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

ยซ Previous Version 36 Next ยป

[00:30:02 EDT(-0400)] * lfuller (~sparhk@ip68-98-56-21.ph.ph.cox.net) has joined ##uportal
[01:06:13 EDT(-0400)] * mkuraja (~mkuraja@ip68-3-107-145.ph.ph.cox.net) has joined ##uportal
[07:51:47 EDT(-0400)] * michelled (~michelled@CPE001310472ade-CM0011aefd3ca8.cpe.net.cable.rogers.com) has joined ##uportal
[08:51:03 EDT(-0400)] * jessm (~Jess@c-71-232-3-151.hsd1.ma.comcast.net) has joined ##uportal
[09:46:07 EDT(-0400)] * jessm (~Jess@c-71-232-3-151.hsd1.ma.comcast.net) has joined ##uportal
[10:28:52 EDT(-0400)] * EricDalquist (~dalquist@2607:f388:e:0:221:9bff:fe37:e768) has joined ##uportal
[10:37:36 EDT(-0400)] * athena (~athena@adsl-99-97-29-187.dsl.wlfrct.sbcglobal.net) has joined ##uportal
[10:43:29 EDT(-0400)] * michelled (~michelled@142.150.154.101) has joined ##uportal
[10:47:36 EDT(-0400)] * holdorph (~holdorph@wsip-72-215-204-133.ph.ph.cox.net) has joined ##uportal
[10:57:07 EDT(-0400)] * lfuller (~sparhk@wsip-72-215-204-133.ph.ph.cox.net) has joined ##uportal
[11:14:14 EDT(-0400)] * awills (~awills@wsip-72-215-204-133.ph.ph.cox.net) has joined ##uportal
[11:41:32 EDT(-0400)] * b-rock (~cdd07bf2@gateway/web/freenode/x-kxxbczkxjgmcbgzb) has joined ##uportal
[11:43:08 EDT(-0400)] <EricDalquist> can anyone here think of a reason it would be bad to include each person's jasig username on this page: http://www.ja-sig.org/wiki/display/LIC/Completed+Contributor+Agreements
[11:43:36 EDT(-0400)] <EricDalquist> I'm thinking it would be handy to have something like this: http://people.apache.org/~jim/committers.html
[11:43:37 EDT(-0400)] <b-rock> Greetings uPortal devs: http://uportal.pastebin.com/pHS5x5WH I've found a login issue from the trace I think it may be missing data in the portal database. Has anyone here seen this trace by chance?
[11:44:01 EDT(-0400)] <EricDalquist> " THE SITE IS IN READ-ONLY MODE AS WE ARE CURRENTLY MOVING HOSTER, SOME PASTES MIGHT BE MISSING FOR A FEW HOURS"
[11:44:10 EDT(-0400)] <EricDalquist> thats all I see at pastebin right now (sad)
[11:44:32 EDT(-0400)] <b-rock> oh. ok. sorry for the interruption. I'll chekc w/ the dba's on site.
[11:44:42 EDT(-0400)] <EricDalquist> not your issue
[11:44:53 EDT(-0400)] <EricDalquist> there may be other pastebin like services out there
[11:45:09 EDT(-0400)] <b-rock> oh ok. I' can put it on another one and post. just a moment...
[11:46:14 EDT(-0400)] <b-rock> here on http://pastie.org/967806 ..
[11:46:56 EDT(-0400)] <b-rock> this is only in development and not production server.
[11:47:03 EDT(-0400)] <EricDalquist> hrm
[11:47:10 EDT(-0400)] <EricDalquist> there have been a few reports in the past of a missing guest user
[11:47:18 EDT(-0400)] <EricDalquist> not sure if anyone has figured out why it has gone missing
[11:47:33 EDT(-0400)] <EricDalquist> and the fix has been to re-import the guest.user file I believe
[11:47:44 EDT(-0400)] <b-rock> I thought org.jasig.portal.RDBMUserIdentityStore.__getPortalUID(RDBMUserIdentityStore.java:327) may mean database
[11:47:51 EDT(-0400)] <EricDalquist> yeah
[11:48:15 EDT(-0400)] <athena> yeah, sounds like the guest user or something important about it is missing from the db
[11:48:43 EDT(-0400)] <holdorph> EricDalquist (regarding the committer ids) it would probably be worth individually asking each of the existing committers, it's a hard thing to take back, if anyone objects. I would be fine with my id, next to my name.
