2008-09-22T01:28:57  *** Mark__T has joined #openils-evergreen
2008-09-22T07:08:59  *** kgs has joined #openils-evergreen
2008-09-22T08:26:53  *** bag has joined #openils-evergreen
2008-09-22T08:44:50  *** csharp_LT has joined #openils-evergreen
2008-09-22T08:59:55  *** csharp_LT has quit IRC
2008-09-22T09:04:47  *** dbs has joined #openils-evergreen
2008-09-22T09:06:47  <dbs> holy crow, phasefx - you know that you're supposed to work during the week, not on weekends, right?
2008-09-22T09:09:10  <phasefx> well, I'm supposed to do migrations on weekends :)
2008-09-22T09:11:51  <dbs> phasefx: heh! and from the hours on those commits, looks like most of them fell on Monday anyway :)
2008-09-22T09:13:54  <phasefx> :D
2008-09-22T09:18:52  *** wlayton has joined #openils-evergreen
2008-09-22T09:26:54  *** sarabee has quit IRC
2008-09-22T09:35:00  <lisppaste6> phasefx pasted "script tags and xulrunner 1.9" at http://paste.lisp.org/display/67270
2008-09-22T09:37:17  <dbs> phasefx: you know it's invalid to have an empty script tag, right?
2008-09-22T09:37:31  <dbs> I guess the question is, does XUL know that...
2008-09-22T09:38:53  <phasefx> you mean <script src="blah"></script> is invalid?
2008-09-22T09:39:29  <dbs> no, <script src="blah" /> is invalid (in xhtml -- maybe in XUL it's the reverse)
2008-09-22T09:40:00  <phasefx> I thought XUL accepted both, but apparently not now
2008-09-22T09:41:36  <dbs> that is shitty
2008-09-22T09:42:09  <phasefx> well, that can't be right
2008-09-22T09:42:22  <phasefx> unless it's only true for remote xul
2008-09-22T09:42:31  <phasefx> the chrome has <script/> and doesn't mind at all
2008-09-22T09:44:54  <dbs> phasefx: you could roll back to a 1.2 client but keep the XMLENT settings the same
2008-09-22T09:45:26  <dbs> then turn off the XMLENT settings, and see if that is indeed where the problem is being introduced
2008-09-22T09:46:12  <phasefx> with trunk client and 1.8 xulrunner, no problems
2008-09-22T09:46:25  <phasefx> with 1.2 client and 1.9 xulrunner, problems
2008-09-22T09:46:46  <phasefx> let me dig up a 1.2 server
2008-09-22T09:46:58  <phasefx> or
2008-09-22T09:47:12  <phasefx> I can turn xmlent off in trunk server
2008-09-22T09:47:24  <phasefx> but I'll get entity errors
2008-09-22T09:47:26  <dbs> exactly
2008-09-22T09:47:34  <phasefx> gotcha
2008-09-22T09:55:44  <lisppaste6> phasefx annotated #67270 with "a 1.2 server with xmlent off for xul files" at http://paste.lisp.org/display/67270#1
2008-09-22T09:56:29  <dbs> weird - do we have any entities defined in 1.2?
2008-09-22T09:57:17  <phasefx> looks like it.. that's &staff.auth.title; :)  I'll rip them all out for this test
2008-09-22T09:57:34  <dbs> oh yeah - lang.dtd hardcoded to en-US
2008-09-22T09:58:51  <phasefx> in server/main/data.xul, with the way we're ssi'ing the dtd, I don't think we even need xmlent
2008-09-22T09:59:03  <phasefx> well, no, that's obviously not true
2008-09-22T09:59:11  <phasefx> because it barfed.. but.. weird
2008-09-22T09:59:27  <phasefx> it should be just inlining the entity definitions, right?
2008-09-22T10:00:44  <lisppaste6> phasefx annotated #67270 with "with xmlent off and that one entity removed from server/main/data.xul" at http://paste.lisp.org/display/67270#2
2008-09-22T10:00:46  <dbs> phasefx: right, in theory if you wget the file you would see the entities defined there
2008-09-22T10:01:10  <dbs> heh
2008-09-22T10:01:23  <dbs> and if you change that to <script blah></script> ?
