uPortal IRC Logs-2009-12-04

Owned by Apereo Infrastructure
Last updated: Dec 04, 2009Version comment
6 min read

[00:01:45 EST(-0500)] * lennard1 (n=sparhk@ip68-98-56-21.ph.ph.cox.net) has joined ##uportal
[00:27:38 EST(-0500)] * lennard1 (n=sparhk@ip68-98-56-21.ph.ph.cox.net) has left ##uportal
[09:07:58 EST(-0500)] * philosophia (n=rjm@wireless-128-135-241-159.uchicago.edu) has joined ##uportal
[09:19:51 EST(-0500)] * athena (n=athena@adsl-75-58-127-15.dsl.wlfrct.sbcglobal.net) has joined ##uportal
[09:36:08 EST(-0500)] * michelled (n=michelle@142.150.154.193) has joined ##uportal
[09:36:25 EST(-0500)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[09:45:08 EST(-0500)] * Sememmon (n=Sememmon@unaffiliated/sememmon) has joined ##uportal
[09:53:42 EST(-0500)] * colinclark (n=colin@bas2-toronto09-1176131079.dsl.bell.ca) has joined ##uportal
[09:59:50 EST(-0500)] * lennard1 (n=sparhk@ip68-98-56-21.ph.ph.cox.net) has joined ##uportal
[10:03:24 EST(-0500)] * holdorph (n=holdorph@wsip-72-215-204-133.ph.ph.cox.net) has joined ##uportal
[10:13:40 EST(-0500)] * philosophia (n=rjm@wireless-128-135-241-159.uchicago.edu) has joined ##uportal
[10:41:14 EST(-0500)] <athena> wow, i'd totally forgotten about dbloader
[10:41:23 EST(-0500)] <EricDalquist> heh
[10:41:24 EST(-0500)] <EricDalquist> yeah
[10:52:32 EST(-0500)] <athena> ok, we've got some problems in the 3.0 branch
[10:52:48 EST(-0500)] <EricDalquist> what's up?
[10:53:13 EST(-0500)] <athena> during initportal: http://uportal.pastebin.com/m284bf904
[10:53:28 EST(-0500)] <athena> only a couple tables got created, looks like
[10:53:36 EST(-0500)] <EricDalquist> hrm
[10:53:39 EST(-0500)] <athena> maybe the table creation goal just isn't running
[10:53:44 EST(-0500)] <EricDalquist> yeah ...
[10:53:46 EST(-0500)] <athena> the only ones there are the hibernate-managed ones
[10:56:14 EST(-0500)] <athena> looks like there are a lot of merged revisions from trunk in the history
[10:56:29 EST(-0500)] <EricDalquist> yeah, but I can't think of one that should affect dbloader
[10:56:33 EST(-0500)] <athena> ah
[10:56:37 EST(-0500)] <EricDalquist> are you see it run at all during initportal?
[10:56:51 EST(-0500)] <EricDalquist> can you capture the initportal output against a clean db and post that?
[10:58:41 EST(-0500)] <athena> yeah
[11:00:59 EST(-0500)] <athena> oooh i may have left some lines out of dbloader
[11:01:02 EST(-0500)] <athena> lets try this again
[11:04:06 EST(-0500)] <athena> it's like my brain emptied itself of all 3.0 knowledge
[11:04:11 EST(-0500)] <EricDalquist> lol
[11:04:14 EST(-0500)] <EricDalquist> that's not all bad (wink)
[11:05:06 EST(-0500)] <athena> lol
[11:05:14 EST(-0500)] <athena> no, it probably contributes to me feeling better about the world
[11:05:43 EST(-0500)] <athena> but having kept around a test instance would be useful right now (smile)
[11:05:59 EST(-0500)] * michelled (n=michelle@142.150.154.193) has joined ##uportal
[11:21:31 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[11:55:09 EST(-0500)] <EricDalquist> great dailwtf today
[11:55:10 EST(-0500)] <EricDalquist> http://thedailywtf.com/Articles/Special-Delivery.aspx
[11:57:47 EST(-0500)] <athena> oh man
[11:57:49 EST(-0500)] <athena> lol
[11:59:02 EST(-0500)] <athena> i'll have to forward that one to kris
[12:07:10 EST(-0500)] <athena> so, for that IE8 issue
[12:07:29 EST(-0500)] <athena> for me it generates an error, but if i step through it the screen still fills in with content
[12:07:40 EST(-0500)] <EricDalquist> yeah
[12:07:42 EST(-0500)] <EricDalquist> same fore me
[12:07:53 EST(-0500)] <EricDalquist> I think if you step through it in IE8 it does different error handling
[12:08:03 EST(-0500)] <EricDalquist> so I don't think it is the xml retrieval
[12:08:19 EST(-0500)] <EricDalquist> I think it is some css attribute setting that is causing the problem
[12:09:24 EST(-0500)] <athena> yeah, you're right - i see the reference to z-index
[12:09:32 EST(-0500)] <athena> from the debugger, looks like it's getting set to NaN
[12:10:06 EST(-0500)] <EricDalquist> yeah
[12:10:09 EST(-0500)] <athena> or maybe that's the current value
[12:10:11 EST(-0500)] <athena> need to look
[12:10:24 EST(-0500)] <EricDalquist> and it appears to be caused by simply calling dialog("open") on jqui
[12:15:54 EST(-0500)] <athena> yeah . . .