[11:48:46 EDT(-0400)] <b-rock> ok. Thanks EricDalquist. Its possible our dba group has modified something in the dev db
[11:51:14 EDT(-0400)] <awills> EricDalquist i think it would be fine... the names themselves are already public, only the link to real name is new... and 90%+ are inferable
[11:51:30 EDT(-0400)] <awills> *the SVN names are already public
[11:51:51 EDT(-0400)] <EricDalquist> right and since we have the People Directory enabled in confluence
[11:52:05 EDT(-0400)] <athena> i have no problem with having my commiter username posted
[11:52:10 EDT(-0400)] <EricDalquist> that makes your confluence displayname->usename public
[11:52:30 EDT(-0400)] <EricDalquist> and once we get this migration to contegix done we'll finally have people on a single username for all services
[11:52:38 EDT(-0400)] <holdorph> i doubt anyone would object
[11:52:44 EDT(-0400)] <awills> i also have no issue with that... in fact i took the time to make that link explicitly on Ohloh
[11:52:49 EDT(-0400)] * athena did too
[11:52:53 EDT(-0400)] <holdorph> but if someone did, it's a real PITA to eliminate all the sites that could have 'captured' that data
[11:53:08 EDT(-0400)] <EricDalquist> UHG ... crowd only lets me add users to a group by email or username
[11:53:10 EDT(-0400)] <athena> it seems like it'd be hard to expect anonymity when contributing to an open source project
[11:53:11 EDT(-0400)] <EricDalquist> ugly
[11:53:11 EDT(-0400)] <holdorph> like google, etc.
[11:53:17 EDT(-0400)] <athena> unless you use an alias on email lists and such or something
[12:35:58 EDT(-0400)] <athena> heh
[12:36:18 EDT(-0400)] <athena> so i figured i'd need to add logic so that DLM fragment admins could delete portlets that were marked unremovable
[12:36:32 EDT(-0400)] <athena> turns out you actually can't do that w/ the DLM user preferences channel
[12:36:36 EDT(-0400)] <athena> have to mark it deletable first
[12:41:18 EDT(-0400)] <EricDalquist> yup
[12:41:20 EDT(-0400)] <EricDalquist> which is a PITA
[12:41:24 EDT(-0400)] <athena> yeah
[12:41:36 EDT(-0400)] <athena> course, i suspect that's an issue on the back-end, not just the front
[12:41:42 EDT(-0400)] <EricDalquist> back on in a little bit
[12:41:53 EDT(-0400)] <athena> though maybe at least if the permissions are easy to edit and don't require a refresh it won't be such a pain
[13:19:58 EDT(-0400)] * Sememmon (~Sememmon@unaffiliated/sememmon) has joined ##uportal
[13:47:13 EDT(-0400)] * michelled_ (~michelled@142.150.154.101) has joined ##uportal
[14:03:09 EDT(-0400)] * EricDalquist (~apollo@static-210-59.vpn.wisc.edu) has joined ##uportal
[14:12:36 EDT(-0400)] <athena> EricDalquist: so spring 3 next week or so, you said?
[14:13:00 EDT(-0400)] <EricDalquist> crosses fingers
[14:13:00 EDT(-0400)] <EricDalquist> yes
[14:13:05 EDT(-0400)] <EricDalquist> nick and I are on furlough on Friday
[14:13:10 EDT(-0400)] <EricDalquist> so we're spending all day on pluto 2
[14:13:11 EDT(-0400)] <athena> ah right
[14:13:13 EDT(-0400)] <athena> yeah (smile)
[14:13:15 EDT(-0400)] <athena> awesome!
[14:13:19 EDT(-0400)] <EricDalquist> so right now
[14:13:24 EDT(-0400)] <athena> trying to decide what to do w/ the DLM fragment stuff
[14:13:25 EDT(-0400)] <EricDalquist> the pluto 2 branch renders portlets
[14:13:43 EDT(-0400)] <EricDalquist> one big bug I need to fix is for portlets request.contextPath always returns /uPortal
[14:13:46 EDT(-0400)] <athena> might add stuff to UpdatePreferencesServlet for now, since i'll have to work through the DLM code even w/ the REST change
[14:13:54 EDT(-0400)] <athena> ah (smile)
[14:13:55 EDT(-0400)] <EricDalquist> and Nick is going to work on merging the URL branch into the pluto 2 branch
[14:13:59 EDT(-0400)] <EricDalquist> which we'll then merge into trunk
[14:14:00 EDT(-0400)] <athena> awesome (smile)
[14:14:02 EDT(-0400)] <EricDalquist> lots of merging fun
[14:14:03 EDT(-0400)] <athena> that sounds great!
