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 32 Next »

[08:32:52 CDT(-0500)] * agherna (n=argherna@singularity.ci.uiuc.edu) has joined ##uportal
[08:32:59 CDT(-0500)] * agherna (n=argherna@singularity.ci.uiuc.edu) has left ##uportal
[08:34:08 CDT(-0500)] * EricDalquist (n=dalquist@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[08:57:03 CDT(-0500)] * ChanServ sets mode +o JASIGLogBot
[08:59:00 CDT(-0500)] * bszabo (n=bszabo@uni1.unicon.net) has joined ##uportal
[09:18:01 CDT(-0500)] * agherna (n=argherna@singularity.ci.uiuc.edu) has joined ##uportal
[10:07:59 CDT(-0500)] * dmccallum (n=dmccallu@uni1.unicon.net) has joined ##uportal
[10:08:09 CDT(-0500)] <dmccallum> EricDalquist, i hear you're looking for candidate issues to work on for the near-term uP3 release.
[10:08:16 CDT(-0500)] <EricDalquist> yeah
[10:08:20 CDT(-0500)] <EricDalquist> getting to that point
[10:08:35 CDT(-0500)] <EricDalquist> anything that is bothering you guys that you'd like some help on?
[10:08:47 CDT(-0500)] <dmccallum> we lost Allen Haws back to a Cisco project, so he had to leave some things midstream that you might find interesting
[10:09:19 CDT(-0500)] <EricDalquist> ok, what are they?
[10:09:22 CDT(-0500)] <dmccallum> specifically, he was working on portlet mode controls and had made some progress, but had run afoul of some caching problems, i believe
[10:09:54 CDT(-0500)] <dmccallum> i think the closest issue in the ja-sig jira is UPT-191, which doesn't provide much detail....
[10:10:22 CDT(-0500)] <EricDalquist> (smile) that is a bit short on info
[10:10:40 CDT(-0500)] <dmccallum> there's not much detail in our jira either. but, I can either email or check in the code he had cooked up
[10:11:01 CDT(-0500)] <EricDalquist> can you email me a patch off of HEAD of his work?
[10:11:12 CDT(-0500)] <dmccallum> can do
[10:12:15 CDT(-0500)] <dmccallum> the other two candidates i saw were UPT-67, which I don't think bszabo is going to have time to finish up this month, and UPT-193, which was another Allen Haws item
[10:13:13 CDT(-0500)] <EricDalquist> ok
[10:13:22 CDT(-0500)] <dmccallum> I'm also probably not going to get to UPT-99 by the end of the month, and I've run into some issues with WSRP on Oracle that I probably won't get to either
[10:13:25 CDT(-0500)] <peterk> upt-191 is just that - provide ability to set rendering attributes for a portlet at publish time
[10:13:40 CDT(-0500)] <EricDalquist> ah ok
[10:13:52 CDT(-0500)] <EricDalquist> so an addition to the portlet manager to provide that UI?
[10:14:05 CDT(-0500)] <peterk> yes, an optional screen
[10:14:09 CDT(-0500)] <EricDalquist> ok
[10:14:52 CDT(-0500)] <EricDalquist> optional as in not required to have anything filled out?
[10:15:01 CDT(-0500)] <peterk> right
[10:15:04 CDT(-0500)] <EricDalquist> ok
[10:15:46 CDT(-0500)] <peterk> although we may come up with some standard ones (e.g. icon) to fill, or at least have defaults. portlet mode controls that dmccallum mentioned can be implemented this way as well
[10:16:09 CDT(-0500)] <EricDalquist> pl
[10:16:10 CDT(-0500)] <EricDalquist> ok
[10:16:23 CDT(-0500)] <EricDalquist> that shouldn't be too hard once the base UI is there
[10:17:01 CDT(-0500)] <bszabo> Eric, let me know if you look at UPT-67, because I have some incomplete work that fixes some of the test failures.
