Page Comparison

[08:32:23 EDT(-0400)] * athena7 (n=athena7@adsl-99-146-184-32.dsl.wlfrct.sbcglobal.net) has joined ##uportal
[09:17:54 EDT(-0400)] <higpetter> yawn....
[09:20:26 EDT(-0400)] * athena7 seconds that
[09:23:35 EDT(-0400)] <higpetter> although it's like 3pm here...
[09:23:57 EDT(-0400)] <athena7>
[09:26:46 EDT(-0400)] <higpetter> mission of today: learn how to use AbstractWizardFormController
[09:26:54 EDT(-0400)] <higpetter> or do you have any better suggestion?
[09:27:35 EDT(-0400)] <athena7> i haven't used that one much
[09:29:41 EDT(-0400)] <higpetter> gee.. I don't even know how to properly set up a portlet with springframework...
[09:33:01 EDT(-0400)] <higpetter> like I have a non working portlet I'm supposed to fix... and the question is: should I try to fix it right off, or should I learn how to use that wizard somewhere else first?
[09:39:45 EDT(-0400)] <apetro-_> higpetter it depends on how urgent the project is, doesn't it?
[09:40:15 EDT(-0400)] <apetro-_> there's a whole lot to be said for finding some good examples to work through first.
[09:40:25 EDT(-0400)] <apetro-_> Spring in Action? Example apps shipping with Spring?
[09:40:26 EDT(-0400)] <higpetter> may I say very urgent? supposed to deployed next friday...
[09:40:42 EDT(-0400)] <apetro-> _next Friday? That's not urgent.
[09:41:06 EDT(-0400)] <higpetter> I don't even know if the real problem is the wizard or the crazy student data backend
[09:41:38 EDT(-0400)] <higpetter> but I really suspect the latter..
[09:41:40 EDT(-0400)] <apetro-_> unit tests? Stub backend impl you can plug in?
[09:42:37 EDT(-0400)] <apetro-_> and, since I'm drawing a blank, whois higpetter, if you please?
[09:43:09 EDT(-0400)] <higpetter> apetro-_: I think you met me in denver a year ago. One of the swedish guys...
[09:44:55 EDT(-0400)] <apetro-_> ah, very good.
[09:45:36 EDT(-0400)] <higpetter> hmm.. have we come to the point in the agenda where I should say "oups, what did I do to make you remember me?"
[09:47:48 EDT(-0400)] <higpetter> probably that .se thing would have been enough....
[09:48:22 EDT(-0400)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[09:49:58 EDT(-0400)] <athena7> i've found it convenient that the portlet and servlet controllers work fairly similarly
[09:57:36 EDT(-0400)] * bulloche (n=bulloche@134.250.4.77) has joined ##uportal
[10:00:33 EDT(-0400)] * grimesp (n=grimesp@134.250.4.177) has joined ##uportal
[10:16:31 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:21:11 EDT(-0400)] <higpetter> oh, what's up with ja-sig.org now?
[10:21:22 EDT(-0400)] <EricDalquist> ?
[10:21:27 EDT(-0400)] <higpetter> just very tired, ok
[10:21:43 EDT(-0400)] <EricDalquist> is something not working for you?
[10:22:51 EDT(-0400)] <higpetter> is anything working for me?
[10:24:24 EDT(-0400)] <higpetter> log4j is wonderful... thousands of lines telling me nothing, but nowhere I can find the line that tells me WHY it failed to load what...
[10:24:32 EDT(-0400)] <higpetter> Caused by: javax.portlet.PortletException: Unable to find dispatcher for context: null
[10:29:02 EDT(-0400)] <athena7> is this in the portal's logs?
[10:29:28 EDT(-0400)] <higpetter> yep
[10:29:46 EDT(-0400)] <higpetter> I guess log4j is missing a jar...
[10:29:53 EDT(-0400)] <athena7> how do you have the portlet configured? and what version are you using?
[10:31:10 EDT(-0400)] <higpetter> as basic as possible
[10:31:13 EDT(-0400)] <athena7> there should be some channel parameters that specify the path to the portlet
[10:31:15 EDT(-0400)] <higpetter> uPortal 2.6 I think
[10:31:28 EDT(-0400)] <athena7> like PortletName.portletName
[10:31:34 EDT(-0400)] <higpetter> the portlet starts after a re-deployment
[10:31:44 EDT(-0400)] <higpetter> but not upon tomcat start
[10:32:09 EDT(-0400)] <higpetter> sometimes a reload/restart of the portlet works
[10:32:12 EDT(-0400)] <higpetter> some times not
[10:32:24 EDT(-0400)] <athena7> doesn't sound fun
[10:32:55 EDT(-0400)] <higpetter> you don't think? Oh, I love those predictable errors
[10:32:59 EDT(-0400)] <athena7> lol
[10:33:06 EDT(-0400)] <athena7> yeah, i know
[10:35:19 EDT(-0400)] <higpetter> can anyone tell me exactly what jars I need for log4j?
