDT(-0400)] <EricDalquist> so pluto always knows about all available portlets in the container
[14:13:50 EDT(-0400)] <athena7> (smile)
[14:20:07 EDT(-0400)] <athena7> ugh i hate configuring these facebook api keys
[14:21:14 EDT(-0400)] <holdorph> just think if MS buys facebook, how much *better* it will be
[14:21:34 EDT(-0400)] <athena7> haha
[14:21:46 EDT(-0400)] <athena7> are they looking to acquire facebook?
[14:22:03 EDT(-0400)] <holdorph> yeah, since their failed attempt at yahoo
[14:22:17 EDT(-0400)] <athena7> ah, i'd missed that latest move i guess
[14:22:23 EDT(-0400)] <athena7> i'm really glad they failed at taking over yahoo
[14:22:32 EDT(-0400)] <athena7> yahoo has too much cool stuff that i don't want to see killed
[14:22:36 EDT(-0400)] <holdorph> yup. very happy about that.
[14:22:38 EDT(-0400)] <athena7> it's too bad yahoo's been having so much trouble
[14:22:47 EDT(-0400)] <athena7> they have some really neat stuff
[14:22:56 EDT(-0400)] <athena7> beyond even the other cool companies that they've bought
[14:23:19 EDT(-0400)] <athena7> like all the web APIs and YUI toolkit and such
[14:23:31 EDT(-0400)] <athena7> their feeds mashup thing is pretty cool too
[15:17:45 EDT(-0400)] * jessm (n=Jess@142.150.154.101) has joined ##uportal
[16:15:35 EDT(-0400)] <EricDalquist> <jittery>... too ... much ... caffeine ...</jittery>
[16:15:39 EDT(-0400)] <athena7> uhoh!
[16:15:48 EDT(-0400)] <athena7> i had a bit too much on an empty stomach this morning
[16:15:50 EDT(-0400)] <athena7> feeling icky now
[16:16:11 EDT(-0400)] <EricDalquist> I feel "fine" just can't stop moving (tongue)
[16:16:19 EDT(-0400)] <athena7> lol
[16:16:50 EDT(-0400)] <athena7> having you start typing really fast yet? (smile)
[16:17:03 EDT(-0400)] <athena7> i have days like OMG I CAN FINISH EVERYTHING RIGHT NOW
[16:17:44 EDT(-0400)] <EricDalquist> yup
[16:17:47 EDT(-0400)] <EricDalquist> (smile)
[16:19:13 EDT(-0400)] <athena7> i've been trying to take it easy since the conference
[16:19:20 EDT(-0400)] <athena7> apparently i got a concussion, oops (tongue)
[16:20:07 EDT(-0400)] <EricDalquist> uhoh
[16:20:14 EDT(-0400)] <EricDalquist> oh I just replied to your portlets question
[16:20:25 EDT(-0400)] <EricDalquist> can you get the facebook API to POST back to the action url with the parameter?
[16:20:54 EDT(-0400)] <athena7> i don't think so
[16:21:04 EDT(-0400)] <EricDalquist> (sad)
[16:21:10 EDT(-0400)] <athena7> i can try turning this popup=true flag back on and see what that does
[16:21:19 EDT(-0400)] <athena7> although i don't htink it's relevant
[16:21:19 EDT(-0400)] <EricDalquist> the other option would be to render a URL to a servlet with the portlet
[16:21:23 EDT(-0400)] <athena7> yeah
[16:21:28 EDT(-0400)] <EricDalquist> have the servlet put the key in the session
[16:21:35 EDT(-0400)] <EricDalquist> and then have that redirect to an actionurl
[16:21:44 EDT(-0400)] <EricDalquist> though how the servlet gets an actionurl would be a bit trickey
[16:21:59 EDT(-0400)] <EricDalquist> though if you escape it you could pass it as a parameter to the servlet (tongue)
[16:22:08 EDT(-0400)] <athena7> yeah that's what i was about to suggest (tongue)
[16:22:28 EDT(-0400)] <EricDalquist> this is this contstant issue with portlets of assuming URL formats
[16:22:39 EDT(-0400)] <athena7> it's too bad you can't just tell the facebook portlet to call the parameter something else
[16:22:41 EDT(-0400)] <EricDalquist> in up3 I added a prefix to portlet parameters to make finding them easier
[16:22:44 EDT(-0400)] <EricDalquist> yeah
[16:22:46 EDT(-0400)] <EricDalquist> but even then
[16:22:47 EDT(-0400)] <athena7> well, it's more a problem with the facebook service
[16:22:59 EDT(-0400)] <EricDalquist> if someone changes then portlet parameter prefix they break it again
[16:23:01 EDT(-0400)] <athena7> the portlet does a good job of calling getActionUrl instead of manually constructing it, etc.
