uPortal IRC Logs-2010-12-15

[12:04:45 CST(-0600)] <athena> EricDalquist: fixed that fragmentDefinitionDao issue in 3.2.x
[13:16:32 CST(-0600)] <athena> EricDalquist: have you tried fragment administration recently?
[13:16:42 CST(-0600)] <EricDalquist> in which version?
[13:16:57 CST(-0600)] <athena> sorry! trunk
[13:17:02 CST(-0600)] <athena> seems to be throwing errors now
[13:17:05 CST(-0600)] <EricDalquist> nope, hrm
[13:17:16 CST(-0600)] <EricDalquist> Nick did convert that and fragment exit into flows
[13:17:24 CST(-0600)] <EricDalquist> what error are you seeing?
[13:17:24 CST(-0600)] <athena> http://uportal.pastebin.com/Y0rKnb7E
[13:17:34 CST(-0600)] <athena> yeah, that was a while ago though right?
[13:17:41 CST(-0600)] <athena> it seems like this must be pretty recent
[13:17:44 CST(-0600)] <EricDalquist> neat, yeah a few months
[13:17:53 CST(-0600)] <athena> yeah, i think it worked after he made that change
[13:21:14 CST(-0600)] <EricDalquist1> arg … wifi at home appears to be very finicky today
[13:21:31 CST(-0600)] <athena> ick
[13:21:37 CST(-0600)] <athena> my internet went out 2 nights ago
[13:21:39 CST(-0600)] <EricDalquist1> so is there any more to that stack trace?
[13:21:42 CST(-0600)] <athena> luckily right as i was ending work for the day
[13:21:45 CST(-0600)] <athena> no
[13:21:48 CST(-0600)] <athena> not sure what's going on
[13:22:01 CST(-0600)] <athena> hmmmm
[13:22:02 CST(-0600)] <EricDalquist1> well I'm wondering if it is related to the header support nick just added
[13:22:12 CST(-0600)] <EricDalquist1> because that is a NPE in the rendering pipeline code
[13:22:19 CST(-0600)] <athena> before that there is a "session must not be null" error that seems to be coming from the fragment admin portlet
[13:22:32 CST(-0600)] <athena> er, sorry, the exit fragment admin portlet
[13:23:03 CST(-0600)] <EricDalquist1> yeah, I think that is from it starting to render because it is in the structure but the portal doesn't wait for it since it isn't in the theme output
[13:24:07 CST(-0600)] <athena> sorry, it was the exit fragment admin portlet
[13:24:11 CST(-0600)] <athena> that one is output in the theme, right?
[13:24:19 CST(-0600)] <EricDalquist1> yeah
[13:24:33 CST(-0600)] <EricDalquist1> off topic but where is the CMS portlet you wrote in SVN?
[13:25:37 CST(-0600)] <athena> here we go: https://source.jasig.org/portlets/SimpleContentPortlet/trunk
[13:25:40 CST(-0600)] <EricDalquist1> thanks
[13:25:41 CST(-0600)] <athena> was just accepted into incubation
[13:25:46 CST(-0600)] <athena> been thinking about some updates to that
[13:25:51 CST(-0600)] <nickblair> hello
[13:25:53 CST(-0600)] <EricDalquist1> nickblair: this is the stack trace
[13:25:54 CST(-0600)] <EricDalquist1> http://uportal.pastebin.com/Y0rKnb7E
[13:26:05 CST(-0600)] <athena> first of all, we'd talked about letting people add tokens for usernames and such, second, how do we internationalize that
[13:26:05 CST(-0600)] <EricDalquist1> and when are you seeing this athena?
[13:26:06 CST(-0600)] <athena> hey nick
[13:27:22 CST(-0600)] <nickblair> hi - what's up?
[13:27:22 CST(-0600)] <EricDalquist1> I'm going to pull up the ChunkingEventReader code too
[13:27:29 CST(-0600)] <EricDalquist1> see if it is something obvious
[13:27:34 CST(-0600)] <athena> when i attempt to impersonate a fragment owner
[13:27:43 CST(-0600)] <athena> i guess maybe it'd be worth trying the other impersonation protlet
[13:27:53 CST(-0600)] <EricDalquist1> so when do you see this error athena?
[13:27:56 CST(-0600)] <athena> see if it happens there, and if it happens w/ just fragment users or whatever
[13:28:02 CST(-0600)] <EricDalquist1> when you submit the form?
[13:28:06 CST(-0600)] <athena> yeah
[13:28:11 CST(-0600)] <athena> was trying the admin-lo user
[13:28:17 CST(-0600)] <athena> hmm.
[13:28:23 CST(-0600)] <athena> here's a point though, just in case it's relevant
[13:28:28 CST(-0600)] <EricDalquist1> could you try it with live-headers running and see if you bounce of the Login servlet?
