uPortal IRC Logs-2011-07-19

[08:44:20 CDT(-0500)] <Arvids> Eric, are you here?
[08:45:14 CDT(-0500)] <EricDalquist> yes
[08:45:21 CDT(-0500)] <Arvids> i'm looking at different *_locale and *_mdata tables and it seems that there's a big mess
[08:45:29 CDT(-0500)] <EricDalquist> yes
[08:45:49 CDT(-0500)] <EricDalquist> pretty much the only one thats semi-sane is the one for portlet definitions in 4
[08:45:58 CDT(-0500)] <EricDalquist> the others are still part of the legacy DAO structure
[08:46:05 CDT(-0500)] <Arvids> ok
[08:46:20 CDT(-0500)] <Arvids> what can be done in order to improve the situation?
[08:46:23 CDT(-0500)] <EricDalquist> and if you've tried to follow that old JDBC code … well its gross
[08:46:31 CDT(-0500)] <Arvids> indeed
[08:46:36 CDT(-0500)] <Arvids> i'm currently investigating it
[08:46:38 CDT(-0500)] <EricDalquist> I'm not sure for the layout releated suff
[08:46:45 CDT(-0500)] <EricDalquist> our longer term plans, probably for 4.1
[08:46:55 CDT(-0500)] <EricDalquist> is to replace that layout storage code completely
[08:46:59 CDT(-0500)] <EricDalquist> and move it to JPA2
[08:47:17 CDT(-0500)] <Arvids> i'd like to resolve UP-3095
[08:47:45 CDT(-0500)] <Arvids> that would give a lot to the quality of i18n
[08:48:12 CDT(-0500)] <Arvids> those little piecies of hardcoded texts (mostly in javascript) are doable, but this one looks like nightmare
[08:48:32 CDT(-0500)] <EricDalquist> pulling up that issue now … internet is slow at home
[08:48:59 CDT(-0500)] <Arvids> it's about translating tabs
[08:49:18 CDT(-0500)] <Arvids> tab names of layout owners must be translatable
[08:50:03 CDT(-0500)] <EricDalquist> do you have any thoughts on what a UI would look like or how the localized tab names would be edited?
[08:50:24 CDT(-0500)] <Arvids> i thing of special portlet for this purpose
[08:50:34 CDT(-0500)] <Arvids> i see no other solution
[08:51:00 CDT(-0500)] <Arvids> it wouldn't be as nice as any other editing, but at least it would work
[08:51:29 CDT(-0500)] <EricDalquist> yeah
[08:51:50 CDT(-0500)] <EricDalquist> and the problem right now is anything that touches that old layout code is pretty much "whatever works"
[08:52:25 CDT(-0500)] <Arvids> i would be glad to contribute such a portlet, but first i need to resolve this issue
[08:52:26 CDT(-0500)] <EricDalquist> yeah so doing a quick search it looks like up_layout_struct_mdata is only referenced in "public void removePortalUID(final int uPortalUID) throws Exception"
[08:52:35 CDT(-0500)] <Arvids> yes
[08:52:44 CDT(-0500)] <Arvids> and no select's
[08:52:57 CDT(-0500)] <EricDalquist> I'm wondering if doing this in the XSL would be easier than touching the layout management code
[08:53:06 CDT(-0500)] <EricDalquist> how flexible do you need the tab naming to be?
[08:53:19 CDT(-0500)] <EricDalquist> like would it be acceptable to put the tab names in the messages file?
[08:54:02 CDT(-0500)] <Arvids> it would be more desirable to put them in DB
[08:54:40 CDT(-0500)] <Arvids> messages file is acceptable, but only as temporary solution
[08:54:41 CDT(-0500)] <EricDalquist> I'm just thinking that a really quick fix would be to use the tab name as a message key
[08:55:02 CDT(-0500)] <EricDalquist> and in the theme XSLT do a message lookup using the tab name as the key
[08:55:12 CDT(-0500)] <Arvids> hmmm... that could work
[08:55:29 CDT(-0500)] <EricDalquist> our current message approach already falls back to just using the key value if no message is found
[08:55:47 CDT(-0500)] <EricDalquist> so that should "just work" for people that don't stick translations the message files
[08:56:12 CDT(-0500)] <Arvids> ok, that's one of the possible solutions..
