Versions Compared

Old Version 40

changes.mady.by.user Apereo Infrastructure

Saved on

compared with

New Version 41

changes.mady.by.user Apereo Infrastructure

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

[08:30:22 CDT(-0500)] <athena1> http://pastebin.com/kyB3bHnz
[10:33:08 CDT(-0500)] <athena> hey EricDalquist
[10:33:13 CDT(-0500)] <EricDalquist> hi athena
[10:33:26 CDT(-0500)] <athena> so it turns out that the last problem with the etags was that i'm an idiot
[10:33:33 CDT(-0500)] <athena> and that i never noticed that that request was still a POST
[10:33:34 CDT(-0500)] <athena> sigh
[10:33:35 CDT(-0500)] <EricDalquist> lol
[10:33:39 CDT(-0500)] <athena> yeah, go figure.
[10:33:46 CDT(-0500)] <athena> so nick's fix did fix everything, i think
[10:33:48 CDT(-0500)] <athena> which is awesome
[10:33:49 CDT(-0500)] <athena> so.
[10:33:50 CDT(-0500)] <EricDalquist> glad you got it figured out
[10:33:52 CDT(-0500)] <EricDalquist> great
[10:33:53 CDT(-0500)] <athena> now on to the next issue!
[10:33:54 CDT(-0500)] <athena> http://pastebin.com/kyB3bHnz
[10:34:01 CDT(-0500)] <EricDalquist> I still want to get expires and cache-control working
[10:34:04 CDT(-0500)] <EricDalquist> but that's lower priority
[10:34:12 CDT(-0500)] <athena> for some reason i'm sometimes getting this on the first request, but only in the mobile theme
[10:34:16 CDT(-0500)] <athena> no idea what's up with that
[10:34:34 CDT(-0500)] <EricDalquist> man pastebin mangles wide text :/
[10:34:42 CDT(-0500)] <athena> i know
[10:35:16 CDT(-0500)] <EricDalquist> so this is in a vanilla trunk
[10:35:22 CDT(-0500)] <EricDalquist> sometimes on the first request
[10:35:26 CDT(-0500)] <EricDalquist> you just get this stack trace
[10:35:39 CDT(-0500)] <athena> first request specifically to the news reader
[10:35:44 CDT(-0500)] <EricDalquist> ah
[10:35:44 CDT(-0500)] <EricDalquist> ok
[10:35:47 CDT(-0500)] <athena> not sure i've seen it elsewhere, though i haven't done a lot of testing
[10:35:51 CDT(-0500)] <athena> could try it out
[10:36:17 CDT(-0500)] <EricDalquist> that is bizar
[10:36:28 CDT(-0500)] <athena> yes.
[10:36:30 CDT(-0500)] <athena> sigh
[10:36:34 CDT(-0500)] <EricDalquist> hrm
[10:36:42 CDT(-0500)] <EricDalquist> and this is when you click on the news reader
[10:36:44 CDT(-0500)] <EricDalquist> so it is rendering?
[10:36:51 CDT(-0500)] <EricDalquist> or when you render the grid of icons
[10:36:51 CDT(-0500)] <athena> yeah
[10:36:54 CDT(-0500)] <EricDalquist> ok
[10:36:59 CDT(-0500)] <athena> when loading the portlet
[10:37:14 CDT(-0500)] <EricDalquist> and it is on the serve resource request ...
[10:37:16 CDT(-0500)] <athena> i know we had some problems before where the mobile theme was misconfigured
[10:37:20 CDT(-0500)] <athena> thought we'd sorted that out though
[10:37:27 CDT(-0500)] <athena> unless i accidentally reintroduced something
[11:12:55 CDT(-0500)] <athena> and of course now i can't reproduce it
[11:12:59 CDT(-0500)] <athena> whaaatever.
[11:13:15 CDT(-0500)] <EricDalquist> lol
[11:15:54 CDT(-0500)] <athena> yeah.
[11:15:58 CDT(-0500)] <athena> it's going to be one of those days.
[13:38:12 CDT(-0500)] <athena1> looks like we're whittling down the list of bugs at least
[13:39:58 CDT(-0500)] <EricDalquist> yup
[13:40:09 CDT(-0500)] <EricDalquist> working on making things work if you delete a portlet right now
[13:40:24 CDT(-0500)] <athena1> oh terrific
[13:40:49 CDT(-0500)] <athena1> is that that bug that causes the layout to crash if a portlet is missing?
[13:41:06 CDT(-0500)] <EricDalquist> yup
[13:41:14 CDT(-0500)] <athena1> yay (smile)
[13:41:25 CDT(-0500)] <athena1> testing a potential fix for that missing save and configure button right now
