2008-03-31T01:31:55 *** Mark__T has joined #openils-evergreen 2008-03-31T01:43:29 *** dbs has quit IRC 2008-03-31T02:10:10 *** lisppaste6 has quit IRC 2008-03-31T02:10:10 *** pmurray_away has quit IRC 2008-03-31T02:13:35 *** lisppaste6 has joined #openils-evergreen 2008-03-31T02:13:35 *** pmurray_away has joined #openils-evergreen 2008-03-31T08:43:13 *** dbs has joined #OpenILS-Evergreen 2008-03-31T08:54:23 *** edc is now known as ecorrado 2008-03-31T09:18:48 *** pmurray_away is now known as pmurray 2008-03-31T09:21:27 *** rsinger_ is now known as rsinger 2008-03-31T09:43:39 *** Mark__T has quit IRC 2008-03-31T09:47:57 *** atheos has joined #OpenILS-Evergreen 2008-03-31T10:11:56 *** gmcharlt has joined #OpenILS-Evergreen 2008-03-31T10:58:22 *** dbs has quit IRC 2008-03-31T11:09:16 *** Karen_ has joined #OpenILS-Evergreen 2008-03-31T11:53:20 *** dbs has joined #OpenILS-Evergreen 2008-03-31T13:11:19 *** ecorrado is now known as edc_fooding 2008-03-31T14:01:37 *** edc_fooding is now known as ecorrado 2008-03-31T14:56:23 *** lisppaste6 has quit IRC 2008-03-31T15:09:23 *** lisppaste6 has joined #openils-evergreen 2008-03-31T15:50:51 dbs: miker_ tells me yr having search troubles on your trunk build... 2008-03-31T15:51:11 umm, I think I'm good now 2008-03-31T15:51:26 was it the search an empty collection bug maybe? 2008-03-31T15:52:05 seems fine at http://biblio-dev.laurentian.ca (searching McMaster or Windsor, they're both empty) 2008-03-31T15:52:12 something about a bug with low/no hits 2008-03-31T15:52:42 well, I saw problems initially, but then miker_ pumped out a few updates which I quickly applied 2008-03-31T15:52:49 I'm I smoking crack again on accident? 2008-03-31T15:53:05 haven't really seen anything since then, but then again I haven't been playing with it too much 2008-03-31T15:53:24 * miker_ swares there was a "knightrider bar of death" going on with <10 hits ... 2008-03-31T15:53:25 meh 2008-03-31T15:53:37 swears, even 2008-03-31T15:54:20 miker_: there might have been, but I was dealing with some caching issues too 2008-03-31T15:54:30 I might have to go back to our transcript of that fateful night 2008-03-31T15:54:49 * dbs notes that he should see if the site has been crawled yet 2008-03-31T16:05:12 ooh, mod_xmlent is a chatty cathy in debug mode 2008-03-31T16:05:18 * dbs wonders why he has it set to debug 2008-03-31T16:05:45 *** ecorrado is now known as edc_away 2008-03-31T16:05:52 :) 2008-03-31T16:06:19 oh, probably the default in eg.conf 2008-03-31T16:07:46 lisppaste6: help! 2008-03-31T16:07:47 To use the lisppaste bot, visit http://paste.lisp.org/new/openils-evergreen and enter your paste. 2008-03-31T16:07:57 yahoo, google, and msn have all stopped by, if the UAs can be trusted 2008-03-31T16:08:23 rsinger pasted "Jangle JSON response from connector" at http://paste.lisp.org/display/58362 2008-03-31T16:09:41 rsinger: rad! 2008-03-31T16:10:04 Oh dear, what have I wrought: XMLENT XML Parse Error: not well-formed (invalid token) at line 1: parsing /openils/var/web/opac/extras/slimpac/start.html: data \x1f\x8b\b 2008-03-31T16:12:24 miker_: just using vCard would simplify the AtomPub response, too 2008-03-31T16:12:54 (and, theoretically, the script that tried to consume it) 2008-03-31T16:12:55 rsinger: are you thinking that the ILS will return a vcard? or the connector? or that jangle will take a JSON object and map that to a vcard? (the latter, based on connecter-config, is close to what I was imagining after your last trip up) 2008-03-31T16:13:54 dbs: i wonder if it's data in lib_list.inc ... 2008-03-31T16:15:18 miker_: uh, all of the above, maybe 2008-03-31T16:15:35 miker_: definitely the last bit 2008-03-31T16:16:26 miker_: so, in that way 2008-03-31T16:16:30 rsinger: rock. just confirming that my mind has not /completely/ melted from UI work ;) 2008-03-31T16:16:36 miker_: you'd have some hash map laid out somewhere 2008-03-31T16:16:49 miker_: and you'd return some json 2008-03-31T16:17:03 miker_: and jangle would construct the requisite vcard for you 2008-03-31T16:17:04 dbs: interesting, i was able to get a JSON_v1 error on your trunk system 2008-03-31T16:17:17 miker_: or, what i'm doing here is just making a vcard in the connector 2008-03-31T16:17:17 berick: good thought 2008-03-31T16:17:51 miker_: that way no jangle/connector mapping has to occur 2008-03-31T16:17:56 but that should be a config option 2008-03-31T16:18:02 (in my mind) 2008-03-31T16:18:13 miker_: that's what we talked about, right? 