[08:56:28 CDT(-0500)] <EricDalquist> I'd have to look at the code or check with athena but I bet you could even make the message source a composite
[08:56:32 CDT(-0500)] <EricDalquist> where some come from a flat file
[08:56:37 CDT(-0500)] <EricDalquist> and some come from a db table
[08:56:41 CDT(-0500)] <EricDalquist> and have a portlet to edit that db table
[08:56:56 CDT(-0500)] <Arvids> anyway... i'll try to do it in layout management code tomorrow, but if i won't succeed, then it looks like we'll have to stick to file-based approach
[08:57:03 CDT(-0500)] <EricDalquist> sounds good
[08:58:00 CDT(-0500)] <Arvids> composite solution also sounds good
[08:58:05 CDT(-0500)] <EricDalquist> I just hate doing work in that layout code
[08:58:11 CDT(-0500)] <EricDalquist> its really freaking fragile
[08:58:22 CDT(-0500)] <EricDalquist> and its all going to be replaced in the next 12 months
[08:58:24 CDT(-0500)] <Arvids> today i understand your feelings
[08:59:37 CDT(-0500)] <EricDalquist> yeah that code is just gross primarily because there is no separation between DAO and logic
[09:00:12 CDT(-0500)] <Arvids> btw, looks like i'll need a commit access
[09:00:32 CDT(-0500)] <EricDalquist> sounds good
[09:00:36 CDT(-0500)] <EricDalquist> I'll propose it on the dev list
[09:00:45 CDT(-0500)] <Arvids> ok
[09:02:45 CDT(-0500)] <Arvids> i believe i must talk to Jen about those message_*.propoerties files - he said that he's using a script in order to translate missing keys
[09:03:36 CDT(-0500)] <EricDalquist> yeah, she wrote a maven plugin that finds missing keys and uses google's (soon to be dead) translation service to translate them
[09:03:43 CDT(-0500)] <EricDalquist> with the thought that its better than nothing
[09:04:02 CDT(-0500)] <EricDalquist> and hopefully a starting point for an actual native speaker to finish them
[09:04:39 CDT(-0500)] <Arvids> as i understand that plugin is only available to her (locally)
[09:05:45 CDT(-0500)] <EricDalquist> I believe this is it: https://source.jasig.org/sandbox/maven-translate-plugin/
[09:06:10 CDT(-0500)] <Arvids> ahh... nice
[09:07:04 CDT(-0500)] <EricDalquist> also, what time zone are you in?
[09:07:14 CDT(-0500)] <EricDalquist> just so I can have some idea for planning communications
[09:07:14 CDT(-0500)] <Arvids> +2
[09:07:18 CDT(-0500)] <EricDalquist> thanks
[09:07:30 CDT(-0500)] <Arvids> work time finished right now
[09:07:53 CDT(-0500)] <EricDalquist> yeah, since we're 7 hours different
[09:08:13 CDT(-0500)] <Arvids> ok, i'll try to do something about those tab translations tomorrow
[09:08:16 CDT(-0500)] <EricDalquist> once the vote passes I'll get your svn access setup
[09:08:21 CDT(-0500)] <Arvids> but now i'm going home
[09:08:24 CDT(-0500)] <EricDalquist> it will be the same username/password you use for jira
[09:08:31 CDT(-0500)] <Arvids> ok
[09:08:33 CDT(-0500)] <EricDalquist> sounds good
[09:08:35 CDT(-0500)] <EricDalquist> have a good evening
[11:10:56 CDT(-0500)] <jeffbcross> athena, per your email I'm going to try deploying the Android app from my trunk project
[11:11:50 CDT(-0500)] <athena> ok
[11:11:57 CDT(-0500)] <athena> i think that maybe my android emulator is just horked
[11:12:03 CDT(-0500)] <athena> i can't even delete the app that was there before
[11:12:07 CDT(-0500)] <athena> though maybe i'm just doing it wrong
[11:12:30 CDT(-0500)] <athena> don't think i've really used it since i switched to ti studio
[11:12:34 CDT(-0500)] <jeffbcross> well, TiStudio is also a little quirky
[11:12:42 CDT(-0500)] <athena> yeah, might be i have something misconfigured
[11:12:51 CDT(-0500)] <jeffbcross> but the app deleting problem is odd
[11:14:32 CDT(-0500)] <athena> yeah
[11:14:40 CDT(-0500)] <athena> i mean i assume you just drag it over the trash can right?