[13:41:30 CDT(-0500)] <EricDalquist> there are like a dozen little bugs that all contribute to it
[13:41:36 CDT(-0500)] <athena1> ah
[13:41:37 CDT(-0500)] <EricDalquist> essentially a bunch of missing null checks
[13:41:42 CDT(-0500)] <athena1> yeah, makes sense
[13:41:43 CDT(-0500)] <EricDalquist> oh great
[13:43:14 CDT(-0500)] <athena1> think it might be a really simple fix, but we'll see
[13:43:58 CDT(-0500)] <EricDalquist> I'm attempting to put down in words why I think we should move uPortal to git
[13:44:10 CDT(-0500)] <EricDalquist> I'm guessing "because its awesome" isn't a good answer (tongue)
[13:44:12 CDT(-0500)] <athena1> lol
[13:44:22 CDT(-0500)] <athena1> so it'll make it easier for non-committers to submit patches
[13:44:28 CDT(-0500)] <athena1> make it easier for us to work with other communities
[13:44:49 CDT(-0500)] <athena1> like say we wanted to work w/ grouper or kuali on a group store implementation, life would be a lot easier
[13:44:54 CDT(-0500)] <EricDalquist> yup
[13:45:00 CDT(-0500)] <athena1> hopefully should make it easier to maintain other branches - like the umobile project
[13:45:09 CDT(-0500)] <athena1> and the uhhh git is awesome, whatever, etc.
[13:45:16 CDT(-0500)] <EricDalquist> (smile)
[13:45:19 CDT(-0500)] <athena1> could cite the recent successful migration by fluid
[13:45:24 CDT(-0500)] <EricDalquist> yes
[13:45:25 CDT(-0500)] <athena1> i know CAS is also planning to move to git
[13:45:26 CDT(-0500)] <EricDalquist> I am
[13:57:11 CDT(-0500)] <drewwills1> good afternoon folks... quick sanity check... the portlet windowId is unique to that portlet AND that user, right? so 2 users using the same portlet on the same fragment would have different IDs?
[14:06:11 CDT(-0500)] <athena1> EricDalquist: you have any thoughts about how to turn off XML escaping for JSR-286 portlets by default?
[14:06:42 CDT(-0500)] <holdorph> Drew, if i'm remembering the right 'thing', i believe window id is unique to the user AND even that particular login session for that user.
[14:51:16 CDT(-0500)] <EricDalquist> holdorph is right
[14:51:31 CDT(-0500)] <EricDalquist> athena1: we had talked about disabling it by default for 168 portlets
[14:51:49 CDT(-0500)] <EricDalquist> for 286 portlets the spec says they have to explicitly request it be disabled in their portlet.xml file
[14:52:11 CDT(-0500)] <drewwills1> that's good, thanks... but I got a bit ahead of myself... forgot the getWindowId method is portlet 2.0
[14:52:19 CDT(-0500)] <EricDalquist> (smile)
[14:53:17 CDT(-0500)] <drewwills1> eric... suggestion for the ehcache-spring-annotations lib... add a new annotation @RefreshMeIfTrue
[14:53:49 CDT(-0500)] <EricDalquist> can you give a functional example?
[14:54:03 CDT(-0500)] <EricDalquist> also if you're interested in caching anotations: https://github.com/jsr107/jsr107spec
[14:54:12 CDT(-0500)] <EricDalquist> and our first firm draft of the API: https://jsr107.ci.cloudbees.com/job/jsr107api/ws/target/apidocs/index.html
[14:54:13 CDT(-0500)] <drewwills1> so i could do something like...public NotificationResponse getNotifications(@PartialCacheKey String notificationsContextName, @PartialCacheKey String remoteUser, @RefreshMeIfTrue boolean doRefresh)
[14:54:43 CDT(-0500)] <drewwills1> avoid having to write a triggersRemove method
[14:54:45 CDT(-0500)] <EricDalquist> ah so if that last parameter is true a cached value is ignored
[14:54:51 CDT(-0500)] <drewwills1> yeppers
[14:55:04 CDT(-0500)] <EricDalquist> the 107 spec adds @CachePuit
[14:55:09 CDT(-0500)] <EricDalquist> er @CachePut
[14:55:22 CDT(-0500)] <EricDalquist> which is like a combo of @CacheRemove/@Cacheable
[14:55:44 CDT(-0500)] <drewwills1> hmmm... i should look
