uPortal IRC Logs-2007-10-16

Owned by Apereo Infrastructure
Last updated: Oct 16, 2007Version comment

[09:23:53 EDT(-0400)] * agherna (n=argherna@cites-agherna01.ci.uiuc.edu) has joined ##uportal
[09:33:43 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[09:37:16 EDT(-0400)] * esm (n=esm@asdf.dkc.jhu.edu) has joined ##uportal
[10:13:12 EDT(-0400)] * ac_chan_ (n=alex@tempoutsidepix.pratt.edu) has joined ##uportal
[10:30:51 EDT(-0400)] <EricDalquist> hey esm ... any idea if they made PortletMode & WindowState serializable in JSR-286?
[10:33:07 EDT(-0400)] * EiNZTEiN (n=einztein@check.your.ignorelist.com) has joined ##uportal
[10:33:26 EDT(-0400)] <EiNZTEiN> hi guys
[10:33:38 EDT(-0400)] <EricDalquist> hello
[10:34:14 EDT(-0400)] <agherna> good morning
[10:35:24 EDT(-0400)] <EiNZTEiN> (smile)
[10:36:14 EDT(-0400)] <EiNZTEiN> I'm going to get involved in developing a customized look and feel for a uPortal application, the same one on ISU (Iowa State Univ)
[10:36:57 EDT(-0400)] <EiNZTEiN> I've never worked/heard of uPortal up until now, and I'm guessing that's because of your orientation towards "higher education institutions" (smile)
[10:37:17 EDT(-0400)] <ac_chan_> hello
[10:37:32 EDT(-0400)] <EiNZTEiN> I'm trying to get a "feel" of how it's built, especially on the frontend side
[10:37:32 EDT(-0400)] <EiNZTEiN> (smile)
[10:38:23 EDT(-0400)] <EiNZTEiN> and, don't take me wrong, but doesn't it "feel... old"?? (tongue)
[10:38:39 EDT(-0400)] <EiNZTEiN> (the frontend)
[10:38:49 EDT(-0400)] <EricDalquist> which version have you looked at most recently?
[10:39:19 EDT(-0400)] <esm> EricDalquist:
[10:39:21 EDT(-0400)] <esm> er
[10:39:21 EDT(-0400)] <EricDalquist> also for a more well rounded response I would recommend jumping on the uportal-user email list http://www.ja-sig.org/wiki/display/JSG/uportal-user and asking some questions there as well, this IRC channel is a very small sampling of uPortal users and developers
[10:39:42 EDT(-0400)] <esm> not that I know of: the public review doesn't have them implementing Serializable.
[10:39:42 EDT(-0400)] <EiNZTEiN> EricDalquist: I'm an IRC guy, thus came here first (tongue)
[10:40:07 EDT(-0400)] <EiNZTEiN> EricDalquist: I'll subscribe the mailing-list anyway (wink)
[10:40:10 EDT(-0400)] <EricDalquist> cool
[10:40:15 EDT(-0400)] <EricDalquist> like I said, we'll do our best here
[10:40:26 EDT(-0400)] <EiNZTEiN> (wink)
[10:40:28 EDT(-0400)] <EricDalquist> but there are 800+ people on the uportal-user list
[10:40:35 EDT(-0400)] <EricDalquist> versus 6 here (smile)
[10:40:36 EDT(-0400)] <EiNZTEiN> I love the IRC, that's why (wink)
[10:40:37 EDT(-0400)] <EiNZTEiN> hehe
[10:40:43 EDT(-0400)] <agherna> the uportal-users list is a great source. you can get lots of good feedback there
[10:40:46 EDT(-0400)] <EiNZTEiN> EricDalquist: I'll work with both (wink)
[10:41:06 EDT(-0400)] <EricDalquist> esm: grr ... I have to have custom methods for serialization for my PortletWindow because of those two objects
[10:41:07 EDT(-0400)] <agherna> this channel usu. has about 10 people on it at the most (and I like to lurk) (smile)
[10:41:08 EDT(-0400)] <EricDalquist> (tongue)
[10:41:17 EDT(-0400)] <EiNZTEiN> EricDalquist: with "old" I meant it's still implementing tables in the design
[10:41:23 EDT(-0400)] <EricDalquist> ah
[10:41:23 EDT(-0400)] <EricDalquist> well
[10:41:40 EDT(-0400)] <EricDalquist> you can change that without too much effort if you're comfortable with XSL
[10:42:34 EDT(-0400)] <EricDalquist> I'm not a web UI person so I can't really give much insight there other than in my experience tables work very well for data organized in columns so that is what is being used