[16:23:05 EDT(-0400)] <athena7> right
[16:23:38 EDT(-0400)] <athena7> i think we may just have to go the servlet route
[16:23:56 EDT(-0400)] <EricDalquist> yeah I think that would be the most 'correct'
[16:24:07 EDT(-0400)] <EricDalquist> though it is quite a bit clunkier
[16:25:44 EDT(-0400)] <EricDalquist> wow ... that debian bug is BAD
[16:25:49 EDT(-0400)] <EricDalquist> it had a keyspace of 262148
[16:26:11 EDT(-0400)] <athena7> (sad)
[16:59:06 EDT(-0400)] <athena7> so are we intentionally making issues be resolved with both 3.1 and 3.0.1 fix versions?
[16:59:25 EDT(-0400)] <EricDalquist> yes
[16:59:35 EDT(-0400)] <athena7> ok
[16:59:36 EDT(-0400)] <athena7> can do
[17:42:16 EDT(-0400)] <deuce> eric?
[17:42:20 EDT(-0400)] <EricDalquist> hey deuce
[17:42:52 EDT(-0400)] <deuce> does the tomcat ear deployer support putting libraries into common/endorsed?
[17:43:13 EDT(-0400)] <EricDalquist> it supports a single 'shared' location
[17:43:55 EDT(-0400)] <EricDalquist> and Anthony's commit this morning fixes that support so you can target that shared location anywhere instead of just under the server root
[17:44:26 EDT(-0400)] <EricDalquist> I guess the better way to state it is it can put all the .war files in the EAR in one directory and all of the .jar files in the EAR in another
[17:45:21 EDT(-0400)] <deuce> hmm.. i'm having issues needing xalan in endorsed. just putting it in shared/lib no workie
[17:45:30 EDT(-0400)] <EricDalquist> really?
[17:45:32 EDT(-0400)] <deuce> yes
[17:45:41 EDT(-0400)] <EricDalquist> what version of Java and what code is having issues?
[17:46:10 EDT(-0400)] <deuce> 1.5.0_13
[17:46:13 EDT(-0400)] <deuce> toro portlets
[17:46:30 EDT(-0400)] <EricDalquist> ah, they require a special version of xalan?
[17:46:45 EDT(-0400)] <deuce> yeah they require 2.7.0
[17:47:13 EDT(-0400)] <EricDalquist> strange ... uportal does too and it seems to work just including it in /uPortal/web-inf/lib
[17:47:37 EDT(-0400)] * EricDalquist has never fully understood xalan/xerces/endorsed issues
[17:47:47 EDT(-0400)] <deuce> me either
[17:48:21 EDT(-0400)] <deuce> is the uPortal xalan the combined version or it's missing xsltc?
[17:48:42 EDT(-0400)] <deuce> it's the combined
[17:48:45 EDT(-0400)] <EricDalquist> not sure ... it is whatever is in the public maven repo
[17:49:36 EDT(-0400)] <deuce> hmm.. i'm not sure why I cannot just place it in my webapp's lib .. or even in shared!
[17:50:08 EDT(-0400)] <EricDalquist> yeah, WEB-INF/lib seems like the best place
[17:50:17 EDT(-0400)] <EricDalquist> I've grown quite the aversion to stuff in shared
[17:50:54 EDT(-0400)] <EricDalquist> and since you're on JDK5 you shouldn't need it in endorsed since the JDK dom objects should all be the dom3 versions
[17:51:18 EDT(-0400)] <EricDalquist> I think that was the reason it was needed there in JDK4, the DOM version for the included XML apis was wrong
[17:52:07 EDT(-0400)] <deuce> well our goofy portlets only work when xalan is in endorsed (smile)
[17:52:22 EDT(-0400)] <EricDalquist> lol
[17:52:37 EDT(-0400)] <deuce> i suppose i'll have to extend the ear deployer
[17:52:47 EDT(-0400)] <EricDalquist> do you have a stack trace or something you could share if the jar isn't here?