2008-03-31T16:18:33 right ... UID... ... completely contrived and wrong, but you get what I mean :) 2008-03-31T16:18:43 * rsinger is procrastinating going to the convenience and getting ganked while inflating his car tires 2008-03-31T16:19:36 miker_: yeah, or YAML or something, but yes, i think i think either way can work 2008-03-31T16:19:52 and, like i said, it's the choice of the connector implementor 2008-03-31T16:19:55 right, yaml++ 2008-03-31T16:19:59 right 2008-03-31T16:20:27 dbs: when you can, could you paste the activity log entry for the staged search request created when going to this page: http://ln-s.net/1kjy The JS is not happy with the JSON response 2008-03-31T16:20:29 even per-... er ...method (but not method ... resource?) 2008-03-31T16:20:48 miker_: yes, definitely 2008-03-31T16:21:15 it would probably be helpful to set up a gui in the prototype jangle to demonstrate this for people 2008-03-31T16:21:24 s/people/connector developers/ 2008-03-31T16:21:30 so, there's really a cloud of small connectors (which may happen to live in a single file) which are configured per-resource 2008-03-31T16:21:38 s/connector developers/suckers/ 2008-03-31T16:22:10 yeah, i think things are going to have to be configured per resource 2008-03-31T16:22:35 although, realistically, that would be done at the connector level, not at the jangle installation level 2008-03-31T16:22:45 so if asnagy produces a voyager connector, say 2008-03-31T16:23:06 everybody can use his config 2008-03-31T16:23:06 (just change the URLs) 2008-03-31T16:23:48 we have our new jangle.org webhost 2008-03-31T16:24:04 i'm trying to figure out why the dns hasn't updated 2008-03-31T16:24:11 oh, fun! 2008-03-31T16:24:32 i think i'll wait until april 2nd to announce this great new project to get all your data out of your ils though :) 2008-03-31T16:25:22 ok, time to go get ganked 2008-03-31T16:25:26 good call 2008-03-31T16:25:28 have fun 2008-03-31T16:25:32 if i'm not back in 20 minutes, call 911 :) 2008-03-31T16:25:33 dbs pasted "searching for fiction - non-happiness" at http://paste.lisp.org/display/58363 2008-03-31T16:26:41 * dbs heads home, I should be back in a bit 2008-03-31T16:26:43 dbs: mind running the following SQL directly? 2008-03-31T16:26:53 SELECT array_accum(distinct id) FROM actor.org_unit_descendants( 12, 0 ); 2008-03-31T16:27:45 dbs annotated #58363 with "miker_'s direct sql wish granted" at http://paste.lisp.org/display/58363#1 2008-03-31T16:27:58 wth 2008-03-31T16:28:20 haven't run any proximity crud yet 2008-03-31T16:28:28 (as a completely irrelevant data point) 2008-03-31T16:28:38 hrm.. 2008-03-31T16:29:03 yeah ... no prox used here 2008-03-31T16:29:07 okay - really going now 2008-03-31T16:29:11 lator 2008-03-31T16:29:26 *** dbs has quit IRC 2008-03-31T16:42:18 *** kados has quit IRC 2008-03-31T16:42:24 *** kados has joined #openils-evergreen 2008-03-31T16:45:19 * rsinger returns, notices the large metal thing sticking from tire :( 2008-03-31T16:45:54 better tire than kidney, i suppose 2008-03-31T17:01:25 true 2008-03-31T17:01:42 *** pmurray is now known as pmurray_away 2008-03-31T17:56:11 *** greg-g has joined #openils-evergreen 2008-03-31T18:03:39 *** Karen_ has left #OpenILS-Evergreen 2008-03-31T18:14:44 *** jmellis has joined #openils-evergreen 2008-03-31T18:36:34 *** djfiander has joined #OpenILS-Evergreen 2008-03-31T18:43:25 *** jmellis has left #openils-evergreen 2008-03-31T19:11:38 *** gmcharlt is now known as gmcharlt-away 2008-03-31T20:22:17 dojo submit buttons do not work as well as one might like :( 2008-03-31T20:22:34 djfiander: how's that 2008-03-31T20:22:58 they fail to set the value of the button: http://trac.dojotoolkit.org/ticket/6215 2008-03-31T20:23:14 so, the button gets poked, but the value attached to the button is not returned to the server 2008-03-31T20:24:00 that bug report is odd, but I am seeing the reported behaviour 2008-03-31T20:24:04 with 1.1 2008-03-31T20:26:11 and I know it's dojo because if I delete the dojoType decl from the button, and change NOTHING ELSE, it works perfectly 2008-03-31T20:27:15 odd indeed 2008-03-31T20:27:25 I think I need to step back from dojo and go back to working on the vanilla workflow for creating POs, so we have some experience with the whole process 2008-03-31T20:27:38 do you need the button's value? (real question ... I usually throw it away ;) ) 2008-03-31T20:27:54 yes. I'm cheating badly. 2008-03-31T20:28:03 well, I won't try to stand in your way :) 2008-03-31T20:28:22 when looking at the picklist, there's an "update" button for each item, to set the number of copies of that item 2008-03-31T20:29:00 each title in the picklist is a separate form. All the forms have the same action. I figure out which one to update by using the button value. 2008-03-31T20:29:09 I could just as easily use a hidden input field I guess. 2008-03-31T20:29:13 *sigh* 2008-03-31T20:29:43 I see... 2008-03-31T20:30:13 so, you're using a