2008-09-22T10:01:50  *** rsinger has quit IRC
2008-09-22T10:02:51  <lisppaste6> phasefx annotated #67270 with "maybe it's the chrome url in @src?" at http://paste.lisp.org/display/67270#3
2008-09-22T10:04:16  <phasefx> dbs: yeah, it doesn't complain about the script tag pointing to an http file, but does complain about the chrome
2008-09-22T10:04:20  <dbs> or maybe it's the preceding <scripts  id="openils_util_scripts"/> tag
2008-09-22T10:04:45  <phasefx> I'm actually surprised there are chrome url's in there
2008-09-22T10:05:11  <phasefx> I think you can only go from remote to chrome in very limited circumstances, like with stylesheets
2008-09-22T10:05:22  <dbs> http://www.xulplanet.com/references/elemref/ doesn't define "scripts" as an element, where is that coming from?
2008-09-22T10:05:30  <phasefx> those were probably ignored scripts in xr 1.8
2008-09-22T10:06:18  <phasefx> scripts gets overlayed with a set of explicit script tags.. any element name is fair game with xul.. it just gets styled as box if anything displays
2008-09-22T10:06:55  <dbs> okay
2008-09-22T10:07:37  <phasefx> well, at least one of those chrome script tags was actually being used in 1.8.. the one loading JSAN.. I'll try switching it to a remove JSAN
2008-09-22T10:07:39  <dbs> I just find it weird that the value of an attribute would trigger an XML parsing error
2008-09-22T10:07:40  <phasefx> remote
2008-09-22T10:07:49  <phasefx> that's true too
2008-09-22T10:07:50  <phasefx> hrmm
2008-09-22T10:07:55  <phasefx> let me do this test
2008-09-22T10:08:38  <dbs> but I suppose if the schema is defined via XSD then that could be triggered
2008-09-22T10:10:00  <phasefx> removing those script/chrome tags got us past the error
2008-09-22T10:10:02  <dbs> you can see why I didn't have a hope in hell of making this work with xulrunner 1.9
2008-09-22T10:12:25  <dbs> heh, this might be useful at some point: http://developer.mozilla.org/en/XUL_improvements_in_Firefox_3
2008-09-22T10:12:37  <phasefx> bleh :)  Security Error: Content at http://blah may not load or link to chrome://blah
2008-09-22T10:12:47  <phasefx> yeah, I've been perusing that
2008-09-22T10:13:32  <phasefx> so it looks like even my chrome stylesheets are out of bounds for remote xul now.. I wonder about bindings and messagecatalog
2008-09-22T10:13:33  <dbs> datepicker and timepicker, eh? Maybe we won't need to dojoize it all
2008-09-22T10:13:47  <phasefx> maybe
2008-09-22T10:14:48  <phasefx> this is just wrong.. they make an exception for chrome://global/skin/, apparently
2008-09-22T10:15:46  <dbs> or maybe we have to dojo-ize the whole damned thing, if security "improvements" are screwing you over
2008-09-22T10:16:36  <phasefx> well, I'm going to declare 1.2 not compatible with xr 1.9 and focus on 1.4 :)
2008-09-22T10:18:34  <phasefx> dbs: I say ssh and ncurses for any new staff client :)
2008-09-22T10:18:43  <dbs> sounds reasonable to me - is your plan to move to xr 1.9 for 1.4?
2008-09-22T10:19:12  <dbs> that will impact whether I go with Dojo date/time pickers or native XUL option
2008-09-22T10:19:35  <phasefx> if the roadblocks aren't too terrible.. evaluating now
2008-09-22T10:20:18  <dbs> phasefx: fantabulous - I won't be able to do squat for a few days anyway
2008-09-22T10:20:45  *** Mark__T has left #openils-evergreen
2008-09-22T10:21:31  <phasefx> dbs: even if we didn't, we could probably take the xbl/css for those widgets and use them anyway.. unless they make heavy use of xpcom, and then I'm in foreign territory
2008-09-22T10:27:57  <phasefx> http://developer.mozilla.org/en/Chrome_Registration#contentaccessible
2008-09-22T10:28:18  *** wlayton has quit IRC
2008-09-22T10:30:11  <dbs> huh
2008-09-22T10:46:12  <lisppaste6> phasefx pasted "Error: Could not load 'openils.I18N'" at http://paste.lisp.org/display/67271
2008-09-22T10:53:56  *** phase_bb has quit IRC
2008-09-22T11:07:18  <dbs> phasefx: that's the internationalization Dojo widget
2008-09-22T11:08:04  <_bott_> Who wants to talk SIP server?  I've never run it in a cluster, but I'm guessing it just needs to live on one of the app servers?