[14:14:18 EDT(-0400)] <athena> yes lol
[14:14:21 EDT(-0400)] <athena> sounds great though!
[14:14:59 EDT(-0400)] <EricDalquist> but that means the end result is pluto 2 + the new URL syntax
[14:17:35 EDT(-0400)] <athena> yeah
[14:17:38 EDT(-0400)] <athena> really exciting!
[14:17:42 EDT(-0400)] <athena> that's this friday?
[14:17:54 EDT(-0400)] <EricDalquist> well thats when the work is happening
[14:18:01 EDT(-0400)] <EricDalquist> the question is do we get far enough to merge back
[14:18:02 EDT(-0400)] <athena> gotcha
[14:18:04 EDT(-0400)] <athena> yeah
[14:18:06 EDT(-0400)] <EricDalquist> keep in mind when we do that merge
[14:18:09 EDT(-0400)] <EricDalquist> all channels stop working
[14:18:12 EDT(-0400)] <athena> yep
[14:18:14 EDT(-0400)] <EricDalquist> and there is no error portlet yet
[14:18:17 EDT(-0400)] <athena> haha
[14:18:17 EDT(-0400)] <EricDalquist> so you just get an error in the logs
[14:18:20 EDT(-0400)] <athena> yeah
[14:18:21 EDT(-0400)] <EricDalquist> and an empty portlet space
[14:18:22 EDT(-0400)] <athena> gotcha
[14:18:24 EDT(-0400)] <athena> (smile)
[14:18:32 EDT(-0400)] <EricDalquist> and probably lots of rough edges
[14:18:32 EDT(-0400)] <athena> better than crashing the whole portal at least though (smile)
[14:18:35 EDT(-0400)] <athena> yeah
[14:18:36 EDT(-0400)] <athena> that's ok
[14:20:17 EDT(-0400)] <EricDalquist> yeah
[14:20:26 EDT(-0400)] <EricDalquist> the new rendering code already is better than the old
[14:20:46 EDT(-0400)] <EricDalquist> if a portlet times out it does a thread.interrupt to try and free up the thread
[14:20:57 EDT(-0400)] <EricDalquist> so hopefully that means less stuff hanging on stuck UI
[14:20:59 EDT(-0400)] <EricDalquist> er IO
[14:23:01 EDT(-0400)] * michelled (~michelled@2002:8e01:9964:b:21e:52ff:fe75:8997) has joined ##uportal
[14:29:02 EDT(-0400)] <awills> nice
[14:30:58 EDT(-0400)] <EricDalquist> also I think I have code in progress that will log the stack trace of the thread when it times out
[14:31:19 EDT(-0400)] <EricDalquist> and the last thing to do is a proxy for all portlet APIs that upon timeout any portlet API call throws a PortletTimedoutException
[14:41:42 EDT(-0400)] <athena> oh nice
[14:41:45 EDT(-0400)] <athena> speaking of such things
[14:42:03 EDT(-0400)] <athena> it looks like a lot of the HttpSession errors people are reporting are the result of AJAX calls taking more than 500 ms
[14:44:08 EDT(-0400)] <EricDalquist> 500ms?