[14:55:51 CDT(-0500)] <drewwills1> i really like this library
[14:55:57 CDT(-0500)] <EricDalquist> the 107spec is at 0.2
[14:56:13 CDT(-0500)] <EricDalquist> I'm supposed to be writing a spring impl of the annotations support this week (tongue)
[14:56:28 CDT(-0500)] <drewwills1> don't you have enough to do?
[14:56:36 CDT(-0500)] <EricDalquist> and once it is finalized the 2.0 release of ehache-spring-annotations will implement it
[14:56:39 CDT(-0500)] <EricDalquist> apparently not
[14:56:44 CDT(-0500)] <EricDalquist> couldn't say no to being on a JSR though
[14:56:51 CDT(-0500)] <drewwills1> yeah i follow that
[14:57:53 CDT(-0500)] <drewwills1> you should see the notifications portlet... i'll post screenshots when i can get my ehad above water
[14:59:38 CDT(-0500)] <athena1> EricDalquist: yeah, remember the conversation - was wondering if that was something i could help out wiht
[15:00:00 CDT(-0500)] <EricDalquist> huh?
[15:00:24 CDT(-0500)] <athena1> sorry, about disabling the XML escaping for JSR-168 portlets (smile)
[15:01:21 CDT(-0500)] <EricDalquist> brb
[15:09:55 CDT(-0500)] <EricDalquist> sjflak.shdfajk.sdhfljkashdfn
[15:10:01 CDT(-0500)] <EricDalquist> so much random stuff going on
[15:10:03 CDT(-0500)] <EricDalquist> ok
[15:10:20 CDT(-0500)] <EricDalquist> so I don't really have any concrete ideas for that issue
[15:10:25 CDT(-0500)] <EricDalquist> other than:
[15:10:46 CDT(-0500)] <EricDalquist> - need to figure out where in pluto this XML escaping flag is handled
[15:11:05 CDT(-0500)] <athena1> ok (smile)
[15:11:13 CDT(-0500)] <EricDalquist> - then need to figure out if there is a way to tell a 1.0 portlet app from a 2.0 portlet app in the pluto data model
[15:11:32 CDT(-0500)] <EricDalquist> - then need to figure out if we can override that xml escaping flag either in the data model or where it is uses for 1.0 apps
[15:11:34 CDT(-0500)] <athena1> gotcha
[15:11:42 CDT(-0500)] <athena1> sounds irritating
[15:11:48 CDT(-0500)] <EricDalquist> also, need to double check in the spec and see what it says the behavior should be for 1.0 portlets
[15:11:49 CDT(-0500)] <EricDalquist> yes
[15:11:53 CDT(-0500)] <EricDalquist> I'm sure it will be (smile)
[15:11:57 CDT(-0500)] <athena1> lol
[15:13:19 CDT(-0500)] <athena1> oh interesting
[15:13:29 CDT(-0500)] <athena1> looks like maybe we can set it as a container runtime option
[15:13:38 CDT(-0500)] <athena1> http://developers.sun.com/portalserver/reference/techart/jsr286/jsr286_2.html#Container_runtime_options
[15:13:58 CDT(-0500)] <EricDalquist> yes
[15:14:00 CDT(-0500)] <athena1> does that seem like what we want?
[15:14:04 CDT(-0500)] <EricDalquist> for 286 portlets
[15:14:09 CDT(-0500)] <EricDalquist> you can disable it portlet-app wide
[15:14:22 CDT(-0500)] <EricDalquist> but container-runtime-option is a 286 specific element
[15:14:40 CDT(-0500)] <EricDalquist> so it doesn't help us with 168 portlets
[15:14:45 CDT(-0500)] <athena1> so you can't use it w/o upgrading?
[15:14:46 CDT(-0500)] <athena1> that's annoying
[15:14:50 CDT(-0500)] <EricDalquist> right
[15:14:51 CDT(-0500)] <athena1> paragraph made it sound more useful
[15:15:01 CDT(-0500)] <athena1> guess that's only if you upgrade the jar and don't want to port your code?
[15:15:12 CDT(-0500)] <EricDalquist> there is no portlet jar in portlets
[15:15:18 CDT(-0500)] <EricDalquist> you just have to update the portlet.xml file to the 2.0 XSD
[15:15:33 CDT(-0500)] <EricDalquist> but then it will only work on a 2.0 container
[15:15:43 CDT(-0500)] <athena1> ah right
[15:15:44 CDT(-0500)] <athena1> hmm
[15:15:56 CDT(-0500)] <athena1> what about spring? we'd have to upgrade that too, right?