[10:17:02 CDT(-0500)] <peterk> upt-193 - replace the jar with the standard version
[10:17:23 CDT(-0500)] <EricDalquist> yeah
[10:17:50 CDT(-0500)] <EricDalquist> I've assigned 191 and 193 to me
[10:18:12 CDT(-0500)] <EricDalquist> I'm not enough of a hibernate guy to know how to doUPT-99 (I couldn't figure it out when I did the original mappings)
[10:18:14 CDT(-0500)] <dmccallum> ok... it looks like we were tracking Allen's work on portlet mode controls under an issue that we have mapped to UPT-188, which doesn't say anything at all about portlet mode controls. i vote to open a new UPT issue and attach allen's partial patch
[10:18:27 CDT(-0500)] <EricDalquist> sounds good
[10:18:50 CDT(-0500)] <dmccallum> with that one, UPT-99, is it talking about problems like orphaned portlet preferences?
[10:19:14 CDT(-0500)] <EricDalquist> and orphaned portlet domain objects, say you delete a portlet app def
[10:19:24 CDT(-0500)] <EricDalquist> there is a whole tree of dependent objects that need to be removed
[10:20:07 CDT(-0500)] <dmccallum> ok... i think what's going on there is that cascading deletes don't work if you have defined a uni-directional graph of full-blown entities
[10:20:37 CDT(-0500)] <EricDalquist> well the problem I was having is our domain object structure doesn't have any actual object references
[10:20:54 CDT(-0500)] <EricDalquist> we just list the object IDs
[10:24:20 CDT(-0500)] <dmccallum> fair enough. i hadn't looked at the entire domain object structure, just preferences.
[10:25:02 CDT(-0500)] <EricDalquist> there very well may be a way to get hibernate to do the foreign keys and cascading deletes I just don't know enough about it to figure out how
[10:25:07 CDT(-0500)] <dmccallum> i'm still willing to work on upt-99, but it sounds like the issue might be more complicated than i had been thinking. in any event, i probably won't get to it this month.
[10:25:16 CDT(-0500)] <EricDalquist> ok
[10:25:28 CDT(-0500)] <peterk> if you figure that out, let me know - I have a similar problem in portlet window mapping dao
[10:25:34 CDT(-0500)] <dmccallum> will do
[10:25:38 CDT(-0500)] <EricDalquist> well I'll also add UPT-67 to my plate
[10:26:07 CDT(-0500)] <EricDalquist> I should be able to address all 3 of those next week (at least get the two easy ones done)
[10:26:24 CDT(-0500)] * peterk is glad to have Eric back on uP3 (smile)
[10:26:48 CDT(-0500)] <dmccallum> bszabo, did you check in any patches for UPT-67?
[10:27:17 CDT(-0500)] <EricDalquist> I'm glad to be back on it
[10:27:47 CDT(-0500)] <bszabo> No I have not. I will get that stuff in... Is Monday early enough?
[11:14:34 CDT(-0500)] * JASIGLogBot (n=PircBot@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[12:45:06 CDT(-0500)] * EricDalquist (n=dalquist@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has left ##uportal
[12:45:06 CDT(-0500)] * EricDalquist (n=dalquist@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[12:43:49 CDT(-0500)] <EricDalquist> so what problems have you been having with WSRP?
[12:46:44 CDT(-0500)] <dmccallum> getting a stack trace in the initportal target
[12:46:51 CDT(-0500)] <EricDalquist> ah yeah
[12:46:54 CDT(-0500)] <EricDalquist> I think andrew fixed that
[12:47:05 CDT(-0500)] <dmccallum> this is a different issue, i believe
[12:47:10 CDT(-0500)] <EricDalquist> hrm
[12:47:20 CDT(-0500)] <dmccallum> have not looked closely yet, but i think it's oracle lob problems
[12:47:24 CDT(-0500)] <EricDalquist> ah
[12:47:29 CDT(-0500)] <EricDalquist> The WSRP persistence is something I was hoping to defer
[12:47:41 CDT(-0500)] <EricDalquist> but perhaps I should really look at dumping jdom
[12:47:50 CDT(-0500)] <EricDalquist> and using the latest WSRP4j snapshot
[12:48:53 CDT(-0500)] <dmccallum> i think i've got the lob problem solved for DbLoader, but jpox is a whole 'nuther can of worms
[12:49:14 CDT(-0500)] <dmccallum> want me to email you the trace?
