uPortal IRC Logs-2011-04-06
[10:56:41 CDT(-0500)] <EricDalquist> so liferay doesn't return the username for getRemoteUser?
[11:01:53 CDT(-0500)] <athena> no, i believe it returns the internal ID
[11:02:31 CDT(-0500)] <athena> which is some long number
[11:02:40 CDT(-0500)] <EricDalquist> nice
[11:02:50 CDT(-0500)] <EricDalquist> yeah I don't think the spec is all that ambiguous there
[11:03:25 CDT(-0500)] <athena> yeah, i agree
[11:03:39 CDT(-0500)] <athena> i do think that there's potential ambiguity in what you actually want in some portlets
[11:03:49 CDT(-0500)] <EricDalquist> right
[11:03:56 CDT(-0500)] <EricDalquist> 286 helps a little there
[11:04:01 CDT(-0500)] <athena> like say for an email portlets, there's a good chance that you want the institutional string id, and a good chance that is what the user logged in using
[11:04:02 CDT(-0500)] <EricDalquist> by also giving you access to a "windowId"
[11:04:23 CDT(-0500)] <athena> but actually maybe your email system requires you to log in with your email address instead of your username
[11:04:30 CDT(-0500)] <EricDalquist> yup
[11:04:36 CDT(-0500)] <athena> most of our portlets get the username for cases like that, rather than actually determining the user
[11:05:00 CDT(-0500)] <EricDalquist> yeah those cases USER_INFO makes more sence
[11:05:10 CDT(-0500)] <EricDalquist> I'd still like to get the features we talked about in the long-dead 3.x code
[11:05:24 CDT(-0500)] <EricDalquist> of having the option to map data from the portlet.xml at publish time
[11:05:47 CDT(-0500)] <EricDalquist> so like if a portlet.xml declares a user attribute "emailLoginName"
[11:05:58 CDT(-0500)] <EricDalquist> you could map that to a user attribute that uPortal knows about when publishing the portlet
[11:06:13 CDT(-0500)] <athena> interesting
[11:06:16 CDT(-0500)] <athena> yeah that'd be nice
[11:06:18 CDT(-0500)] <EricDalquist> do the same thing with roles, etc
[11:06:27 CDT(-0500)] <EricDalquist> that is the intent in the spec
[11:06:34 CDT(-0500)] <athena> oh interesting
[11:06:43 CDT(-0500)] <EricDalquist> that portals provide a way to map between declared role/attribute/etc keys
[11:06:46 CDT(-0500)] <EricDalquist> and what the portal supports
[11:07:00 CDT(-0500)] <athena> yeah that'd be useful
[11:07:28 CDT(-0500)] <EricDalquist> so something to stick on the todo list
[11:08:05 CDT(-0500)] <athena> i bet there's only one item on there
[11:08:07 CDT(-0500)] <athena>
[11:08:38 CDT(-0500)] <EricDalquist> lol
[12:34:30 CDT(-0500)] <athena> ergh, found another JAXB issue with JSTL
[12:34:37 CDT(-0500)] <EricDalquist>
[12:34:42 CDT(-0500)] <EricDalquist> whats that?
[12:34:47 CDT(-0500)] <athena> for xs:boolean types, they produce getters/setters that use Booleans
[12:34:56 CDT(-0500)] <athena> and they name the methods isWhatever/setWhatever
[12:35:04 CDT(-0500)] <athena> which apparently is only legal with the primitive boolean type
[12:35:10 CDT(-0500)] <EricDalquist> oi
[12:36:47 CDT(-0500)] <EricDalquist> http://s-weber.blogspot.com/2011/03/jaxb2-primitive-fixer-plugin.html
[12:41:21 CDT(-0500)] <athena> ah, intereting
[12:41:49 CDT(-0500)] <EricDalquist> its under the Apache 2 license
[12:42:29 CDT(-0500)] <EricDalquist> so if we can always repackage that code a little and publish it under org.jasig
[12:42:59 CDT(-0500)] <EricDalquist> heck the guy may even be happy that someone with central access did that for him
[12:44:27 CDT(-0500)] <athena> sounds good
[12:44:37 CDT(-0500)] <athena> for now i'll probably wrap the objects, since i need a couple convenience methods anyway
[13:49:38 CDT(-0500)] <EricDalquist> running into core XML/XSLT bugs is NOT fun: https://issues.apache.org/jira/browse/XALANJ-2169
[14:00:28 CDT(-0500)] <athena> hey eric i'm a bit confused by this thread: http://www.java.net/node/664374
[14:00:36 CDT(-0500)] <athena> how do i know which version of jaxb we're using?
[14:01:35 CDT(-0500)] <EricDalquist> we're using xjc 2.2.1 ...
[14:01:59 CDT(-0500)] <EricDalquist> not sure if that helps
[14:02:07 CDT(-0500)] <athena> yes, thanks!
[14:02:11 CDT(-0500)] <athena> wanted to make sure
[14:02:17 CDT(-0500)] <EricDalquist> thats declared in the uportal-war pom
[14:02:22 CDT(-0500)] <EricDalquist> in the jaxb2 maven plugin
[14:02:28 CDT(-0500)] <athena> ah
[14:02:40 CDT(-0500)] <athena> not sure i've declared it in my portlet
[14:10:32 CDT(-0500)] <athena> oh interesting
[14:10:43 CDT(-0500)] <athena> so i'd put my classes in a package that was a parent of the target package
[14:10:51 CDT(-0500)] <athena> changing to a totally separate package made it magically work
[14:10:52 CDT(-0500)] <athena> go figure.