[11:14:49 CDT(-0500)] <athena> it got red
[11:14:50 CDT(-0500)] <jeffbcross> no, actually
[11:14:51 CDT(-0500)] <athena> but then it's still there
[11:14:53 CDT(-0500)] <athena> oh.
[11:14:57 CDT(-0500)] <jeffbcross> that only removes it from the home screen
[11:15:01 CDT(-0500)] <athena> oh!
[11:15:04 CDT(-0500)] <athena> ok so how do i delete it?
[11:15:14 CDT(-0500)] <jeffbcross> go to menu > settings > applications
[11:15:20 CDT(-0500)] <athena> ah!
[11:15:22 CDT(-0500)] <jeffbcross> select the application from the list
[11:15:23 CDT(-0500)] <athena> ok, will try that
[11:15:31 CDT(-0500)] <jeffbcross> and then remove
[11:16:19 CDT(-0500)] <jeffbcross> I'm experiencing the same tistudio build issue I had yesterday with quickstart, now with my trunk project.
[11:16:24 CDT(-0500)] <jeffbcross> trying TiDeveloper
[11:16:33 CDT(-0500)] <athena> ah
[11:17:48 CDT(-0500)] <athena> that error looked like it was missing some android build resources
[11:20:17 CDT(-0500)] <jeffbcross> I'm going to experiment with changing Android SDK build targets
[11:21:01 CDT(-0500)] <athena> ok
[11:21:18 CDT(-0500)] <athena> make sure in your run configurations that you're building the project you think you're building, too
[11:21:30 CDT(-0500)] <athena> TI doesn't always seem to switch between projects like i expect it to
[11:21:35 CDT(-0500)] <athena> at least in the studio version
[11:21:58 CDT(-0500)] <jeffbcross> k
[11:22:49 CDT(-0500)] <EricDalquist> athena: Arvids and I were talking about i18n for tab names eariler
[11:23:13 CDT(-0500)] <EricDalquist> he pointed out that up_layout_struct_mdata is apparently never touched now
[11:23:27 CDT(-0500)] <athena> ah
[11:23:32 CDT(-0500)] <EricDalquist> and is trying to figure out a reasonable way to get i18n working for tab names
[11:23:34 CDT(-0500)] <athena> are we using any of the mdata stuff right now?
[11:23:38 CDT(-0500)] <athena> that'd be awesome
[11:23:40 CDT(-0500)] <EricDalquist> only the portlet definition stuff
[11:23:50 CDT(-0500)] <athena> well, glad we're at least doing that - wasn't sure
[11:23:55 CDT(-0500)] <EricDalquist> the quick and dirty idea we had was to do it in the theme xsl
[11:23:55 CDT(-0500)] <athena> guess there's no UI way to add that
[11:24:03 CDT(-0500)] <EricDalquist> use the tab name in the db as a message key
[11:24:07 CDT(-0500)] <athena> ah
[11:24:15 CDT(-0500)] <EricDalquist> since that should fail back to the key value if there is no translation
[11:24:23 CDT(-0500)] <athena> makes sense
[11:24:30 CDT(-0500)] <athena> will probably work most of the time
[11:24:30 CDT(-0500)] <EricDalquist> he's going to take a look at the layout store code
[11:24:45 CDT(-0500)] <EricDalquist> but I'm not hopeful for a reasonable solution there
[11:24:48 CDT(-0500)] <EricDalquist> until we get that code replaced
[11:24:54 CDT(-0500)] <athena> yeah
[11:26:52 CDT(-0500)] <athena> how is arvids actually getting data into the portlet title mdata right now?
