Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 34 Next »

[01:00:53 EDT(-0400)] * Sememmon (~Sememmon@unaffiliated/sememmon) has joined ##uportal
[08:59:17 EDT(-0400)] * awills (~awills@c-68-42-121-215.hsd1.mi.comcast.net) has joined ##uportal
[09:21:44 EDT(-0400)] * athena (~athena@adsl-99-89-93-113.dsl.wlfrct.sbcglobal.net) has joined ##uportal
[09:35:34 EDT(-0400)] * jessm (~Jess@c-71-232-3-151.hsd1.ma.comcast.net) has joined ##uportal
[09:50:06 EDT(-0400)] * EricDalquist (~apollo@76.210.71.208) has joined ##uportal
[09:50:25 EDT(-0400)] <EricDalquist> so I stumbled onto this project yesterday: https://jawr.dev.java.net/
[09:50:46 EDT(-0400)] <EricDalquist> seems to be a much more thorough version of our css/js aggregator in 3.2
[09:52:02 EDT(-0400)] <EricDalquist> something to look at for 3.3 perhaps
[09:52:36 EDT(-0400)] <EricDalquist> oh and this is tons of fun: http://www.sonatype.com/people/2010/03/introduction-to-maven-shell/
[09:52:50 EDT(-0400)] <EricDalquist> even has growl support to tell me when tasks complete
[10:35:31 EDT(-0400)] * colinclark (~colin@142.150.154.101) has joined ##uportal
[10:48:52 EDT(-0400)] <athena> oh neat
[10:49:00 EDT(-0400)] <athena> those both sound interesting (smile)
[10:52:12 EDT(-0400)] * michelled (~michelled@142.150.154.101) has joined ##uportal
[11:02:05 EDT(-0400)] <athena> do we want to have shared message keys for common strings like "next" and "save"?
[11:02:38 EDT(-0400)] <athena> it seems like it might just make it harder to maintain and do internationalization if we keep repeating stuff like that all over the place
[11:04:03 EDT(-0400)] <EricDalquist> I don't think we do
[11:04:22 EDT(-0400)] <EricDalquist> and we should probably have a chat on list about how and where to locate our i18n files
[11:04:30 EDT(-0400)] <EricDalquist> right now we have a bunch of .properties files for the flows
[11:04:36 EDT(-0400)] <EricDalquist> and an XML file for the theme
[11:09:38 EDT(-0400)] <athena> yeah, that sounds like a great idea
[11:09:52 EDT(-0400)] <athena> i wonder if they were more unified if it'd be easier for people to work with
[11:10:05 EDT(-0400)] <athena> would be good to get some input from anyone who's actually doing i18n
[11:10:10 EDT(-0400)] <athena> i'll make a note to email the list about that
[11:10:11 EDT(-0400)] <EricDalquist> well since java Properties supports XML
[11:10:33 EDT(-0400)] <EricDalquist> we could perhaps switch from .properties to .xml for flows
[11:10:38 EDT(-0400)] <athena> interesting
[11:10:40 EDT(-0400)] <EricDalquist> and then see if we can also use those from the theme xsl
[11:10:41 EDT(-0400)] <athena> no objection here
[11:10:43 EDT(-0400)] <athena> yeah
[11:10:45 EDT(-0400)] <athena> that'd be cool
[11:11:04 EDT(-0400)] <EricDalquist> if we do a decent enough job naming properties and using comments we could even have just one big XML file
[11:11:18 EDT(-0400)] <athena> yes
[11:11:41 EDT(-0400)] <athena> turns out liferay actually builds their translated files through machine translation
[11:11:47 EDT(-0400)] <athena> and has volunteers QA them later
[11:11:59 EDT(-0400)] <EricDalquist> well if we had one big XML file that would potentially be easier
[11:12:04 EDT(-0400)] <athena> yep
[11:12:11 EDT(-0400)] <athena> i suspect that's why they have just one file
[11:13:01 EDT(-0400)] <athena> so another question i had was whether we want to just eliminate the ChannelCategory class and all the category methods in IChannelRegistryStore
[11:13:10 EDT(-0400)] <EricDalquist> fine by me
[11:13:17 EDT(-0400)] <EricDalquist> honestly once we get to portlet only code
[11:13:23 EDT(-0400)] <athena> ChannelCategory is just this weird alternate class that doesn't implement IEntity
[11:13:24 EDT(-0400)] <EricDalquist> IChannelREgistryStore may go away too
[11:13:38 EDT(-0400)] <athena> and the IChannelRegistryStore stuff seems like wrappers for the real groups service code
[11:13:42 EDT(-0400)] <athena> that sounds good then
[11:13:42 EDT(-0400)] <EricDalquist> then you just use a combo of the portlet defininition registry and the groups APIs to get your info
[11:13:53 EDT(-0400)] <athena> yeah
[11:14:16 EDT(-0400)] <athena> at what point do we want to tear out the groups channel?