[10:35:37 EDT(-0400)] <athena7> should be the log4j jar
[10:35:45 EDT(-0400)] <athena7> and depending on what you're doing, commons logging
[10:35:46 EDT(-0400)] <athena7> i think?
[10:36:22 EDT(-0400)] <EricDalquist> yeah that's it
[10:36:29 EDT(-0400)] <EricDalquist> log4j doesn't have any dependencies
[10:36:43 EDT(-0400)] <EricDalquist> another option is to turn on verbose classloading logging
[10:36:55 EDT(-0400)] <EricDalquist> the JVM will write to stdout every class file it loads and where it loads it from
[10:37:05 EDT(-0400)] <EricDalquist> which may help you figure out why you're getting these issues
[10:37:12 EDT(-0400)] <EricDalquist> but it is a lot of data to go through
[10:37:28 EDT(-0400)] <EricDalquist> honestly I've never seen an issue with portlet loading where it works sometimes
[10:37:29 EDT(-0400)] <higpetter> EricDalquist: in what log4j.xml/properties should I enable that? tomcat or uportal or that damned portlet?
[10:38:13 EDT(-0400)] <higpetter> oh, why do I always get those errors noone else get? like "google didn't find any answers for xxx"
[10:38:41 EDT(-0400)] <EricDalquist> verbose classloading in a java option
[10:38:57 EDT(-0400)] <EricDalquist> so you'd add -verbose:class to your tomcat options
[10:38:58 EDT(-0400)] <higpetter> gee... that sounds fun
[10:39:52 EDT(-0400)] <EricDalquist> you could always also email the user list and post a link to the portlet you're having problems with to see if it works for anyone else
[10:44:40 EDT(-0400)] <higpetter> thanks for the classloader thing.. I'll be back in a moment
[10:44:46 EDT(-0400)] * higpetter is reading logs...
[10:53:16 EDT(-0400)] <higpetter> NOTHING failed
[10:53:23 EDT(-0400)] <higpetter> ...in the classloading
[10:59:13 EDT(-0400)] <higpetter> argh! now my mouse died
[10:59:53 EDT(-0400)] * higpetter successfully stole mouse from laptop
[11:00:44 EDT(-0400)] <athena7> day.restart()?
[11:01:01 EDT(-0400)] <higpetter> I'd prefer year.restart()...
[11:01:12 EDT(-0400)] <athena7> aw
[11:01:33 EDT(-0400)] <higpetter> or, make it a couple of years...
[11:44:47 EDT(-0400)] * holdorph (n=holdorph@wsip-98-174-242-39.ph.ph.cox.net) has joined ##uportal
[12:10:17 EDT(-0400)] * jessm (n=Jess@rrcs-70-63-141-143.midsouth.biz.rr.com) has joined ##uportal
[12:17:19 EDT(-0400)] * awills (n=awills@wsip-98-174-242-39.ph.ph.cox.net) has joined ##uportal
[12:54:27 EDT(-0400)] <higpetter> FSCK FSCK FSCK
[12:57:58 EDT(-0400)] <holdorph> ok
[12:59:33 EDT(-0400)] <higpetter> sure... I commented out the configloaderlistener, then the shitty portlet starts
[13:03:28 EDT(-0400)] <awills> anyone here know what might cause uP3 to use the portlet <title> (from portlet.xml) as the channel title in the chrome window occasionally?
[13:03:47 EDT(-0400)] <EricDalquist> awills: yes, there is a know bug
[13:03:51 EDT(-0400)] <holdorph> that was talked about on the uportal list, pretty extensively
[13:04:02 EDT(-0400)] <EricDalquist> haven't gotten to looking into it yet
[13:04:16 EDT(-0400)] <EricDalquist> http://www.ja-sig.org/issues/browse/UP-2099
[13:05:34 EDT(-0400)] <holdorph> discussion was with the subject "Channel title used within uPortal 3 theme" last post was on June 22, 2008.
[13:08:44 EDT(-0400)] <awills> this JIRA is a bit confusing... "On first render [...] the titles are correctly set. However, if the portlet is rendered from cache [...] the title ends up being whatever was set [...] in Channel Manager."