[12:49:22 CDT(-0500)] <EricDalquist> sure
[12:50:04 CDT(-0500)] <EricDalquist> \join ##jasig-bot-test
[12:50:21 CDT(-0500)] <EricDalquist> friggin \ vs / always messes me up on irc
[12:54:46 CDT(-0500)] <peterk> Michael I. said he'll finish up WSRP hibernate dao impl soon
[12:54:57 CDT(-0500)] <EricDalquist> ok
[12:55:11 CDT(-0500)] <peterk> not sure what soon means, but perhaps we can defer working on it
[12:55:51 CDT(-0500)] <peterk> that said, we might as well turn of jdom and consider wsrp broken until new daos are in
[12:55:59 CDT(-0500)] <peterk> ^of^off
[12:56:31 CDT(-0500)] <EricDalquist> true
[12:56:41 CDT(-0500)] <EricDalquist> it may be better to just pull all that out
[12:56:55 CDT(-0500)] <EricDalquist> so we don't have to use a hacked version of xalan
[12:58:40 CDT(-0500)] <dmccallum> orm can be complicated stuff.... i think there's something to be said for standardizing on a single approach within a given project so you dont end up with little islands of exclusive ownership.
[12:58:54 CDT(-0500)] <EricDalquist> yup
[13:02:33 CDT(-0500)] <peterk> well, hopefully that's what we'll have after WSRP daos are migrated
[13:04:49 CDT(-0500)] <peterk> but what would also help if we new hibernate well enough to do all the things we need (smile)
[13:04:59 CDT(-0500)] <EricDalquist> (smile)
[15:15:23 CDT(-0500)] <EricDalquist> is there any other work being done on the portlet manager right now?
[15:15:52 CDT(-0500)] <bszabo> We are planning on integrating the Permissions Manager into it
[15:15:58 CDT(-0500)] <EricDalquist> ok
[15:16:08 CDT(-0500)] <EricDalquist> but no converting it to spring portlet MVC right now?
[15:17:00 CDT(-0500)] <bszabo> Not at the moment (current Sprint)
[15:17:44 CDT(-0500)] <EricDalquist> ok
[15:18:27 CDT(-0500)] <bszabo> Are you thinking about working on that?
[15:18:47 CDT(-0500)] <EricDalquist> not right now, I'm starting on adding rendering attributes to portlets at publish time
[15:19:00 CDT(-0500)] <bszabo> Ah.
[15:19:04 CDT(-0500)] <EricDalquist> but I didn't want to make changes to the portlet manager if someone else was making big changes to the same code
[15:20:37 CDT(-0500)] <dmccallum> we should be able to work on rendering attributes and the perm mgr delegation at the same time, correct?
[15:21:29 CDT(-0500)] <EricDalquist> yes
[15:24:12 CDT(-0500)] <EricDalquist> peterk ... how do you see these rendering parameters being used by the rest of the framework. Will they be accessed by something like IRenderingAttributeProvider ?
[15:26:03 CDT(-0500)] <peterk> Eric: it's the other way around - portlet window manager will have to figure out what providers are available (configured) and set the values for them. The rendering pipeline is then configured with these providers (in uP2 TransformerImpl initates the incorporation of theme and structure- scoped attributes)
[15:26:15 CDT(-0500)] <EricDalquist> ah ok
[15:26:18 CDT(-0500)] <EricDalquist> that makes more sence
[15:26:32 CDT(-0500)] * JASIGLogBot (n=PircBot@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[15:26:32 CDT(-0500)] * Topic is 'go uP3 go' set by EricDalquist on 10:24:54 CDT(-0500)
[15:27:26 CDT(-0500)] <peterk> the generalization is that attributes can be associated not only with theme and structure transform, but with other things, for instance portlet definition.