[11:14:28 EDT(-0400)] <EricDalquist> whenever you're comfortable doing so
[11:14:39 EDT(-0400)] <EricDalquist> I'm working on the pluto 2 stuff still
[11:14:46 EDT(-0400)] <EricDalquist> but once that is done all channels will be useless
[11:14:54 EDT(-0400)] <athena> yeah
[11:15:13 EDT(-0400)] <athena> so what i have now is a portlet that lets you browse groups and categories
[11:15:20 EDT(-0400)] <athena> edit their name and description and manage their children
[11:15:26 EDT(-0400)] <athena> add new groups/categories
[11:15:28 EDT(-0400)] <athena> and delete them
[11:15:37 EDT(-0400)] <athena> it doesn't yet have integration w/ the new permissions manager portlet
[11:15:41 EDT(-0400)] <EricDalquist> sounds good enough though
[11:15:44 EDT(-0400)] <athena> but i think hopefully that's the only missing feature
[11:16:04 EDT(-0400)] <athena> the other difference i've noticed is that the groups channel lets you view class-specific details for each entity type
[11:16:11 EDT(-0400)] <athena> so if you click on a channel you can see it's timeout default
[11:16:16 EDT(-0400)] <athena> though not anything much more useful
[11:16:24 EDT(-0400)] <athena> i'm not convinced that's really all that helpful
[11:16:29 EDT(-0400)] <athena> though it might be for people
[11:16:31 EDT(-0400)] <EricDalquist> yeah
[11:16:37 EDT(-0400)] <athena> but we can add that in later if we need to, i think
[11:16:40 EDT(-0400)] <EricDalquist> eventually we can add some cross-flow linking
[11:16:43 EDT(-0400)] <EricDalquist> find a portlet in groups manager
[11:16:45 EDT(-0400)] <athena> hopefully none of the devs are really relying on that
[11:16:47 EDT(-0400)] <athena> yes
[11:16:49 EDT(-0400)] <EricDalquist> click on it to jump into portlet manager
[11:16:51 EDT(-0400)] <athena> that was kind fo my thought to
[11:16:52 EDT(-0400)] <EricDalquist> find a person
[11:16:53 EDT(-0400)] <athena> yep
[11:16:56 EDT(-0400)] <EricDalquist> click to jump into person manager
[11:16:58 EDT(-0400)] <athena> reuse is good (smile)
[11:19:08 EDT(-0400)] <athena> wonder what breaks if i just delete the groupmanager package . . .
[11:19:15 EDT(-0400)] <EricDalquist> (smile)
[11:20:05 EDT(-0400)] <athena> seems like the permission manager might depend on one of the IServants or something
[11:20:29 EDT(-0400)] <EricDalquist> well if we have a new perms manager too the old one can go as well
[11:21:55 EDT(-0400)] <athena> what do we still use CChannelManager for?
[11:21:57 EDT(-0400)] <athena> anything?
[11:22:17 EDT(-0400)] <EricDalquist> nothing
[11:22:24 EDT(-0400)] <athena> we'd need to remove that tooo
[11:22:26 EDT(-0400)] <EricDalquist> I thought it got deleted for 3.2
[11:22:29 EDT(-0400)] <athena> nope
[11:22:40 EDT(-0400)] <athena> probably should have though
[11:23:05 EDT(-0400)] <athena> want me to take all that out today? doing some coop dev work on the groups manager
[11:23:11 EDT(-0400)] <EricDalquist> sure
[11:23:14 EDT(-0400)] <athena> (smile)
[11:25:10 EDT(-0400)] <athena> what does CUserManager do?