[13:08:55 EDT(-0400)] <awills> isn't the channel manager title the one we want?
[13:09:07 EDT(-0400)] <EricDalquist> depends
[13:09:19 EDT(-0400)] <EricDalquist> according to the spec portlets can set their own titles during rendering
[13:09:32 EDT(-0400)] <awills> sure, and that's fine
[13:09:40 EDT(-0400)] <EricDalquist> so what we want from a JSR-168 view is to honor that on each render
[13:09:47 EDT(-0400)] <EricDalquist> problem is right now it is only honored on the first render
[13:09:49 EDT(-0400)] <awills> i believe what I'm seeing is the title from portlet.xml
[13:09:59 EDT(-0400)] <EricDalquist> yup
[13:10:17 EDT(-0400)] <EricDalquist> and if your portlet extends GenericPortlet it will set the title from portlet.xml on render
[13:10:26 EDT(-0400)] <awills> oh will it
[13:10:32 EDT(-0400)] <EricDalquist> all described in that email thread
[13:10:43 EDT(-0400)] <holdorph> yup
[13:11:13 EDT(-0400)] <awills> is <title> optional? what happens when it's omitted?
[13:11:28 EDT(-0400)] <holdorph> it's required.
[13:11:31 EDT(-0400)] <EricDalquist> it is required
[13:11:37 EDT(-0400)] <EricDalquist> setting of the title is not
[13:11:44 EDT(-0400)] <EricDalquist> but you either need to implement Portlet directly
[13:11:53 EDT(-0400)] <EricDalquist> or override the getTitle method in GenericPortlet
[13:12:23 EDT(-0400)] <holdorph> or just set the value in portlet.xml to the same value in channel manager, and never call getTitle
[13:12:27 EDT(-0400)] <awills> the portlet in question is one you're familiar with... UW WebProxy
[13:12:49 EDT(-0400)] <awills> we want tto publish it several times in portlet manager, each time with a different title
[13:13:27 EDT(-0400)] <EricDalquist> probably needs to be filled as a bug in webproxy
[13:13:32 EDT(-0400)] <EricDalquist> and modify the code to not set the title
[13:14:36 EDT(-0400)] <awills> which, if i understand this item, would mean implementing Portlet and not extending GenericPortlet
[13:15:29 EDT(-0400)] <EricDalquist> or just overriding the http://www.jdocs.com/portlet/1.0/javax/portlet/GenericPortlet.html#M-getTitle(RenderRequest) method of GenericPortlet
[13:17:41 EDT(-0400)] <awills> how would that help? (I'm reading, haven't got there yet) Does GP not set the title in render in getTitle() returns null?
[13:18:15 EDT(-0400)] <higpetter> webAppRootKey was the shit!
[13:18:58 EDT(-0400)] <awills> doesn't look that way, from the javadoc
[13:19:22 EDT(-0400)] <EricDalquist> but uPortal would ignore a null title
[13:19:29 EDT(-0400)] <EricDalquist> and just use the configured title
[13:19:50 EDT(-0400)] <holdorph> just override render
[13:19:55 EDT(-0400)] <holdorph> http://www.koders.com/java/fid1B20FABF16720C53294CEA21F69B0A346634FD98.aspx
[13:20:07 EDT(-0400)] <holdorph> call doDispatch() and don't do the setTitle part
[13:20:34 EDT(-0400)] <EricDalquist> that would work even better
[13:20:49 EDT(-0400)] <awills> yes that looks better
[13:21:20 EDT(-0400)] <awills> doesn't rely on portal-specific cooperation
[13:21:35 EDT(-0400)] <EricDalquist> I can double check but I think the handling of a null title is in the spec
[13:21:37 EDT(-0400)] <holdorph> but it's a portal specific bug
[13:21:44 EDT(-0400)] <holdorph> so not sure that point matters much
[13:22:04 EDT(-0400)] <holdorph> the portal specific cooperation point, that is.
[13:25:38 EDT(-0400)] <awills> there's a portal-specific bug in the total picture here... to wit, that uP exhibits the correct behavior only some of the time...
[13:25:54 EDT(-0400)] <EricDalquist> yup
[13:26:04 EDT(-0400)] <EricDalquist> which will hopefully be fixed for 3.0.2
[13:26:06 EDT(-0400)] <awills> but we want the "correct" behavior exactly none of the time
[13:26:11 EDT(-0400)] <EricDalquist>
[13:26:28 EDT(-0400)] <EricDalquist> perhaps while fixing the bug we should add feature to be able to disable dynamic titles across the board
[13:26:59 EDT(-0400)] <awills> as apublish-time parameter? perhaps