2008-09-22T11:10:00  <dmcmorris_esi> _bott_: pretty much... if you have multiple SIP servers, they can be load balanced with iptables
2008-09-22T11:10:02  <dmcmorris_esi> ...
2008-09-22T11:10:22  <_bott_> 31307008866651
2008-09-22T11:10:33  <_bott_> oops, wrong screen!
2008-09-22T11:11:48  <_bott_> dmcmorris_esi: Thanks.  I've only got 7 self-checks to worry about, so load shouldn't be an issue, I'll just have it ready to run elsewhere for redundancy
2008-09-22T11:26:15  <dbs> phasefx: is the fieldmapper service running on your test server? just wondering if that might be why openils/I18N.js isn't loading
2008-09-22T12:18:04  <phasefx> dbs: I can invoke the fieldmapper class viewer in the staff client.. does that help?
2008-09-22T12:18:47  <dbs> phasefx: that doesn't depend on open-ils.fieldmapper though; that just parses fm_IDL.xml directly, doesn't it?
2008-09-22T12:19:31  <phasefx> it uses a service.. lemme see which one
2008-09-22T12:19:46  * berick wonders if dbs means open-ils.fielder
2008-09-22T12:20:03  <dbs> berick: ah, of course!
2008-09-22T12:20:57  <phasefx> the class viewer calls opensrf.open-ils.system.fieldmapper from open-ils.actor
2008-09-22T12:21:32  *** rsinger has joined #OpenILS-Evergreen
2008-09-22T12:22:06  <phasefx> the dojo files.. they should live in /openils/var/web/js/dojo/ right, such that you have a dojo/dojo/dojo.js?
2008-09-22T12:22:21  <dbs> phasefx: yes/yes/yes
2008-09-22T12:22:39  <phasefx> I had them like js/dojo and js/dijit at first
2008-09-22T12:22:56  <phasefx> I can change the wiki
2008-09-22T12:23:34  * dbs tries out the wiki instructinos
2008-09-22T12:23:47  <dbs> I probably screwed that up, yeah. sigh
2008-09-22T12:23:51  <phasefx> k
2008-09-22T12:23:59  <dbs> yep
2008-09-22T12:24:39  <phasefx> but that still leads me to the openils.I18N thing
2008-09-22T12:25:12  <dbs> fixored
2008-09-22T12:25:23  <dbs> so do you have open-ils.fielder running?
2008-09-22T12:25:28  *** Toccoa has joined #openils-evergreen
2008-09-22T12:26:00  <phasefx> introspect says yes
2008-09-22T12:26:16  <Toccoa> I am looking for a place to report that the "Find a Library Near Me" link on the GAPINES catalog page is not working.
2008-09-22T12:27:15  <phasefx> Toccoa: I'll email their webmaster.  Thanks!
2008-09-22T12:27:25  <phasefx> or I can give you his email address if you want easy follow-up
2008-09-22T12:27:36  <Toccoa> Whichever is easier for you!
2008-09-22T12:28:37  <phasefx> alright, I fired that off
2008-09-22T12:28:47  <Toccoa> Thank you.
2008-09-22T12:29:30  <phasefx> incidentally, if you need the functionality now, it appears to be here:  http://georgialibraries.org/directories/publiclibdir.php
2008-09-22T12:29:59  *** Toccoa has quit IRC
2008-09-22T12:30:10  <phasefx> dbs: so yes on open-ils.fielder
2008-09-22T12:30:30  <dbs> phasefx: right, got that. hmm
2008-09-22T12:31:23  <phasefx> dbs: it works with your trunk?
2008-09-22T12:34:20  <dbs> phasefx: I have to update to catch up to you
2008-09-22T12:35:37  <dbs> phasefx: when do you get this JS error/
2008-09-22T12:36:28  <phasefx> doing a search in the opac (outside of the staff client.. it happens in the client too, though)
2008-09-22T12:36:59  <phasefx> also, placing a hold will trigger it
2008-09-22T12:37:16  <phasefx> or trying to place a hold, rather.. because you can't place a hold with that error happening
2008-09-22T12:38:51  <dbs> huh - can't trigger it with staff client against our test server comet (although I do get a few zillion "reference to undefined property root.name" warnings provoked by common/js/utils.js)
2008-09-22T12:39:17  <phasefx> where should that file exist?