[13:28:32 CST(-0600)] <athena> the admin-lo fragment layout seem to have two header elements
[13:28:32 CST(-0600)] <EricDalquist1> before the error
[13:28:50 CST(-0600)] <athena> sorry - you'll have to clue me in to what live-headers is
[13:29:06 CST(-0600)] <EricDalquist1> er live-http-headers or firebug
[13:29:10 CST(-0600)] <nickblair> i have to run to a meeting, I'm sorry; I'll check back in about an hour
[13:29:12 CST(-0600)] <EricDalquist1> something that you can use to watch the requets
[13:29:19 CST(-0600)] <EricDalquist1> np nickblair
[13:29:55 CST(-0600)] <athena> i can probably just set a breakpoint in the login servlet for the moment
[13:30:32 CST(-0600)] <EricDalquist1> sounds good
[13:31:15 CST(-0600)] <athena> yep, going through login
[13:31:24 CST(-0600)] <EricDalquist1> athena
[13:31:26 CST(-0600)] <EricDalquist1> in your log file
[13:31:51 CST(-0600)] <EricDalquist1> do you see a warning that says something like "StartElement foo does not have an ID Attribute. No PortletPlaceholderEvent will be generated.
[13:31:57 CST(-0600)] <EricDalquist1> just before the exception?
[13:32:29 CST(-0600)] <athena> well
[13:32:32 CST(-0600)] <athena> i do see that
[13:32:39 CST(-0600)] <athena> unclear if it's right before the exception
[13:32:55 CST(-0600)] <athena> hang on
[13:33:11 CST(-0600)] <athena> ok, yep, looks related
[13:33:12 CST(-0600)]

[13:33:46 CST(-0600)] <EricDalquist1> ok, great
[13:34:11 CST(-0600)] <athena> that narrow it down?
[13:34:17 CST(-0600)] <EricDalquist1> yeah, I think I know what is happening
[13:34:29 CST(-0600)] <EricDalquist1> should be an easy fix …
[13:34:42 CST(-0600)] <athena> oh yay
[13:34:52 CST(-0600)] <athena> thinking i might try and finish up some of the DLM administration js features
[13:35:12 CST(-0600)] <athena> get them fully factored as fluid components
[13:36:14 CST(-0600)] <athena> is it legal to have two header elements in a layout? i'm sort of confused about that
[13:36:37 CST(-0600)] <EricDalquist1> um, I'm not sure
[13:36:44 CST(-0600)] <EricDalquist1> it probably "works"
[13:36:52 CST(-0600)] <EricDalquist1> but it should get cleaned up
[13:36:55 CST(-0600)] <EricDalquist1> ok just committed the fix
[13:38:21 CST(-0600)] <athena> lol
[13:38:29 CST(-0600)] <athena> works is better than not-works
[13:38:31 CST(-0600)] <athena> i'll try it out
[13:42:36 CST(-0600)] <athena> hurray!
[13:42:37 CST(-0600)] <athena> thanks eric
[13:42:42 CST(-0600)] <EricDalquist1> yup
[13:43:03 CST(-0600)] <EricDalquist1> it was part of a fix we had to make for the header support that apparently wasn't a complete fix
[13:43:07 CST(-0600)] <EricDalquist1> also we have header support now
[13:43:26 CST(-0600)] <EricDalquist1> so if you set your portlet up right you can contribute content at the end of the <head> block now
[13:48:07 CST(-0600)] <EricDalquist1> though sadly spring portlet mvc doesn't have header support as far as nick could find
[14:23:08 CST(-0600)] <athena> yeah i saw that
[14:23:13 CST(-0600)] <athena> that does seem really unfortunate
[14:23:36 CST(-0600)] <athena> i suppose once we figure out how to do that we'll have to make sure it's all working with our javascript strategy
[14:25:07 CST(-0600)] <EricDalquist1> yeah, I think it will work just fine
[14:25:18 CST(-0600)] <EricDalquist1> all the portal JS imports and init happen before any portlet content
[14:26:22 CST(-0600)] <athena> well
[14:26:28 CST(-0600)] <athena> right now i think actually that'll break things
[14:26:54 CST(-0600)] <EricDalquist1> really?
[14:27:05 CST(-0600)] <EricDalquist1> that is essentially the same place in order portlet stuff gets loaded right now
[14:27:13 CST(-0600)] <EricDalquist1> after the head but before the end of the page
[14:27:47 CST(-0600)] <athena> the JS imports each happy inline in the page now though right?
[14:27:58 CST(-0600)] <athena> so if portlet A imports jquery, it immediately uses it and then clears the namespace
[14:28:05 CST(-0600)] <athena> then portlet B does the same w/ another version
[14:28:41 CST(-0600)] <EricDalquist1> right, so if a portlet uses the head support it would just need to move all of its imports and init at once
[14:28:55 CST(-0600)] <EricDalquist1> so then portlet A gets its stuff all loaded, inited and noconflicted in the head
[14:29:13 CST(-0600)] <nickblair> hello - I'm back
[14:29:20 CST(-0600)] <athena> right
[14:29:26 CST(-0600)] <athena> so it can work, but we'll need to make some changes
[14:29:56 CST(-0600)] <athena> each portlet will probably have to create its own dynamic javascript import that ties the jquery variable to the portlet namespace
[14:30:12 CST(-0600)] <athena> hopefully the head resource stuff has access to the portlet namespace
[14:30:14 CST(-0600)] <nickblair> ericD and athena - get the problem with the ChunkingEventReader sorted?