[15:28:09 CDT(-0500)] <peterk> or with the layout
[15:28:14 CDT(-0500)] <EricDalquist> ah ok
[15:30:36 CDT(-0500)] <EricDalquist> is there an IRenderingAttributeProvider registry or something that I need to use to get the available IRenderingAttributeProviders?
[15:30:39 CDT(-0500)] <peterk> so, for example, if you want to define an attribute such as "folderType" that should transcend all structure and theme stylesheet choices, you can define a rendering attribute provider that will associate attributes based on layout node id and will always (no matter what stylesheet is selected) inject them at an early stage in the rendering pipeline.
[15:30:57 CDT(-0500)] <EricDalquist> ok
[15:31:00 CDT(-0500)] <peterk> there isn't a registry ... we'll have to figure this out
[15:31:09 CDT(-0500)] <peterk> they are not dynamic objects - they are configured in spring
[15:31:18 CDT(-0500)] <EricDalquist> ok
[15:31:52 CDT(-0500)] <peterk> I guess it's fair to assume that you'd be defining attributes only for the current portal context
[15:32:43 CDT(-0500)] <EricDalquist> so they are configured explicity in spring. When the pipeline is executing getAttributeValues will be called and the provider will have to return the appropriate value(s)?
[15:34:14 CDT(-0500)] <peterk> the attributes are incorporated using injectors (there are alternative implementations, depending on how you want the syntax to look like) ... injectors can be either placed in the rendering pipeline directly (as SAX filters) or, if you're doing a transformation, configured as a property of the TransformerImpl
[15:34:32 CDT(-0500)] <peterk> btw, this is in the architecture docs (wink)
[15:34:43 CDT(-0500)] <EricDalquist> ah ok ... I'll go read (smile)
[15:35:04 CDT(-0500)] <peterk> that'd be good, because then we could improve them (smile)
[15:38:03 CDT(-0500)] <peterk> also, you shouldn't need much in terms of other rendering attribute providers in uP2 context. Perhaps one that keeps track of portlet-definition-associated attributes (so we can define some things at a publish time).
[15:39:31 CDT(-0500)] * dmccallum (n=dmccallu@uni1.unicon.net) has joined ##uportal
[16:25:04 CDT(-0500)] <EricDalquist> I've got a SVN question unrelated to uP3 if anyone has a moment
[16:25:18 CDT(-0500)] <bszabo> Shoot
[16:25:53 CDT(-0500)] <EricDalquist> So the Atlassian folks created a SVN space for me to put this IRC bot: https://svn.atlassian.com/svn/public/contrib/confluence/ircbot/trunk/
[16:26:09 CDT(-0500)] <EricDalquist> I have the command line client installed ... how do I do the initial import?
[16:26:27 CDT(-0500)] <EricDalquist> the eclipse client will only let me do an import if the directory I specify doesn't already exist
[16:29:28 CDT(-0500)] <bszabo> You are going to want to import the contents of the ircbot root directory into trunk, so 'svn import /path/to/ircbot_rootdir https://svn.atlassian.com/svn/public/contrib/confluence/ircbot/trunk/ '
[16:29:49 CDT(-0500)] <EricDalquist> ok thats what the docs looked like but I wanted to be sure
[16:29:58 CDT(-0500)] <bszabo> I am pretty sure this will import all of the contents under ircbot_rootdir into /trunk
[16:30:05 CDT(-0500)] <EricDalquist> no fun messing up my first svn commit in a nice public repository (smile)
[16:30:31 CDT(-0500)] <bszabo> You may want to test it first
[16:30:40 CDT(-0500)] <EricDalquist> is there a way to do that?
[16:30:47 CDT(-0500)] <bszabo> Let me double check, so you don't blame me (smile)
[16:30:55 CDT(-0500)] <EricDalquist> lol ok
[16:30:57 CDT(-0500)] <bszabo> I will just test the syntax on a local repo
[16:32:11 CDT(-0500)] <EricDalquist> ok, thanks
[16:32:50 CDT(-0500)] <bszabo> I am pretty sure that is correct, but I have messed up and imported the wrong level before...