2008-09-22T12:40:24  <dbs> not sure - do you see the same problem running searches on comet.cs.uoguelph.ca ?
2008-09-22T12:40:39  <phasefx> gee, now I'm getting other weird stuff.. like rec.isbn is not a function
2008-09-22T12:41:11  <dbs> IE6 or something?
2008-09-22T12:41:16  <phasefx> ff3
2008-09-22T12:41:38  <dbs> supah-weird
2008-09-22T12:41:53  <phasefx> so it's looking for /js/dojo/openils/I18N.js
2008-09-22T12:42:03  <dbs> the "root.name" problem is in line 417 of utils.js,  findNodeByName() function
2008-09-22T12:42:13  <dbs> oh, I thought you were asking about root.name - sorry!
2008-09-22T12:42:33  <dbs> yes, that should have been installed as part of an install from trunk
2008-09-22T12:43:49  <dbs> part of Open-ILS/web/js/dojo
2008-09-22T12:44:28  <phasefx> okay, so openils/ should be a peer to dojo/ digit/ and dojox/
2008-09-22T12:44:57  <berick> phasefx:
2008-09-22T12:44:58  <berick> erickson@app07% ls /openils/var/web/js/dojo/
2008-09-22T12:44:59  <berick> dijit  dojo  DojoSRF.js  dojox  fieldmapper  openils  opensrf  OpenSRF.js  util
2008-09-22T12:45:26  <phasefx> berick: thanks man
2008-09-22T12:45:58  <berick> you don't need the 'util' dir
2008-09-22T12:47:03  <berick> and, IIR, OpenSRF.js is old
2008-09-22T12:47:31  <phasefx> that fixed things right up
2008-09-22T12:53:12  <dbs> phasefx: did that stuff not get placed in the right directory by the autotools install? or did you sidestep the official installer?
2008-09-22T12:54:14  <phasefx> dbs: I had followed the wiki instructions.. and then must have broken things when I place dojo/ dijit/ dojox/ into a common dojo/ dir
2008-09-22T12:54:50  <phasefx> or to put it another way, I tried to fix dojo by hand and broke more stuff :)
2008-09-22T13:22:54  <phasefx> berick: re: get_org_tree, I can't reproduce whatever I was seeing before.. so I'm going to write it off as a fluke for now
2008-09-22T13:25:06  <berick> phasefx: k
2008-09-22T13:46:12  *** Branflakes has joined #openils-evergreen
2008-09-22T13:46:46  <phasefx> berick: oh, duh.  The reason why I'm not seeing the get_org_tree error is because I "fixed" it.  See http://svn.open-ils.org/trac/ILS/changeset/10654   I didn't escape my log message properly, but meant to say that I tried invoking it as $self->get_org_tree($client) and still didn't work
2008-09-22T13:53:05  <phasefx> berick: did another change outside of my turf too, unrelated: added an event to ils_events.xml, changeset 10657
2008-09-22T14:09:48  <berick> phasefx: cool
2008-09-22T15:32:55  *** dbs has quit IRC
2008-09-22T15:33:00  *** dbs has joined #openils-evergreen
2008-09-22T16:04:00  *** bag has quit IRC
2008-09-22T16:11:18  <_bott_> I found the event "COPY_NEEDED_FOR_HOLD" in an example circ script and used it (I don't think I just made that up!).   ...it seems there is no permission for "COPY_NEEDED_FOR_HOLD.override".  What kind of voodo hides behind the System Permission Editor?
2008-09-22T16:16:03  <berick> _bott_: not much.  you literally create the perm (and optional description), then apply it to whichever groups you choose
2008-09-22T16:16:36  <_bott_> berick: am I correct in thinking that's what's needed then?
2008-09-22T16:16:49  <berick> yep, that's it
2008-09-22T16:16:59  <_bott_> alrighty, here goes nothin'
2008-09-22T16:41:51  *** wlayton has joined #openils-evergreen
2008-09-22T16:45:18  <Branflakes> berick: It looks like added content caching is done in the master module, OpenILS::WWW::AddedContent, and not it's child modules like .::Amazon, .::Syndetics, etc.