[14:44:21 EDT(-0400)] <EricDalquist> ah the default timeout
[14:44:22 EDT(-0400)] <EricDalquist> yeah
[14:44:34 EDT(-0400)] <EricDalquist> there are A LOT of weird errors you get when interacting with a timed out request
[14:44:42 EDT(-0400)] <EricDalquist> which is the goal of adding that proxy with timeout checking
[14:44:54 EDT(-0400)] <athena> yeah
[14:44:56 EDT(-0400)] <athena> sounds good to me
[14:45:06 EDT(-0400)] <athena> woudl be great to get a more intuitive error message for that
[14:45:18 EDT(-0400)] <athena> course, i bet our ajax portlets will function a lot better without 302 redirects
[14:45:18 EDT(-0400)] <EricDalquist> yeah
[14:45:22 EDT(-0400)] <athena> kinda slow
[14:45:22 EDT(-0400)] <EricDalquist> yes
[14:45:26 EDT(-0400)] <athena> (smile)
[14:45:27 EDT(-0400)] <EricDalquist> resource URLs will be good
[14:45:30 EDT(-0400)] <athena> yes
[14:45:36 EDT(-0400)] <EricDalquist> another thing 'd like to do is have per request type timeouts
[14:45:45 EDT(-0400)] <EricDalquist> like being able to specify render, action and resource timeouts
[14:45:59 EDT(-0400)] <EricDalquist> since you may want to allow render and resource URLs to take a 'long time'
[14:46:03 EDT(-0400)] <EricDalquist> but actions should be short
[14:46:06 EDT(-0400)] <EricDalquist> er renders
[14:46:13 EDT(-0400)] <athena> yeah
[14:46:16 EDT(-0400)] <athena> makes sense to me!
[14:47:53 EDT(-0400)] <athena> well, i've got dialogs working for editing DLM permissions on portlets and tabs
[14:48:02 EDT(-0400)] <athena> need to do the same for columns
[14:48:15 EDT(-0400)] <athena> and then add code to persist the permissions, which shouldn't be too bad
[14:48:26 EDT(-0400)] <EricDalquist> nice
[14:48:42 EDT(-0400)] <athena> so basically what i decided to do is to add a link to the bottom of the tab flyout
[14:48:48 EDT(-0400)] <athena> and an icon to the portlet toolbar
[14:49:04 EDT(-0400)] <athena> both of which launch appropriate permission editng modal dialogs
[14:49:18 EDT(-0400)] <athena> probably not perfect, but i don' tthink it's too terrible, and we can always re-examing the UI later
[14:49:36 EDT(-0400)] <EricDalquist> yeah
[14:49:39 EDT(-0400)] <EricDalquist> functional for now
[14:49:47 EDT(-0400)] <EricDalquist> and we can get the UI folks to do some actual review
[14:50:04 EDT(-0400)] <athena> yep
[14:50:07 EDT(-0400)] <athena> progress
[14:50:09 EDT(-0400)] <EricDalquist> yes
[15:14:46 EDT(-0400)] <EricDalquist> crowed 1.3's group editing is PAINFUL
[15:14:49 EDT(-0400)] <EricDalquist> you have to find the user
[15:14:54 EDT(-0400)] <EricDalquist> switch to their groups tab
[15:14:58 EDT(-0400)] <EricDalquist> then add the group
[15:15:06 EDT(-0400)] <EricDalquist> adding 50+ users to one group this way is no fun
[15:15:13 EDT(-0400)] <EricDalquist> and you can't search on full name
[15:15:17 EDT(-0400)] <EricDalquist> just username or email
[15:15:18 EDT(-0400)] <EricDalquist> arg
[15:24:00 EDT(-0400)] <athena> eeep
[15:24:02 EDT(-0400)] <athena> no good
[15:28:54 EDT(-0400)] * michelled (~michelled@user152-189.wireless.utoronto.ca) has joined ##uportal
[15:47:25 EDT(-0400)] <athena> i'm going to be so glad not to have to look at this UpdatePreferencesServlet monstrosity anymore.
[15:47:32 EDT(-0400)] <EricDalquist> (smile)
[15:47:33 EDT(-0400)] <athena> such a pile of spagghetti
[15:47:35 EDT(-0400)] <EricDalquist> yeah
[15:47:40 EDT(-0400)] <EricDalquist> so much of that old code is
[15:47:43 EDT(-0400)] <athena> it really grew into an evil monster (tongue)
[15:48:48 EDT(-0400)] <athena> we'll have to think about what we want the rest services to be like
[15:49:00 EDT(-0400)] <EricDalquist> yeah
[15:49:07 EDT(-0400)] <EricDalquist> we need to start a wiki page to spec them out
[15:49:14 EDT(-0400)] <EricDalquist> also data transfer formats
[15:49:19 EDT(-0400)] <athena> yeah
[15:49:24 EDT(-0400)] <EricDalquist> for example Nexus uses Json going into the server and XML coming out
[15:49:26 EDT(-0400)] <athena> not going to be trivial exactly
[15:49:29 EDT(-0400)] <athena> but it's all good stuff
[15:49:35 EDT(-0400)] <athena> huh, interesting
[15:49:40 EDT(-0400)] <EricDalquist> yeah
[15:49:41 EDT(-0400)] <EricDalquist> not sure why
[15:49:46 EDT(-0400)] <athena> any idea why?