[11:25:29 EDT(-0400)] <EricDalquist> it pretends to allow you to add/delete users
[11:25:31 EDT(-0400)] <EricDalquist> and update passwords
[11:25:40 EDT(-0400)] <EricDalquist> I think there is an item on the 3.3 roadmap to replace it
[11:25:50 EDT(-0400)] <athena> gotcha
[11:25:58 EDT(-0400)] <athena> so we could probably live without that for a few weeks
[11:26:08 EDT(-0400)] <athena> only pretends to let you add/delete? (smile)
[11:26:26 EDT(-0400)] <EricDalquist> I think it works
[11:26:39 EDT(-0400)] <EricDalquist> it seems to work on and off in random versions of uPortal
[11:27:04 EDT(-0400)] <athena> weird.
[11:29:49 EDT(-0400)] <athena> hey, i think it's compiling now
[11:30:03 EDT(-0400)] <athena> so would be removing the groups, permissions, channelmanager, and user manager channels
[11:30:04 EDT(-0400)] <EricDalquist> yay
[11:30:07 EDT(-0400)] <athena> yeah (smile)
[11:30:17 EDT(-0400)] <athena> and if we need a reference for prior behavior, we can always look at 3.2
[11:30:18 EDT(-0400)] <EricDalquist> sounds good to me
[11:30:22 EDT(-0400)] <athena> will do, then!
[11:30:26 EDT(-0400)] <EricDalquist> I'm still hoping to have pluto 2 rendering stuff by the end of the week
[11:30:27 EDT(-0400)] <EricDalquist> yup
[11:50:46 EDT(-0400)] <athena> wow that'd be great to have that so soon (smile)
[11:51:19 EDT(-0400)] <EricDalquist> well it will be very rough
[11:51:24 EDT(-0400)] <EricDalquist> and incomplete
[11:51:31 EDT(-0400)] <EricDalquist> but hopefully minimally functional
[11:52:23 EDT(-0400)] <athena> yes (smile)
[11:52:28 EDT(-0400)] <athena> definitely sounds like a start
[11:52:47 EDT(-0400)] <athena> at what point do you plan to pull it back into the trunk?
[11:53:02 EDT(-0400)] <EricDalquist> I'm not sure
[11:53:23 EDT(-0400)] <EricDalquist> when we get it rendering I'll probably check up with the dev list
[11:53:30 EDT(-0400)] <EricDalquist> and see what people think about the lack of functionality
[11:55:00 EDT(-0400)] <athena> sounds reasonable
[12:35:41 EDT(-0400)] * EricDalquist (~apollo@adsl-76-208-69-152.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[12:50:37 EDT(-0400)] * Sememmon (~Sememmon@unaffiliated/sememmon) has joined ##uportal
[12:58:37 EDT(-0400)] <athena> EricDalquist: what's the current issue w/ window shading in uportal?
[12:59:11 EDT(-0400)] <athena> did you guys have a patch for that?
[13:00:28 EDT(-0400)] <EricDalquist> brb
[13:08:18 EDT(-0400)] <EricDalquist> I added a patch to the jira issue
[13:08:26 EDT(-0400)] <EricDalquist> it works right now
[13:08:34 EDT(-0400)] <EricDalquist> but doesn't persist for DLM managed nodes
[13:08:41 EDT(-0400)] <EricDalquist> so we had created a fix for that
[13:08:51 EDT(-0400)] <EricDalquist> but we haven't moved it out of dev
[13:10:23 EDT(-0400)] <athena> ah ok
[13:10:36 EDT(-0400)] <athena> just wanted to check, since it's on the coop dev list
[13:10:51 EDT(-0400)] <athena> so presumably uw will contribute patch to close that out eventually?