[15:16:01 CDT(-0500)] <EricDalquist> the question I have is: What does the spec say the behavior should be for a 168 portlet in a 286 container with regard to XML escaping
[15:16:02 CDT(-0500)] <EricDalquist> no
[15:16:05 CDT(-0500)] <athena1> because otherwise i'd say lets just use a portlet.xml overlay for now . . .
[15:16:12 CDT(-0500)] <EricDalquist> Spring 2.5 works with 286
[15:16:21 CDT(-0500)] <EricDalquist> I bet a portlet.xml overlay would work
[15:16:40 CDT(-0500)] <EricDalquist> 286 is API compatible with 168
[15:16:40 CDT(-0500)] <athena1> maybe we should just do that for now so that those portlets work for the release, and figure it out for real later?
[15:16:49 CDT(-0500)] <EricDalquist> so there is nothing that a 168 library can do that won't work in 286
[15:16:55 CDT(-0500)] <athena1> yeah guess you just wouldnt' have access to the new features, but that's fine
[15:16:56 CDT(-0500)] <EricDalquist> sounds good
[15:17:01 CDT(-0500)] <athena1> ok, i'll try it out
[15:17:05 CDT(-0500)] <EricDalquist> great
[15:17:05 CDT(-0500)] <athena1> if that works, lets just go with it
[15:17:18 CDT(-0500)] <athena1> don't want these portlets to be so broken, but don't want it to impact our release schedule
[15:17:25 CDT(-0500)] <athena1> google gadgets are just not that cool (tongue)
[15:17:35 CDT(-0500)] <EricDalquist> lol
[15:18:15 CDT(-0500)] <athena1> more worried about simple content, i suppose
[15:18:21 CDT(-0500)] <athena1> especially since it's all cool and search-y (smile)
[15:20:52 CDT(-0500)] <EricDalquist> (smile)
[15:20:54 CDT(-0500)] <EricDalquist> but its 286
[15:21:01 CDT(-0500)] <EricDalquist> so you can just set that container runtime option
[15:21:12 CDT(-0500)] <athena1> oh right!
[15:21:17 CDT(-0500)] <athena1> ok, so it really is only google gadgets
[15:23:37 CDT(-0500)] <EricDalquist> wtf ... this peoplesoft server is returning a MIME encoded email to a SOAP request
[15:24:59 CDT(-0500)] <athena1> errrr . . .
[15:25:00 CDT(-0500)] <athena1> lol
[15:25:07 CDT(-0500)] <EricDalquist> yeah
[15:30:14 CDT(-0500)] <EricDalquist> so they think it is a "cert issue" and that is why I'm getting an email back for a soap request (tongue)
[15:30:20 CDT(-0500)] <EricDalquist> I have little faith
[15:33:26 CDT(-0500)] <athena1> lol
[15:33:28 CDT(-0500)] <athena1> no kidding.
[15:46:57 CDT(-0500)] <athena1> errrr, huh.
[15:46:58 CDT(-0500)] <athena1> IllegalStateException: A flow execution action URL can only be obtained in a RenderRequest or a ResourceRequest
[15:47:02 CDT(-0500)] <athena1> don't remember seeing that one before
[15:47:25 CDT(-0500)] <EricDalquist> weird
[15:47:39 CDT(-0500)] <EricDalquist> what triggered it?
[15:47:57 CDT(-0500)] <athena1> simple content portlet config mode
[15:48:04 CDT(-0500)] <athena1> went to test it's behavior before making any changes
[15:48:14 CDT(-0500)] <EricDalquist> huh
[15:48:31 CDT(-0500)] <athena1> hmm, looks like it is double-escaping though
[15:48:33 CDT(-0500)] <EricDalquist> I think what it is saying is that createActionUrl only exists on Resource and Render responses
[15:48:34 CDT(-0500)] <athena1> so maybe that did somethign weird
[15:48:44 CDT(-0500)] <athena1> odd though
[15:49:53 CDT(-0500)] <athena1> not 100% clear on why that URL is double-escaped
[15:50:07 CDT(-0500)] <athena1> action="/uPortal/p/portlet-admin.ctf3/max/action.uP?plP_execution=e1s3&amp;amp;plP__eventId=configModeAction&amp;amp;plCa=61_dlg-21-ctf3-5_5&amp;amp;plCd_61_dlg-21-ctf3-5_5=21_ctf3_5&amp;amp;plP_61_dlg-21-ctf3-5_5_action=updateConfiguration"
[15:50:16 CDT(-0500)] <athena1> that shouldn't look like that, right?
[15:50:40 CDT(-0500)] <athena1> that's what's actually in the source of hte page
[15:50:41 CDT(-0500)] <EricDalquist> no
[15:50:47 CDT(-0500)] <EricDalquist> it should not look like that