[10:42:41 EDT(-0400)] <EricDalquist> uPortal's rendering is XSLT based
[10:42:56 EDT(-0400)] <EiNZTEiN> (smile)
[10:42:56 EDT(-0400)] <esm> EricDalquist: well you'll need it anyway for 1.0.. they are simple objects though (smile)
[10:43:01 EDT(-0400)] <EricDalquist> yeah
[10:43:04 EDT(-0400)] <esm> why do you need to serialize them
[10:43:22 EDT(-0400)] <EricDalquist> well a while back I did some work for a school to make uPortal play nice with session replication
[10:43:38 EDT(-0400)] <EricDalquist> my plans for the pluto 1.1 work involves tracking PortletWindow objects in the user's session
[10:43:59 EDT(-0400)] <EricDalquist> so the PortletWindow must be serializable to play nice with replication
[10:44:21 EDT(-0400)] <esm> gotcha
[10:44:46 EDT(-0400)] <EricDalquist> EiNZTEiN: I can't think of the url right now but I believe there is some documentation on the wiki related to the rendering pipeline that uportal uses
[10:44:56 EDT(-0400)] <EricDalquist> the basic idea is two XSLT transformations are done
[10:45:45 EDT(-0400)] <EiNZTEiN> okay
[10:45:48 EDT(-0400)] <EricDalquist> the first transforms the users layout DOM, which contains all layout info for all tabs in a generic tree structure, into a structure DOM which is tab/column specific
[10:45:55 EDT(-0400)] <EiNZTEiN> I'll be reading/experimenting a lot
[10:46:09 EDT(-0400)] <EricDalquist> very few people ever make changes with this first XSL (the structure transform)
[10:46:51 EDT(-0400)] <EricDalquist> the second transform takes the tab/column structure DOM and generates the markup the user sees
[10:47:03 EDT(-0400)] <EricDalquist> most people customize this second XSL (the theme transform)
[10:47:26 EDT(-0400)] <EricDalquist> folks on the list will be able to point you to the specific XSL documents that are responsible for this work
[10:47:34 EDT(-0400)] <ac_chan_> is there a hsqldb channel on freenode?
[10:47:39 EDT(-0400)] <esm> You'll need serialization methods anyway for PortletWindow and portletwindowId
[10:47:55 EDT(-0400)] <EricDalquist> they are both interfaces that I am responsible for implementing
[10:48:07 EDT(-0400)] <esm> yes
[10:48:10 EDT(-0400)] <EricDalquist> so I wouldn't if I'm smart on how I impl them
[10:48:28 EDT(-0400)] <EricDalquist> oh and just so you know the framework is intelligent about caching those XSL results and only re-running the XSLT when it is needed
[10:48:55 EDT(-0400)] * ac_chan_ (n=alex@tempoutsidepix.pratt.edu) has left ##uportal
[10:49:06 EDT(-0400)] <EricDalquist> esm: there is still a fair bit of design work ... like I'm still not sure how I'm going to deal with generating portlet window IDs
[10:49:08 EDT(-0400)] <esm> er... if your storing portletwindow's in the session for serialization purposes, doesn't the entire object graph need to be serializable
[10:49:15 EDT(-0400)] <EricDalquist> yeah
[10:49:55 EDT(-0400)] <esm> you at least have to implment Serializable on your portletwindow impl
[10:49:58 EDT(-0400)] <EricDalquist> but I'm responsible for the impl of the PortletWindow ... currently the only member variables that aren't serializable are the state & mode
[10:50:00 EDT(-0400)] <EricDalquist> yes
[10:50:12 EDT(-0400)] <esm> (which is a marker interface ... ) but you would need the read/write object methods iirc
[10:50:17 EDT(-0400)] <EricDalquist> nope
[10:50:36 EDT(-0400)] <EricDalquist> you only need the read/write methods if your object requires things more complicated than just writing out its member variables
[10:50:44 EDT(-0400)] <EricDalquist> and if all of those are marked Serializable ...