[12:15:56 EST(-0500)] <athena> that's not good
[12:16:11 EST(-0500)] <EricDalquist> so I think we're stuck with having to upgrade jqueryui
[12:16:16 EST(-0500)] <EricDalquist> and I don't know how hard that is
[12:17:22 EST(-0500)] <athena> we probably don't want to do that (smile)
[12:17:25 EST(-0500)] <athena> let me try something
[12:19:20 EST(-0500)] <athena> hm.
[12:37:12 EST(-0500)] <athena> this really is looking like a jquery ui bug :/
[12:37:36 EST(-0500)] <EricDalquist> yeah
[12:37:45 EST(-0500)] <EricDalquist> like IE8 changed the default for some property
[12:37:50 EST(-0500)] <EricDalquist> and jqui doesn't handle it
[12:38:08 EST(-0500)] <athena> yeah
[12:38:44 EST(-0500)] <athena> from walking through the stack trace it looks like it's trying to get the current z-index, manipulate it, and re-set it
[12:38:54 EST(-0500)] * holdorph (n=holdorph@wsip-72-215-204-133.ph.ph.cox.net) has joined ##uportal
[12:39:11 EST(-0500)] <EricDalquist> can we fix it by setting a default zindex on the element?
[12:41:07 EST(-0500)] <athena> heh
[12:41:09 EST(-0500)] <athena> well, that worked
[12:41:26 EST(-0500)] <athena> no, we can't set a default because it seems to be an element that jquery is creating itself
[12:41:33 EST(-0500)] <athena> i changed the line in jquery ui from:
[12:42:01 EST(-0500)] <athena> uiDialog.css("z-index",++maxZ || 0)};
[12:42:06 EST(-0500)] <athena> to that from:
[12:42:13 EST(-0500)] <athena> uiDialog.css("z-index",++maxZ)}
[12:42:24 EST(-0500)] <athena> so that the maxZ value just defaults to an actual number if its NaN
[12:42:29 EST(-0500)] <athena> and the error message goes away
[12:42:33 EST(-0500)] <athena> not really ideal though
[12:42:44 EST(-0500)] <EricDalquist> yeah
[12:42:54 EST(-0500)] <EricDalquist> can you find that code in the latest jqui and see what they do?
[12:44:24 EST(-0500)] <athena> yeah, i'll try
[12:44:26 EST(-0500)] <athena> does IE8 have a console?
[12:44:34 EST(-0500)] <EricDalquist> hit f12
[12:47:02 EST(-0500)] <athena> i don't think my fusion instance knows to make that go to the mac (smile)
[12:48:30 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[12:54:35 EST(-0500)] <athena> this code doesnt' really exist in 1.7.2
[12:54:37 EST(-0500)] <athena> it's just too different
[12:54:58 EST(-0500)] <EricDalquist> :/
[12:57:13 EST(-0500)] <athena> well
[12:57:34 EST(-0500)] <athena> to be honest, i think we'd probably be ok just using the change i had above, because we're manually setting the z-index after that code runs anyway
[12:57:46 EST(-0500)] <EricDalquist> yeah
[12:57:49 EST(-0500)] <EricDalquist> I think that's fine too
[12:59:31 EST(-0500)] <athena> it's not ideal, but it'd at least make the error away
[12:59:31 EST(-0500)] <athena> do
[12:59:32 EST(-0500)] <athena> cu
[12:59:36 EST(-0500)] <athena> document that we did it, etc.
[12:59:45 EST(-0500)] <EricDalquist> yeah
[12:59:53 EST(-0500)] <EricDalquist> and probably safer than upgrading jqueryui
[13:00:14 EST(-0500)] <athena> yeah, we can't really do that
[13:00:22 EST(-0500)] <athena> because we'd have to upgrade jquery and fluid as well
[13:00:53 EST(-0500)] <athena> and it's especially bad because we didn't have the resource server stuff back then so people's portlets are written directly to the portal's js
[13:01:47 EST(-0500)] <EricDalquist> yup
[13:09:32 EST(-0500)] <athena> interesting
[13:09:39 EST(-0500)] <athena> so the problem happens on firefox too
[13:09:45 EST(-0500)] <athena> basically the difference is just in error handling
[13:09:49 EST(-0500)] <EricDalquist> ah
[13:09:52 EST(-0500)] <EricDalquist> ff keeps going
[13:09:54 EST(-0500)] <EricDalquist> ie stops
[13:09:55 EST(-0500)] <EricDalquist> ?