[16:36:40 CDT(-0500)] <EricDalquist> so go ahead with that command?
[16:37:58 CDT(-0500)] <bszabo> Yes. The contents of the local directory you specify will be imported, not the directory itself.
[16:38:08 CDT(-0500)] <EricDalquist> cool thanks
[16:38:18 CDT(-0500)] <bszabo> np
[16:41:16 CDT(-0500)] <shawnlonas> Who knows something about SerializingUserLayoutDao.java?
[16:42:06 CDT(-0500)] * JASIGLogBot (n=PircBot@adsl-71-150-245-185.dsl.mdsnwi.sbcglobal.net) has joined ##uportal
[16:42:06 CDT(-0500)] * Topic is 'go uP3 go' set by EricDalquist on 10:24:54 CDT(-0500)
[16:42:10 CDT(-0500)] <shawnlonas> peterk?
[16:42:32 CDT(-0500)] <shawnlonas> Eric?
[16:42:37 CDT(-0500)] <EricDalquist> let me look at it
[16:43:37 CDT(-0500)] <EricDalquist> I think it was just the quick and dirty layout DAO that just serialized the layout to an XML string and persisted the whole layout as that string
[16:44:02 CDT(-0500)] <peterk> (smile)
[16:44:18 CDT(-0500)] <peterk> why's it a pain ?
[16:44:31 CDT(-0500)] <shawnlonas> Yes, it seems that when I change it to point to the new location for the DTD the portal throws an error
[16:44:51 CDT(-0500)] <peterk> what did you change, and what error do you get?
[16:45:24 CDT(-0500)] <shawnlonas> because it then tries to look for the value layout XML stored in the DB which happens to be http://www.ja-sig.org/products/uportal/dtds/up3userlayout.dtd
[16:45:51 CDT(-0500)] <shawnlonas> I changed line #39 and line #70
[16:45:56 CDT(-0500)] <peterk> right, you have to change the XML in the properties/data.xml - it's referencing the old dtd location
[16:46:00 CDT(-0500)] <shawnlonas> to point to the new location
[16:46:22 CDT(-0500)] <shawnlonas> I suppose I need to reload the DB then?
[16:46:39 CDT(-0500)] <EricDalquist> or you could go in with some DB tool and just edit the XML by hand
[16:46:52 CDT(-0500)] <peterk> yes. you're changing the hibernate version of the SerializingUserLayoutDao right?
[16:47:01 CDT(-0500)] <shawnlonas> yes
[16:47:47 CDT(-0500)] <peterk> btw. line 70 means that the local file (in the dtds dir) needs to be renamed to have the same as the string you're passing on line 70
[16:48:11 CDT(-0500)] <peterk> Eric: but then it will break upon commit (smile)
[16:48:35 CDT(-0500)] <shawnlonas> Right. I have the local file with that name but it is still breaking.
[16:49:12 CDT(-0500)] <peterk> well, fix these data.xml occurrences (there should be two I think), run ant db and if you get an error, show us the strack trace
[16:49:30 CDT(-0500)] <shawnlonas> OK. thx
[16:53:15 CDT(-0500)] <EricDalquist> bszabo Yay that command worked!
[16:54:34 CDT(-0500)] <bszabo> Good stuff
[16:58:30 CDT(-0500)] * agherna (n=argherna@singularity.ci.uiuc.edu) has left ##uportal
[17:39:28 CDT(-0500)] * JASIGLogBot (n=PircBot@mammoth.doit.wisc.edu) has joined ##uportal
[17:39:29 CDT(-0500)] * Topic is 'go uP3 go' set by EricDalquist on 2006-09-19 10:24:54 CDT(-0500)
[19:02:41 CDT(-0500)] * bszabo (n=bszabo@uni1.unicon.net) has left ##uportal
[19:10:08 CDT(-0500)] * shawnlonas (n=shawn@uni1.unicon.net) has left ##uportal
[20:04:04 CDT(-0500)] * andrewpetro (n=microcli@uni1.unicon.net) has joined ##uportal

  • No labels