[14:11:18 CDT(-0500)] <EricDalquist> weird
[14:13:00 CDT(-0500)] <b-rock> Greetings uPortal devs: we run uportal on an oracle version 10.2.03. we need to go up to 10.2.05 or possible an 11.x version. Can uPortal run on the 11x versions?
[14:13:26 CDT(-0500)] <EricDalquist> we're running uPortal on 11x just fine
[14:13:44 CDT(-0500)] <EricDalquist> Its been a while since I've seen a report of any version of any major DB that uPortal hasn't worked on
[14:13:49 CDT(-0500)] <b-rock> cool. thats what I was looking for THanks EricDalquits!!
[14:15:49 CDT(-0500)] <athena> nope, spoke too soon
[14:15:52 CDT(-0500)] <athena> this crap just doens't work
[14:16:22 CDT(-0500)] <EricDalquist>
[14:31:18 CDT(-0500)] <athena> ok, EricDalquist - i finally tracked this one down
[14:31:23 CDT(-0500)] <EricDalquist> yay!
[14:31:26 CDT(-0500)] <EricDalquist> what is it?
[14:31:37 CDT(-0500)] <athena> to get the bugfix, you have to explicitly include the current jaxb-impl in your runtime classpath
[14:31:43 CDT(-0500)] <EricDalquist> ah
[14:31:49 CDT(-0500)] <athena> the stuff in the plugin configuration is for the plugin
[14:31:57 CDT(-0500)] <EricDalquist> ok
[14:32:00 CDT(-0500)] <EricDalquist> that should be fine
[14:32:11 CDT(-0500)] <athena> but since this is happening when the XML is deserialized, we need the jar so it'll override what's in java 1.6
[14:32:17 CDT(-0500)] <EricDalquist> yup
[14:32:23 CDT(-0500)] <athena> so this only affects my portlet right now, but might want to add it to uportal too
[14:32:35 CDT(-0500)] <athena> or at least keep it in mind if we ever run into this problem
[14:32:40 CDT(-0500)] <EricDalquist> yeah
[14:32:53 CDT(-0500)] <athena> that was painful
[14:33:04 CDT(-0500)] <EricDalquist> must be the day of painful XML stuff
[14:33:39 CDT(-0500)] <EricDalquist> I'm having "fun" updating the urlTemplates.xsl stuff to deal with the changes to the url builder model I'm finishing up
[14:35:11 CDT(-0500)] <athena> hm i bet.
[14:36:03 CDT(-0500)] <EricDalquist> I was trying to be all cute
[14:36:13 CDT(-0500)] <EricDalquist> and pass in a full xml structure that represented the url data
[14:36:30 CDT(-0500)] <EricDalquist> and use jaxb to convert it and then map it onto the actual API
[14:36:34 CDT(-0500)] <EricDalquist> but there is a bug in the JVM
[14:36:42 CDT(-0500)] <EricDalquist> around doing that
[14:36:50 CDT(-0500)] <EricDalquist> and I'm guessing very few people ever actually do this
[14:37:06 CDT(-0500)] <EricDalquist> but at least we got a url XSD out of it
[14:37:23 CDT(-0500)] <EricDalquist> so in the theme XSL you can use the auto-complete features of the xsd to build urls
[14:40:32 CDT(-0500)] <athena> oh, sad
[14:40:36 CDT(-0500)] <athena> but that's good at least
[14:40:41 CDT(-0500)] <athena>
[14:40:55 CDT(-0500)] <EricDalquist> yeah
[14:41:08 CDT(-0500)] <EricDalquist> and I was all excited last night when I thought of doing it that way too
[14:41:44 CDT(-0500)] <athena> i bet!!
[14:43:12 CDT(-0500)] <EricDalquist> I could make it work but I'd have to use com.sun. classes
[14:43:15 CDT(-0500)] <EricDalquist> which would be bad
[14:44:31 CDT(-0500)] <athena> hm, yeah
[14:44:52 CDT(-0500)] <EricDalquist> I have another solution
[14:45:03 CDT(-0500)] <EricDalquist> its just a little more work to get implemented
[14:46:33 CDT(-0500)] <EricDalquist> do you know how to verify with easymock that a particular API has been called?
[14:49:22 CDT(-0500)] <EricDalquist> sory
[14:49:27 CDT(-0500)] <EricDalquist> not easymock
[14:49:29 CDT(-0500)] <EricDalquist> mockito
[14:49:41 CDT(-0500)] <EricDalquist> trying to verify that some setters are called on a mock object
[14:52:13 CDT(-0500)] <athena> yes, let me find it
[14:52:16 CDT(-0500)] <athena> i really like mockito
[14:52:19 CDT(-0500)] <athena> much easier than easymock
[14:52:20 CDT(-0500)] <EricDalquist> found it
[14:52:30 CDT(-0500)] <EricDalquist> verify(portalUrlBuilder).addParameter("remove_target", "foo");
[14:52:33 CDT(-0500)] <EricDalquist> after the test is complete
[14:52:45 CDT(-0500)] <athena> yep!
[14:52:46 CDT(-0500)] <athena> http://mockito.googlecode.com/svn/branches/1.8.4/javadoc/org/mockito/Mockito.html
[14:52:57 CDT(-0500)] <athena> also possible to do partial mocks, which comes in useful sometimes
[16:19:37 CDT(-0500)] <athena> EricDalquist: is maven 2.2.1 still OK for uPortal trunk?
[16:52:42 CDT(-0500)] <DrewWills1> EricDalquist the sonatype access isn't working for me... the ticket was closed and marked "Fixed" but I still get 401
[19:54:29 CDT(-0500)] <EricDalquist> DrewWills1: I saw your post on the jira issue, asking for help there is all we can do for now.