2008-09-22T16:45:49  <Branflakes> Is the idea that the content objects you're caching should be generic regardless of the added content backend you're querying?
2008-09-22T16:46:20  <berick> Branflakes: that's the idea
2008-09-22T16:47:54  <Branflakes> Which gets to my actual question: I have an 'available content' object that I'd like to cache for Content Cafe, but it never gets used publicly, just to determine whether content's available.
2008-09-22T16:48:04  <Branflakes> And other modules won't have an analogue.
2008-09-22T16:48:30  * berick recalls such a thing
2008-09-22T16:48:56  <Branflakes> (analog for yankees). Is it ok to cache that in the child module, or should I try to understanding the caching logic a little more and push the cache calls up into OpenILS::WWW::AddedContent?
2008-09-22T16:49:17  <Branflakes> Or is working code better than good code, and I should just get it done? :)
2008-09-22T16:49:22  <berick> i would recommend caching that in the child module, since it's never served
2008-09-22T16:49:33  <berick> and only relevent to that plugin
2008-09-22T16:51:24  <berick> which, as you've already seen, is trivial.  $cache = OpenSRF::Utils::Cache->new; $cache->put_cache(key, value, timeout); $obj = $cache->get_cache
2008-09-22T16:52:27  <berick> the cached value can be any Perl JSON-ifiable object (hash/array/string/number)
2008-09-22T16:53:21  <berick> Branflakes++ # not afraid to getrdone
2008-09-22T16:55:06  *** skmurphy has joined #openils-evergreen
2008-09-22T17:02:47  *** wlayton has quit IRC
2008-09-22T17:19:45  *** bag has joined #openils-evergreen
2008-09-22T17:28:14  *** skmurphy_ has joined #openils-evergreen
2008-09-22T17:28:41  *** skmurphy has quit IRC
2008-09-22T17:28:56  *** skmurphy_ is now known as skmurphy
2008-09-22T17:38:18  *** dbs has quit IRC
2008-09-22T17:48:50  *** error_23 has joined #openils-evergreen
2008-09-22T17:52:26  *** error_23_ has joined #openils-evergreen
2008-09-22T17:52:54  *** error_23 has quit IRC
2008-09-22T17:53:08  *** error_23_ is now known as error_23
2008-09-22T17:56:36  *** error_23_ has joined #openils-evergreen
2008-09-22T17:57:06  *** error_23 has quit IRC
2008-09-22T17:57:19  *** error_23_ is now known as error_23
2008-09-22T18:00:24  *** error_23_ has joined #openils-evergreen
2008-09-22T18:00:48  *** error_23 has quit IRC
2008-09-22T18:00:52  *** error_23_ is now known as error_23
2008-09-22T18:23:45  *** phase_bb has joined #openils-evergreen
2008-09-22T18:32:12  *** lstara has joined #openils-evergreen
2008-09-22T18:33:20  *** lstara has quit IRC
2008-09-22T18:41:03  *** error_23 has quit IRC
2008-09-22T19:40:18  *** rsinger has quit IRC
2008-09-22T20:02:06  *** sarabee has joined #openils-evergreen
2008-09-22T20:38:10  *** kgs_ has joined #openils-evergreen
2008-09-22T20:44:05  *** rsinger has joined #OpenILS-Evergreen
2008-09-22T20:47:32  *** rsinger has quit IRC
2008-09-22T20:50:01  *** Branflakes has quit IRC
2008-09-22T21:24:24  *** dbs has joined #openils-evergreen
2008-09-22T21:31:13  *** rsinger has joined #OpenILS-Evergreen
2008-09-22T21:35:09  <dbs> kewl, there's movement on the BOTS list
2008-09-22T21:42:26  <dbs> okay, going to try setting up upei's ils-contrib accounts, so if warning bells flash on your pagers, you'll know it's me doing something boneheaded
2008-09-22T21:43:11  *** rsinger has quit IRC
2008-09-22T21:56:55  *** rsinger has joined #OpenILS-Evergreen
2008-09-22T22:25:05  *** rsinger has quit IRC
2008-09-22T22:35:58  *** kgs_ has quit IRC
2008-09-22T23:53:40  *** dbs has quit IRC