[11:26:58 CDT(-0500)] <athena> does our import/export format support that?
[11:30:03 CDT(-0500)] <athena> jeffbcross: ugh, i dont' get any errors that i can see, but it just doesn't install
[11:30:58 CDT(-0500)] <EricDalquist> I don't think it does
[11:31:02 CDT(-0500)] <EricDalquist> but I was going to fix that for the ga
[11:31:11 CDT(-0500)] <athena> ah, cool
[11:31:12 CDT(-0500)] <EricDalquist> since witht he jaxb import its really easy to tweak
[11:31:15 CDT(-0500)] <athena> yeah
[11:31:22 CDT(-0500)] <EricDalquist> we can even just use the existing xml:lang attribute
[11:32:34 CDT(-0500)] <athena> nice
[11:37:58 CDT(-0500)] <jeffbcross> athena I did two things to get it to build. 1, shut down the emulator that TiStudio had spawned (don't think it was the appropriate API level)
[11:38:10 CDT(-0500)] <jeffbcross> 2: commented out the android manifest in tiapp.xml
[11:38:18 CDT(-0500)] <jeffbcross> however, I don't think 2 was the problem
[11:40:28 CDT(-0500)] <jeffbcross> the android manifest was NOT the problem, the emulator was
[11:42:24 CDT(-0500)] <athena> hm, ok
[11:42:33 CDT(-0500)] <athena> so should i launch an emulator manually?
[11:42:52 CDT(-0500)] <jeffbcross> well, in studio, you can set run configurations
[11:43:14 CDT(-0500)] <athena> yes
[11:43:15 CDT(-0500)] <jeffbcross> and the emulator settings for the app should be Google APIs Android 2.2
[11:43:20 CDT(-0500)] <athena> that's what i have
[11:43:22 CDT(-0500)] <jeffbcross> before it was without Google APIs
[11:43:31 CDT(-0500)] <athena> knew that needed to be changed
[11:43:36 CDT(-0500)] <athena> never installs though
[11:43:43 CDT(-0500)] <jeffbcross> hmm
[11:43:54 CDT(-0500)] <athena> and i don't see any errors, though there's so much garbage in the log it's hard to tell
[11:44:01 CDT(-0500)] <athena> no idea why the log level isn't being respected
[11:44:20 CDT(-0500)] <jeffbcross> what's the last bit you see in the log?
[11:44:28 CDT(-0500)] <jeffbcross> does it get to the point of trying to install?
[11:44:39 CDT(-0500)] <athena> i have absolutely no idea
[11:44:51 CDT(-0500)] <athena> my log is full of memory allocation crap
[11:45:10 CDT(-0500)] <athena> going to do a ti studio update, see if that helps at all
[11:45:34 CDT(-0500)] <jeffbcross> might do more harm than good
[11:45:37 CDT(-0500)] <athena> might
[11:45:42 CDT(-0500)] <athena> but it doens't work now, so . . .
[11:45:58 CDT(-0500)] <athena> possible there's a problem w/ my android install too, i suppose
[11:46:08 CDT(-0500)] <jeffbcross> well, if you launch DDMS from the android sdk tools folder, that will show some detailed logging
[11:46:23 CDT(-0500)] <jeffbcross> but that will only show good info IF the problem is to do with the emulator
[11:48:21 CDT(-0500)] <athena> http://pastebin.com/0XwUJn3A
[11:49:25 CDT(-0500)] <athena> think i'm going to just reinstall the android stuff
[11:49:39 CDT(-0500)] <athena> Unable to start service Intent sounds bad
[11:49:42 CDT(-0500)] <jeffbcross> k, looking at it now…first glance I don't see any Titanium activity
[11:55:57 CDT(-0500)] <jeffbcross> once I set the run configuration to use APIs instead of standard SDK, TiStudio launched emulator and deployed fine
[11:56:09 CDT(-0500)] <jeffbcross> using TI SDK 1.6.2
[11:56:21 CDT(-0500)] <jeffbcross> so, I agree it might be something with your Android SDK, athena
[11:56:42 CDT(-0500)] <athena> could have gotten corrupted in that hd crash
[11:57:41 CDT(-0500)] <jeffbcross> ahh, the "geniuses" at the genius bar must've sabotaged it
[11:57:52 CDT(-0500)] <athena> clearly
[11:57:58 CDT(-0500)] <jeffbcross> "iOS SDK is sufficient"
[11:58:26 CDT(-0500)] <athena> lol
[11:58:31 CDT(-0500)] <athena> actually that broke too
[11:58:35 CDT(-0500)] <athena> remember that file you had to send me?