[14:30:15 CST(-0600)] <EricDalquist1> yeah … I guess that is what all the portlets I've done already do
[14:30:17 CST(-0600)] <EricDalquist1> yeah
[14:30:29 CST(-0600)] <EricDalquist1> it was a bug in the bugfix I had sent you
[14:30:36 CST(-0600)] <EricDalquist1> never updated the element name variable
[14:30:46 CST(-0600)] <nickblair> ah ok
[14:32:12 CST(-0600)] <EricDalquist1> athena: it does, you have access to everything you do during a normal portlet render
[14:32:35 CST(-0600)] <athena> excellent
[14:32:37 CST(-0600)] <nickblair> eric's right
[14:32:41 CST(-0600)] <nickblair> the RenderRequest is identical
[14:32:51 CST(-0600)] <nickblair> with the exception of the PortletRequest.RENDER_PART request attribute
[14:32:54 CST(-0600)] <athena> so we'll just need to come up with a strategy for creating a dynamic javascript import file for each project that does what we're doing inline in the page now
[14:32:58 CST(-0600)] <EricDalquist1> http://uportal.pastebin.com/yZqGKeeG
[14:33:00 CST(-0600)] <nickblair> it's like the portlet gets it's render fired 2x
[14:33:04 CST(-0600)] <EricDalquist1> that is what I generally do for portlet JS
[14:33:11 CST(-0600)] <athena> makes sense
[14:33:15 CST(-0600)] <athena> maybe we should move that direction anyway
[14:33:25 CST(-0600)] <EricDalquist1> you don't need to make them js import files
[14:33:32 CST(-0600)] <EricDalquist1> you can write out a <script> tag with content
[14:33:38 CST(-0600)] <EricDalquist1> you can actually write out anything you want
[14:33:43 CST(-0600)] <nickblair> very true
[14:33:45 CST(-0600)] <athena> nice
[14:33:48 CST(-0600)] <EricDalquist1> the portlet spec makes no restrictions on what a portlet writes during header rendering
[14:33:50 CST(-0600)] <athena> i haven't had a chance to play with it yet
[14:33:53 CST(-0600)] <athena> really glad we have support now
[14:34:00 CST(-0600)] <athena> would it be hard to add that to spring mvc?
[14:34:01 CST(-0600)] <nickblair> the FunctionalTestsPortlet trunk prints out some CSS
[14:34:01 CST(-0600)] <EricDalquist1> you simply have access to a Writer
[14:34:08 CST(-0600)] <athena> oh cool
[14:34:10 CST(-0600)] <athena> that makes sense
[14:34:13 CST(-0600)] <EricDalquist1> nickblair: I think you looked at that righ?
[14:34:17 CST(-0600)] <nickblair> i was looking at Spring and thinking about what it would take
[14:34:19 CST(-0600)] <athena> i think i get it now
[14:34:37 CST(-0600)] <athena> interesting of course that the yslow tool still advocates not putting js in the head
[14:34:43 CST(-0600)] <nickblair> I'd like to see renderPart added to the RenderMapping annotation, with a default value of RENDER_MARKUP
[14:34:45 CST(-0600)] <athena> don't know if that's worth considering
[14:34:54 CST(-0600)] <EricDalquist1> yeah
[14:34:55 CST(-0600)] <EricDalquist1> perhaps
[14:35:08 CST(-0600)] <EricDalquist1> not sure if the impact is worth the extra complexity
[14:35:24 CST(-0600)] <nickblair> if your annotated method set renderPart to RENDER_HEADERS, it would fire the Portlet#doHeaders method instead of Portlet#doRender
[14:35:46 CST(-0600)] <nickblair> I plan on looking a little closer at it next week
[14:41:17 CST(-0600)] <rickard> athena: Is this the fix for the issue I talked about yesterday? https://developer.jasig.org/source/changelog/jasigsvn/uPortal/branches/rel-3-2-patches/uportal-impl/src/main/resources/properties/contexts/persistenceContext.xml?cs=22673
[14:41:37 CST(-0600)] <rickard> 3-2-patches not deploying that is.
[14:43:15 CST(-0600)] <athena> rickard: yes!
[14:43:18 CST(-0600)] <athena> is it working for you?
[14:46:44 CST(-0600)] <rickard> Great. Haven't tried it yet though
[14:47:52 CST(-0600)] <athena> let me know if that doesn't get things working
[14:58:52 CST(-0600)] <athena> today i am going to remove ajax-preferences-jquery from the trunk!
[14:58:54 CST(-0600)] <athena> huzzah!
[15:00:32 CST(-0600)] <EricDalquist1> yay!
[18:30:20 CST(-0600)] <ajan> Hi, anyone have an idea of a quick fix for Simple Content Management Portlet to support UTF characters? I already tried using the spring charsetEncodingFilter in front of the portlet and also in front of uportal, but it does not help. I also tried to put request and response.setencoding in the include.jsp of the portlet