[13:11:00 EDT(-0400)] <EricDalquist> the patch is contributed to the jira issue
[13:11:07 EDT(-0400)] <EricDalquist> it needs further review and testing though
[13:11:12 EDT(-0400)] <EricDalquist> we didn't get that far with it
[13:11:15 EDT(-0400)] <EricDalquist> so it works
[13:11:45 EDT(-0400)] <EricDalquist> but no testing around dealing with what happens when the node is removed, moved, etc
[13:12:40 EDT(-0400)] <athena> i gotcha
[13:12:57 EDT(-0400)] <athena> so maybe we could ask the community to test it
[13:13:23 EDT(-0400)] <EricDalquist> and someone from coop dev to review it as well
[13:13:55 EDT(-0400)] <athena> sounds good
[13:20:50 EDT(-0400)] <athena> hmm
[13:21:04 EDT(-0400)] <athena> so the permission "owners" are all registered in the database
[13:21:11 EDT(-0400)] <athena> under UPC_PERM_MGR
[13:21:32 EDT(-0400)] <athena> except we haven't actually been keeping to that structure
[13:22:01 EDT(-0400)] <athena> we've got stuff registered in UP_PERMISSION now that has an owner that's not in the other table
[13:24:16 EDT(-0400)] <athena> ah, i see
[13:24:20 EDT(-0400)] <athena> it's the token that's mapped
[13:25:38 EDT(-0400)] <athena> anyway, one side effect of deleting CGroupsManager and such is going to be that the IPermissible for groups management is suddenly gone
[13:25:54 EDT(-0400)] <athena> probably going to have to write new IPermissable implementations
[13:46:13 EDT(-0400)] * michelled (~michelled@142.150.154.101) has joined ##uportal
[13:46:27 EDT(-0400)] * colinclark (~colin@142.150.154.101) has joined ##uportal
[14:30:54 EDT(-0400)] * EricDalquist (~apollo@76.208.69.152) has joined ##uportal
[14:43:47 EDT(-0400)] <athena> ugh this PermissibleRegistry stuff is a mess
[14:44:26 EDT(-0400)] <athena> have you seen much of it EricDalquist?
[14:46:00 EDT(-0400)] <EricDalquist> no
[14:46:04 EDT(-0400)] <EricDalquist> but I can imagine
[14:46:45 EDT(-0400)] <athena> yeah (tongue)
[14:46:50 EDT(-0400)] <athena> just trying to figure out what we actually want
[14:46:56 EDT(-0400)] <athena> i think it's functionality worth keeping though?
[14:47:10 EDT(-0400)] <EricDalquist> what does it do right now?
[14:47:15 EDT(-0400)] <EricDalquist> I don't have the code handy to take a look
[14:47:17 EDT(-0400)] <athena> basically allows you to define the permission owner systems and sort of use them as categories
[14:47:29 EDT(-0400)] <athena> that's basically what keeps groups permissions separate from framework permissions, etc.
[14:48:17 EDT(-0400)] <athena> and the permissions table references a key defined by one of the owners to track what system it's from
[14:48:39 EDT(-0400)] <athena> thinking maybe the thing to do is re-write it w/ JPA not using static methods
[14:48:51 EDT(-0400)] <athena> and actually maybe someday add a foreign key from the permission table to the owners table
[14:49:25 EDT(-0400)] <EricDalquist> yeah
[14:49:30 EDT(-0400)] <EricDalquist> I think that sounds like a sane approach
[14:49:45 EDT(-0400)] <EricDalquist> adding the FK when we get to the rest of the GAP code with the refactor-hammer
[14:49:51 EDT(-0400)] <athena> yeah
[14:50:10 EDT(-0400)] <athena> it's also kind of useful because each system tracks all of the allowed activities and such
[14:50:17 EDT(-0400)] <athena> so that's where the list of valid permissions comes in
[14:50:34 EDT(-0400)] <athena> and it tracks the current targets in use for each activity, which is how that auto-populating of the permissions channel happens
[14:50:54 EDT(-0400)] <athena> though that clearly needs to get wrapped with something that'll refresh the cache - looks like it might never get updated right now
[14:51:13 EDT(-0400)] <athena> wondering if we want to someday move the list of activities to the database as well - right now its hardcoded into each permission owner class
[14:51:20 EDT(-0400)] <athena> and of course none of that is internationalizable
[14:51:31 EDT(-0400)] <athena> small steps for now though . . .
[14:56:33 EDT(-0400)] <athena> what is CarResources anyway?