[17:53:02 EDT(-0400)] <EricDalquist> yeah, that is an option, hopefully it is fairly easy to extend
[17:53:07 EDT(-0400)] <EricDalquist> that was the goal at least
[17:53:11 EDT(-0400)] <deuce> Caused by: java.lang.RuntimeException: Unknown error in XPath.
[17:53:11 EDT(-0400)] <deuce> at org.apache.xalan.transformer.TransformerImpl.run(TransformerImpl.java:3407)
[17:53:11 EDT(-0400)] <deuce> at org.apache.xalan.transformer.TransformerHandlerImpl.endDocument(TransformerHandlerImpl.java:433)
[17:53:11 EDT(-0400)] <deuce> at net.unicon.penelope.store.AbstractEntityStore$ChoiceCollectionSaxParser.endDocument(AbstractEntityStore.java:1781)
[17:53:12 EDT(-0400)] <deuce> at org.jasig.portal.utils.SAX2BufferImpl.outputBuffer(SAX2BufferImpl.java:258)
[17:53:13 EDT(-0400)] <deuce> at org.jasig.portal.utils.SAX2BufferImpl.outputBuffer(SAX2BufferImpl.java:228)
[17:53:15 EDT(-0400)] <deuce> at net.unicon.warlock.fac.xml.XmlWarlockFactory$RenderingEngineImpl.render(XmlWarlockFactory.java:451)
[17:53:18 EDT(-0400)] <deuce> at net.unicon.warlock.portlet.AbstractWarlockPortlet.render(AbstractWarlockPortlet.java:338)
[17:53:20 EDT(-0400)] <deuce> ... 22 more
[17:55:29 EDT(-0400)] <EricDalquist> strange ... not a very helpful error
[17:55:35 EDT(-0400)] <deuce> nope :-/
[17:55:53 EDT(-0400)] <deuce> bb
[18:11:26 EDT(-0400)] <deuce> so any thoughts before i venture into the deployer?
[18:11:47 EDT(-0400)] <EricDalquist> not really other than trying to figure out why exactly you're getting that XPath error
[18:11:55 EDT(-0400)] <EricDalquist> any idea as to what in the XSL is causing it to barf?
[18:12:19 EDT(-0400)] <deuce> it's an identity transform :-/
[18:12:47 EDT(-0400)] <EricDalquist> wow
[18:12:57 EDT(-0400)] <EricDalquist> and that is the root cause in the stack ...
[18:13:04 EDT(-0400)] <deuce> yeah
[18:13:24 EDT(-0400)] <EricDalquist> uhg, yeah other than looking into how the code loads the XML/XSLT APIs it uses
[18:13:40 EDT(-0400)] <EricDalquist> does it code directly against Xalan or use the JDK apis?
[18:14:33 EDT(-0400)] <deuce> i'm trying to remember
[18:14:38 EDT(-0400)] <deuce> lemme check
[18:14:49 EDT(-0400)] <EricDalquist> btw, I'm just randomly guessing here (smile)
[18:15:11 EDT(-0400)] <deuce> einstein randomly guessed from time to time too
[18:16:11 EDT(-0400)] <EricDalquist> lol, I'll take that compliment (wink)
[18:17:29 EDT(-0400)] <deuce> (smile)
[18:17:58 EDT(-0400)] <deuce> it doesn't use xalan directly. it uses the TransformerFactory
[18:18:24 EDT(-0400)] <deuce> similar to how uportal does in XSLT
[18:18:35 EDT(-0400)] <deuce> yeah i saw that..
[18:19:23 EDT(-0400)] <EricDalquist> though the debugger may not show you that ...
[18:19:29 EDT(-0400)] <EricDalquist> verbose classloading would though
[18:19:35 EDT(-0400)] <deuce> indeed
[18:19:47 EDT(-0400)] <EricDalquist> see if some 'bad' version of xalan is sneaking in from somewhere else
[18:20:08 EDT(-0400)] <deuce> xalan has been quite a pain from time to time
[18:20:25 EDT(-0400)] <EricDalquist> which I have seen before ... other libs packaging parts of xalan in themselves
[18:20:31 EDT(-0400)] <EricDalquist> which causes loads of fun like this
[18:20:47 EDT(-0400)] <deuce> thanks for the tip. i will try

  • No labels