[13:10:50 EST(-0500)] <athena> yeah, IE generates a debug message when you try and assign the NaN variable there
[13:10:52 EST(-0500)] <athena> weird
[13:11:00 EST(-0500)] <athena> anyway, i think i've found the right spot to fix this
[13:11:24 EST(-0500)] <athena> it has code that attempts to parse the current value, and then selects the max of that value and the current value (initialized to 0)
[13:11:44 EST(-0500)] <athena> so when it parses, we can just have it use the parse value, or if none exists, use the current value
[13:12:29 EST(-0500)] <athena> so we get:
[13:12:32 EST(-0500)] <athena> maxZ=Math.max(maxZ,parseInt($(this).css("z-index"),10)||maxZ)});
[13:12:38 EST(-0500)] <EricDalquist> great
[13:12:52 EST(-0500)] <athena> seems to work fine, and i'm pretty confident that's really what the code's intention is, and that it won't cause side effects
[13:13:03 EST(-0500)] <athena> so i'll check that in and we can move on with life (smile)
[13:13:28 EST(-0500)] <athena> and that'll fix it everywhere (the edit layout and select skin dialogs were affected by the same bug)
[13:15:04 EST(-0500)] <EricDalquist> sounds good
[13:20:21 EST(-0500)] <athena> looks like we have a bug-free 3.0.5 (smile)
[13:20:42 EST(-0500)] <EricDalquist> woohoo!
[13:20:46 EST(-0500)] <EricDalquist> and 3.1 is getting close
[13:21:22 EST(-0500)] <athena> awesome (smile)
[13:54:26 EST(-0500)] <athena> so for the delegated publishing permissions in UP-2047
[13:54:53 EST(-0500)] <athena> it seems like maybe we should count any channel that has an auto-publish date set as requiring publish-level permissions to edit it?
[14:09:55 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[14:57:19 EST(-0500)] * poofman (i=80876350@gateway/web/freenode/x-byvyvkxfqtszyqlk) has joined ##uportal
[14:59:48 EST(-0500)] <poofman> Greetings uportal devs: I'm running the stats module created by EricDalquist. http://uportal.pastebin.com/m53a0b2f8 I have a trace here in the logs and am not sure if it means bad data or maybe just an issue with the tools cleanup process?
[15:01:19 EST(-0500)] <EricDalquist> perhaps the db contsraints weren't created with the 'on delete cascade' that they should have been?
[15:01:33 EST(-0500)] <EricDalquist> I'm not sure if Hibernate does that automatically or not
[15:01:57 EST(-0500)] <EricDalquist> but all the foreign key constraints in the STATS_ tables need to have 'on delete cascade'
[15:02:34 EST(-0500)] <poofman> ok is that something I can modify in the create script then?
[15:02:50 EST(-0500)] <EricDalquist> those tables are created by uPortal via Hibernate
[15:03:04 EST(-0500)] <EricDalquist> I can't remember if it is hibernate config or a manual thing
[15:03:54 EST(-0500)] <poofman> ok I'll dig a little to try and find it. I also just noticed this one http://uportal.pastebin.com/m8f418c2 . Do I maybe have an older version of the sql startup files?
[15:04:12 EST(-0500)] <poofman> looks like a missing view.
[15:04:15 EST(-0500)] <EricDalquist> looks like a table space issue
[15:04:32 EST(-0500)] <EricDalquist> you probably need to find where it is looking for the schema name PROD_STATS and replace it with your own
[15:04:44 EST(-0500)] <poofman> yep. I see that now . thanks.
[16:25:19 EST(-0500)] * colinclark_ (n=colin@142.150.154.130) has joined ##uportal
[16:46:50 EST(-0500)] * poofman (i=80876350@gateway/web/freenode/x-vlxgmfdcrfivfntm) has joined ##uportal
[16:49:20 EST(-0500)] <poofman> Greetings EricDalquist. not sure if your'e gone for the day: http://uportal.pastebin.com/m35f69333 I was working with an oracle dba about the 'cascade on delete' I need to add to the stats_ tables for the statistics.
[16:50:01 EST(-0500)] <poofman> The can add the extra clause, but not sure where it needs to be applied. I'm not sure as well and wanted to ask you.
[16:50:17 EST(-0500)] <EricDalquist> ah
[16:50:20 EST(-0500)] <EricDalquist> I'm just on my way out
[16:50:29 EST(-0500)] <EricDalquist> it is on ever FK constraint used by the STATS_* tables though
[16:50:37 EST(-0500)] <poofman> figured. this is very late in the day to ask you about this.
[16:50:45 EST(-0500)] <EricDalquist> that allows the tool to delete a stats session and have all related events go with it
[16:51:23 EST(-0500)] <poofman> ok. I'll relay this info back to him. I think he just wanted to make sure if was for every stats_ table. Thanks again for your help. have a great weekend.
[16:51:47 EST(-0500)] <EricDalquist> thanks, you too
[17:44:29 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined ##uportal
[18:31:29 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined ##uportal