[10:53:22 EDT(-0400)] <EricDalquist> so yeah ... it isn't a big issue to implement the read/write methods ... just more to maintain
[10:53:37 EDT(-0400)] <EricDalquist> I still have to figure out stuff like how I'm going to generate the PortletWindowIDs
[10:53:51 EDT(-0400)] <EricDalquist> since if I want session replication to work they need to be unique cross-server
[10:57:55 EDT(-0400)] <esm> maybe a hash of the server's ip or mac addy?
[10:58:04 EDT(-0400)] <EricDalquist> yeah
[10:58:13 EDT(-0400)] <EricDalquist> I was thinking something like that
[10:58:24 EDT(-0400)] <EricDalquist> serverhash-incrementednumber
[10:59:43 EDT(-0400)] <esm> oh
[10:59:43 EDT(-0400)] <esm> well
[11:00:05 EDT(-0400)] <esm> somthing else to consider - they have to be unique per jvm... so i guess that is where the incrementing counter may come from
[11:00:10 EDT(-0400)] <esm> er may be used
[11:00:18 EDT(-0400)] <EricDalquist> ah true
[11:00:27 EDT(-0400)] <EricDalquist> so what if there are 2+ jvms on one server
[11:00:30 EDT(-0400)] <EricDalquist> oh joy
[11:00:32 EDT(-0400)] <esm> exactly
[11:00:41 EDT(-0400)] <EricDalquist> sounds like I need to hide this logic behind an interface
[11:02:41 EDT(-0400)] <esm> can you just use a uuid library and hash the uuid everytime the jvm starts
[11:03:01 EDT(-0400)] <EricDalquist> yeah ...
[11:03:05 EDT(-0400)] <EricDalquist> annoying
[11:03:17 EDT(-0400)] <EricDalquist> our portlet window ids are going to be huge (tongue)
[11:05:00 EDT(-0400)] <esm> just get the hashcode of the uuid objecrt
[11:05:03 EDT(-0400)] * ac_chan_ (n=alex@tempoutsidepix.pratt.edu) has joined ##uportal
[11:05:23 EDT(-0400)] <esm> look at http://jug.safehaus.org/curr/javadoc/index.html
[11:05:50 EDT(-0400)] <EricDalquist> nice
[11:05:54 EDT(-0400)] <EricDalquist> thanks
[11:06:10 EDT(-0400)] <esm> np - I've used http://jug.safehaus.org/ that librar in the past
[11:06:52 EDT(-0400)] <esm> of course mac addy stuff requires native libs
[11:06:54 EDT(-0400)] <esm> so...
[11:07:16 EDT(-0400)] <EricDalquist> eh ... I'll stick it behind an interface and figure that out on its own eventually
[11:07:24 EDT(-0400)] <esm> but there are apparently two additional methods (besides mac addies) to generate uuids
[11:07:25 EDT(-0400)] <esm> yeah
[11:07:41 EDT(-0400)] <EricDalquist> I'm getting pretty excited about pluto 1.1 ... I think I have a pretty good integration plan figured out
[11:07:47 EDT(-0400)] <esm> cool
[11:14:44 EDT(-0400)] <ac_chan> hi
[11:16:10 EDT(-0400)] <ac_chan> for every ldap user ... is there a entry for it in the database table up_user ?