[11:58:41 CDT(-0500)] <jeffbcross> oh yeah
[11:59:10 CDT(-0500)] <jeffbcross> do you know if there's a way to package the run configuration with the quickstart?
[11:59:42 CDT(-0500)] <jeffbcross> not sure if that's an eclipse thing, or a TiStudio thing
[12:00:18 CDT(-0500)] <athena> dunno
[12:00:23 CDT(-0500)] <athena> probably wouldn't try it
[12:00:32 CDT(-0500)] <jeffbcross> looks like it's possible
[12:01:07 CDT(-0500)] <athena> nice
[12:02:25 CDT(-0500)] <jeffbcross> why wouldn't you try it?
[12:02:40 CDT(-0500)] <athena> oh i assumed it had to be packaged outside the project or something
[12:02:44 CDT(-0500)] <athena> but if not, cool
[12:02:58 CDT(-0500)] <jeffbcross> typically is
[12:04:59 CDT(-0500)] <athena> sounds good
[12:05:15 CDT(-0500)] <athena> hoping having a zip of the project really works
[12:05:27 CDT(-0500)] <athena> when we tried it before it never really worked for android because of some weirdness in the build directories
[12:05:30 CDT(-0500)] <athena> but hopefully we can get that right
[12:12:02 CDT(-0500)] <jeffbcross> well, it worked for me aside from that run config
[12:12:41 CDT(-0500)] <jeffbcross> but would be good to have someone try it with fresh install who's never used uMobile/Appcelerator
[12:21:01 CDT(-0500)] <athena> that's great
[12:21:02 CDT(-0500)] <athena> and yes
[12:22:59 CDT(-0500)] <jeffbcross> I'll recruit someone here to test it
[12:23:18 CDT(-0500)] <jeffbcross> do you have any updates to the package, other than my run config?
[12:53:44 CDT(-0500)] <athena> not at the moment
[13:48:43 CDT(-0500)] <athena> i'd never actually played w/ the assembly plugin before
[13:48:44 CDT(-0500)] <athena> pretty cool
[13:48:54 CDT(-0500)] <athena> now succeeding in generating a umobile native app zip
[13:55:00 CDT(-0500)] <EricDalquist> yay
[13:55:00 CDT(-0500)] <EricDalquist> neat
[13:55:40 CDT(-0500)] <athena> yeah
[13:55:45 CDT(-0500)] <athena> still have to install an emulator and titanium
[13:55:51 CDT(-0500)] <athena> but hopefully this will make it easier
[15:06:02 CDT(-0500)] <athena> EricDalquist: is there a good way to call a maven goal on a submodule as part of the uportal build process?
[15:06:10 CDT(-0500)] <athena> either from build.xml or even just as part of the assembly script?
[15:06:21 CDT(-0500)] <EricDalquist> hrm
[15:06:32 CDT(-0500)] <EricDalquist> well the build.xml calls install on each sub module directly
[15:06:34 CDT(-0500)] <athena> if so i could actually generate test data for some of the portlets like the calendar
[15:06:39 CDT(-0500)] <athena> oh ok
[15:06:42 CDT(-0500)] <athena> that sounds positive then
[15:08:41 CDT(-0500)] <EricDalquist> yeah look at the maven-artifact-macro
[15:08:49 CDT(-0500)] <EricDalquist> it call <antcall target="mvn">
[15:08:54 CDT(-0500)] <EricDalquist> and passes in:
[15:08:55 CDT(-0500)]