[15:49:47 EDT(-0400)] <athena> ah, ok
[15:50:00 EDT(-0400)] <athena> seems like a pretty weird decision
[15:50:04 EDT(-0400)] <athena> i think we'd generally want JSON out
[15:50:08 EDT(-0400)] <EricDalquist> yeah
[15:50:09 EDT(-0400)] <athena> though it'd be nice to get to pick
[15:50:18 EDT(-0400)] <EricDalquist> well in theory you could
[15:50:22 EDT(-0400)] <athena> eyah
[15:50:23 EDT(-0400)] <EricDalquist> with the spring 3 stuff
[15:50:27 EDT(-0400)] <EricDalquist> just switch the view
[15:50:28 EDT(-0400)] <athena> yep (smile)
[15:50:35 EDT(-0400)] <EricDalquist> er view renderer
[15:50:56 EDT(-0400)] <athena> ugh i hate trying to track down code in the old channels
[15:51:07 EDT(-0400)] <athena> trying to figure out how it's saving DLM permissions
[15:52:01 EDT(-0400)] <EricDalquist> yeah
[15:52:05 EDT(-0400)] <EricDalquist> some of that stuff seems so complex
[15:52:15 EDT(-0400)] <athena> yeah
[15:52:22 EDT(-0400)] <athena> luckily i think this is pretty straightforward
[15:52:36 EDT(-0400)] <athena> just finding the code to confirm my suspicions that's hard (tongue)
[15:53:17 EDT(-0400)] <EricDalquist> (smile)
[15:53:56 EDT(-0400)] <athena> oh cool
[15:54:00 EDT(-0400)] <athena> it really is just what i thought
[15:54:05 EDT(-0400)] <athena> node.setMoveAllowed(true);
[15:54:05 EDT(-0400)] <athena> (smile)
[15:54:09 EDT(-0400)] <EricDalquist> nice
[15:54:13 EDT(-0400)] <athena> easy enough
[15:56:47 EDT(-0400)] <athena> i'm guessing we may want a number of REST services to replace this one servlet
[15:56:48 EDT(-0400)] <athena> we'll see
[15:57:06 EDT(-0400)] <athena> but things like "update the permissions for a layout note" seem like a reasonable thing to have be a REST call
[15:57:18 EDT(-0400)] <EricDalquist> yeah
[15:58:03 EDT(-0400)]

<EricDalquist> issue a POST to /layouts/

Unknown macro: {id}

/nodes/


[15:58:03 EDT(-0400)] <athena> this pile of methods is sort of without logic or tests
[15:58:07 EDT(-0400)] <athena> yeah (smile)
[15:58:14 EDT(-0400)] <athena> seems pretty reasonable
[15:58:48 EDT(-0400)] <athena> though dunno if we really want to require the layout id
[15:59:19 EDT(-0400)] <EricDalquist> what if the user has multiple layouts?
[15:59:39 EDT(-0400)] <athena> yeah, we have to handle that
[15:59:44 EDT(-0400)] <athena> the current code doens't really know the id though
[16:00:14 EDT(-0400)] <athena> i think it gets stuff based on the request? not sure
[16:01:12 EDT(-0400)] <EricDalquist> ah yeah
[16:01:53 EDT(-0400)] <athena> doing it based on request seems reasonable
[16:02:12 EDT(-0400)] <athena> though there might be some sort of use case where we'd build an admin use case around targeting a specific layout
[16:02:18 EDT(-0400)] <athena> would want some permissions checks for that one
[16:02:33 EDT(-0400)] <EricDalquist> right
[16:08:38 EDT(-0400)] <athena> so what oustanding importan channels do we have?
[16:08:41 EDT(-0400)] <athena> DLM Fragments
[16:08:43 EDT(-0400)] <athena> Error
[16:08:44 EDT(-0400)] <athena> CUser
[16:08:49 EDT(-0400)] <athena> anything else?
[16:09:22 EDT(-0400)] <EricDalquist> sounds about right
[16:09:34 EDT(-0400)] <EricDalquist> error is going to need a lot of รขย€ยฆ thought

  • No labels