[11:17:29 EDT(-0400)] <EricDalquist> for every user that has logged into uPortal yes
[11:17:52 EDT(-0400)] <ac_chan> hrm.. interesting.. i am looking at the database i don't see entry for myself
[11:20:14 EDT(-0400)] * colinclark (n=atrcwrk2@142.150.154.101) has joined ##uportal
[11:20:26 EDT(-0400)] <ac_chan> maybe i am pointing to the wrong urlid with the hsqldb client?
[11:21:03 EDT(-0400)] * michelled (n=michelle@142.150.154.114) has joined ##uportal
[12:49:45 EDT(-0400)] * colinclar1 (n=atrcwrk2@142.150.154.101) has joined ##uportal
[13:47:08 EDT(-0400)] * colinclark (n=atrcwrk2@142.150.154.101) has joined ##uportal
[13:52:43 EDT(-0400)] <ac_chan> hi
[13:55:25 EDT(-0400)] <ac_chan> we just added PAGS configuration to our development uportal setup..
[13:55:35 EDT(-0400)] <ac_chan> how do we know that it worked?
[13:56:24 EDT(-0400)] <EricDalquist> well
[13:56:37 EDT(-0400)] <EricDalquist> you should be able to look up a user in groups manager that should be in a PAGs defined group
[13:56:46 EDT(-0400)] <EricDalquist> and see them in that group in their info
[14:05:03 EDT(-0400)] <EricDalquist> hey esm do you have much luck with the m2eclipse plugin and pluto?
[14:09:02 EDT(-0400)] <ac_chan> thanks.. i am going to change names of the groups to be a little less generic to see if everything is ok
[14:09:15 EDT(-0400)] <EricDalquist> good plan
[14:09:28 EDT(-0400)] <EricDalquist> you should be able to search for the PAGs groups by name as well
[14:09:46 EDT(-0400)] <EricDalquist> what I usually recommend is to add the root PAGs group to the Everyone group in groups manager
[14:10:00 EDT(-0400)] <EricDalquist> otherwise people that only exist in a PAGs group won't be considered part of Everyone
[14:15:32 EDT(-0400)] <esm> EricDalquist: heh
[14:15:40 EDT(-0400)] <esm> you have to comment out something in the container pom... let me find it
[14:15:43 EDT(-0400)] <EricDalquist> yeah
[14:15:47 EDT(-0400)] <EricDalquist> I just figured it out
[14:15:51 EDT(-0400)] <EricDalquist> well I had two issues
[14:15:56 EDT(-0400)] <esm> ok cool yeah it is a pita
[14:15:57 EDT(-0400)] <EricDalquist> first I had to comment out the assembler
[14:16:00 EDT(-0400)] <EricDalquist> in the testsuit
[14:16:01 EDT(-0400)] <esm> hmm
[14:16:10 EDT(-0400)] <EricDalquist> then the META-INF resource directory in the container
[14:16:33 EDT(-0400)] <EricDalquist> the assembler code was failing trying to look up its logger
[14:16:40 EDT(-0400)] <esm> i've always had to do the latter, but never the former...
[14:16:41 EDT(-0400)] <esm> hmm
[14:16:51 EDT(-0400)] <EricDalquist> could just be something in my eclipse env
[14:17:00 EDT(-0400)] <esm> yeah maybe
[14:17:12 EDT(-0400)] <esm> the assembler needs to get stripped out into its own sub-project
[14:17:49 EDT(-0400)] <esm> would make depending on the assembler code much more feasible
[14:18:14 EDT(-0400)] <esm> (esp if it is a shared dep)
[14:20:51 EDT(-0400)] <EricDalquist> yeah
[14:43:51 EDT(-0400)] <ac_chan> the pags group does not show up on the left hand side..
[14:44:14 EDT(-0400)] <ac_chan> of the groups manager
[14:44:37 EDT(-0400)] <EricDalquist> what happens if you search for one of the groups named in the PAGS config?