[14:56:46 EDT(-0400)] <EricDalquist> you remember what CARs are?
[14:58:30 EDT(-0400)] <athena> nope.
[14:58:40 EDT(-0400)] <EricDalquist> stuff from SCT
[14:58:41 EDT(-0400)] <athena> not sure i was ever subjected to that
[14:58:44 EDT(-0400)] <EricDalquist> WARs for channels
[14:58:51 EDT(-0400)] <athena> ahhh
[14:58:59 EDT(-0400)] <athena> seems like this is using that class to get a classloader
[14:59:01 EDT(-0400)] <EricDalquist> drop in your CAR and uportal would extract it, deploy it, do classloader magic and everything
[14:59:09 EDT(-0400)] <EricDalquist> it can be delted
[14:59:16 EDT(-0400)] <EricDalquist> since we won;t have channels any more
[14:59:23 EDT(-0400)] <EricDalquist> so much gets simplified with no channels (tongue)
[14:59:33 EDT(-0400)] <athena> (smile)
[14:59:33 EDT(-0400)] <athena> yes
[15:01:20 EDT(-0400)] <athena> are we using spaces instead of tabs in our java files as well?
[15:01:25 EDT(-0400)] <EricDalquist> yes
[15:01:26 EDT(-0400)] <athena> i'm not sure my eclipse workspace is set up right
[15:01:39 EDT(-0400)] <athena> i'll make sure i get that fixed then if it's not currently that way
[15:02:10 EDT(-0400)] <EricDalquist> oh we should have the jasig parent pom in central later today
[15:02:31 EDT(-0400)] <EricDalquist> and we can probably start publishing the resource server there too, I think it doesn't have any non-central dependencies
[15:02:33 EDT(-0400)] <athena> oh nice!
[15:02:35 EDT(-0400)] <athena> sounds great (smile)
[15:02:43 EDT(-0400)] <EricDalquist> sonatype for maven hosting is so nice
[15:07:43 EDT(-0400)] <athena> yeah it sounds like it should work out great
[15:11:16 EDT(-0400)] <EricDalquist> grr
[15:11:22 EDT(-0400)] <EricDalquist> updated resource server to ehcache 2
[15:11:31 EDT(-0400)] <EricDalquist> found a bug in their page caching filter
[15:11:37 EDT(-0400)] <EricDalquist> it doesn't actually cache the headers any more
[15:12:04 EDT(-0400)] <athena> ohh, eep (sad)
[15:12:14 EDT(-0400)] <EricDalquist> looks like an oversight in some refactoring
[15:12:18 EDT(-0400)] <athena> no good
[15:12:35 EDT(-0400)] <athena> i just discovered that the tokens for the publishing and subscribing owners are the same, interestingly enough
[15:13:26 EDT(-0400)] <athena> so you actually can't tell from looking at a permission which system its from
[15:13:30 EDT(-0400)] <athena> which seems like an error to me
[15:13:35 EDT(-0400)] <EricDalquist> yeah
[15:14:00 EDT(-0400)] <athena> i can fix all this craziness, but we may have to do some extra stuff for data migration to 3.3
[15:14:12 EDT(-0400)] <athena> guess i'm inclined to just change stuff to make sense in the trunk now
[15:14:19 EDT(-0400)] <EricDalquist> thats fine
[15:14:22 EDT(-0400)] <athena> ok
[15:14:23 EDT(-0400)] <EricDalquist> make the change
[15:14:30 EDT(-0400)] <EricDalquist> and we can figure out the data migration tricks
[15:15:37 EDT(-0400)] <athena> you know, i guess actually while i'm changing this stuff there's no reason i couldn't add some extra data to support some of the stuff gary has in this screenshot: http://www.ja-sig.org/wiki/download/attachments/28574411/permissions_manager_main_flat.png
[15:15:47 EDT(-0400)] <athena> if we're making it JPA anyway, i can add a description field and such
[15:15:53 EDT(-0400)] <EricDalquist> yup
[15:16:28 EDT(-0400)] <athena> might want to do the images w/ CSS though, so they could be different in different skins
[15:16:39 EDT(-0400)] <EricDalquist> yeah

  • No labels