[14:45:03 EDT(-0400)] <EricDalquist> unless you explicitly added the root PAGs group as a child of Everyone it won't show up there
[14:45:23 EDT(-0400)] <ac_chan> oh..
[14:45:24 EDT(-0400)] <EricDalquist> but searching for groups by name should find you configured pags groups
[14:45:33 EDT(-0400)] <ac_chan> it did ok
[14:45:41 EDT(-0400)] <EricDalquist> does your PAGs xml still have a PAGs root?
[14:46:06 EDT(-0400)] <ac_chan> how would i make a Pags group child of everyone?
[14:46:29 EDT(-0400)] <EricDalquist> edit the everyone group (click the lock icon to make the group editable, just remember to re-lock it)
[14:49:49 EDT(-0400)] <ac_chan> yeah it does
[14:50:02 EDT(-0400)] <ac_chan> it stills says pags root..
[14:50:05 EDT(-0400)] <ac_chan> should i change it?
[14:50:11 EDT(-0400)] <EricDalquist> nope, that is fine
[14:50:22 EDT(-0400)] <EricDalquist> so what you do is in Groups Manager, edit the everyone group
[14:50:30 EDT(-0400)] <EricDalquist> add a member
[14:50:40 EDT(-0400)] <EricDalquist> search for a group with the name of your root PAGS group
[14:50:43 EDT(-0400)] <EricDalquist> and add that
[14:51:04 EDT(-0400)] <EricDalquist> then in your PAGs config make sure all of your PAGs groups are listed as children of the root PAGs group
[14:53:37 EDT(-0400)] <ac_chan> yeah..
[14:53:40 EDT(-0400)] <ac_chan> i did that
[14:53:57 EDT(-0400)] <ac_chan> for some reason on the searc there were to PAGS Root
[14:54:20 EDT(-0400)] <EricDalquist> hrm, that is strange
[14:54:37 EDT(-0400)] <ac_chan> yeah. but now i added the PAGS Root
[14:54:52 EDT(-0400)] <EricDalquist> I have to run to a meeting right now ... I'll be back in a while but Dan Ellentuck is active on the uportal-user list and is very good and GAP related questions
[14:54:54 EDT(-0400)] <ac_chan> and i see all the child groups i created in the xml file
[14:55:44 EDT(-0400)] <ac_chan> oh okay.
[14:55:47 EDT(-0400)] <ac_chan> thanks eric
[15:33:57 EDT(-0400)] <ac_chan> exit
[15:40:56 EDT(-0400)] * ac_chan (n=alex@tempoutsidepix.pratt.edu) has joined ##uportal
[16:07:38 EDT(-0400)] * michelled (n=michelle@142.150.154.114) has left ##uportal
[16:54:23 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[16:56:55 EDT(-0400)] <agherna> ok, here's a question
[16:57:42 EDT(-0400)] <agherna> what's the deal with the deployPortletApp target in the Ant file? I seem to remember that the issue was centered around how the DTD for web.xml was being resolved, but I also remember that not being the only issue. Is that part of it?
[16:58:40 EDT(-0400)] <esm> i think another issue was that the EntityResolver implementation was not correctly forming absolute paths on *nix based systems
[17:03:22 EDT(-0400)] <EricDalquist> it should all be fixed in 2.6.1-RC2
[17:04:49 EDT(-0400)] <agherna> when will 2.6.1GA be available again?
[17:12:57 EDT(-0400)] <EricDalquist> when I get time to test the proxy portlet bug that was reported and then release it
[17:13:03 EDT(-0400)] <EricDalquist> or when someone else gets time to do that
[17:13:15 EDT(-0400)] <EricDalquist> I probably won't have time until next week
[17:24:04 EDT(-0400)] <esm> later all
[17:28:27 EDT(-0400)] <agherna> have a good evening everyone...
[17:28:30 EDT(-0400)] * agherna (n=argherna@cites-agherna01.ci.uiuc.edu) has left ##uportal