2009-12-03T00:09:11 *** dbs has quit IRC 2009-12-03T00:26:19 *** pmplett has joined #evergreen 2009-12-03T00:49:42 *** error_23 has quit IRC 2009-12-03T01:22:59 *** error_23 has joined #Evergreen 2009-12-03T01:28:24 *** mck9 has left #evergreen 2009-12-03T02:42:42 *** eguest309 has joined #Evergreen 2009-12-03T02:43:43 *** eguest309 has left #Evergreen 2009-12-03T02:44:20 *** eguest309_ has joined #Evergreen 2009-12-03T03:13:01 *** branflakes has left #Evergreen 2009-12-03T04:24:41 *** eguest309 has joined #Evergreen 2009-12-03T04:33:56 *** jamesrf is now known as jamesrf-gone 2009-12-03T04:48:11 *** levani_el has joined #evergreen 2009-12-03T04:48:15 hello 2009-12-03T04:48:24 i have preblem in evergreen 1.6 2009-12-03T04:48:32 autogen.sh dont work 2009-12-03T04:48:42 *** eguest309_ has quit IRC 2009-12-03T04:48:59 when i type ./autogen.sh -c /openils/conf/opensrf_core.xml -u 2009-12-03T04:49:04 get ??? ./autogen.sh -c /openils/conf/opensrf_core.xml -u 2009-12-03T04:50:33 srfsh# login admin open-ils work 2009-12-03T04:50:39 request math work 2009-12-03T04:50:51 but can`t login with staff client 2009-12-03T05:10:05 1.6 work :)\ 2009-12-03T05:10:07 1.6 work :) 2009-12-03T05:10:10 thank all 2009-12-03T05:17:58 *** levani_el has quit IRC 2009-12-03T05:20:55 *** eguest309 has quit IRC 2009-12-03T05:22:00 *** eguest309 has joined #Evergreen 2009-12-03T06:52:20 *** plogik has quit IRC 2009-12-03T06:52:20 *** chrissharp123 has quit IRC 2009-12-03T06:52:20 *** atz has quit IRC 2009-12-03T06:52:20 *** twirlip has quit IRC 2009-12-03T06:52:20 *** dbwells has quit IRC 2009-12-03T06:52:20 *** rsinger has quit IRC 2009-12-03T06:52:20 *** gdunbar has quit IRC 2009-12-03T06:52:20 *** _dkyle_1 has quit IRC 2009-12-03T06:52:20 *** AbizzalsX has quit IRC 2009-12-03T06:52:20 *** error_23 has quit IRC 2009-12-03T06:52:20 *** brendan_ga has quit IRC 2009-12-03T06:52:20 *** Stompro has quit IRC 2009-12-03T06:52:20 *** lisppaste3 has quit IRC 2009-12-03T06:52:20 *** kbeswick has quit IRC 2009-12-03T06:52:20 *** eguest309 has quit IRC 2009-12-03T06:53:43 *** pinesol has joined #evergreen 2009-12-03T06:54:24 *** senator has joined #Evergreen 2009-12-03T06:54:30 *** phase_bb has joined #evergreen 2009-12-03T06:54:30 *** sboyette has joined #evergreen 2009-12-03T07:00:00 *** jamesrf-gone has joined #evergreen 2009-12-03T07:25:34 *** alxp has joined #evergreen 2009-12-03T07:28:32 *** bradl_ is now known as bradl 2009-12-03T07:46:46 *** atheos has joined #Evergreen 2009-12-03T07:53:31 *** eguest309 has quit IRC 2009-12-03T07:59:35 *** mck9 has joined #evergreen 2009-12-03T08:07:05 *** gdunbar has quit IRC 2009-12-03T08:10:38 *** phasefx3 has joined #evergreen 2009-12-03T08:12:54 *** phasefx has joined #evergreen 2009-12-03T08:13:14 *** phasefx__ is now known as phasefx_ 2009-12-03T08:47:35 *** branflakes has joined #Evergreen 2009-12-03T08:56:48 *** bshum has joined #evergreen 2009-12-03T09:01:50 *** bshum has quit IRC 2009-12-03T09:04:05 *** bshum has joined #evergreen 2009-12-03T09:04:13 *** berick_ is now known as berick 2009-12-03T09:15:35 *** jenny has joined #evergreen 2009-12-03T09:25:41 *** r123 has joined #evergreen 2009-12-03T09:27:40 *** gdunbar has joined #evergreen 2009-12-03T09:40:12 *** dcage has joined #evergreen 2009-12-03T09:46:08 *** _dkyle_1 has quit IRC 2009-12-03T09:46:59 *** dbs has joined #evergreen 2009-12-03T09:48:14 Morning everyone. I've been tinkering with trying to create a duplicate database from our 1.4 test server onto a second server so that we can try upgrading it safely to 1.6 but we've been having some troubles with it. 2009-12-03T09:48:39 For some reason, everything seems to load up correctly, but searching for items isn't returning any results. 2009-12-03T09:48:53 Looking up items directly in the staff client works though. 2009-12-03T09:49:09 Though only if we look them up by barcode. 2009-12-03T09:49:27 That seems to indicate to me that I must have missed something with the metabib part of the database 2009-12-03T09:49:51 bshum: do you have any data in metabib.keyword_field_entry? 2009-12-03T09:50:09 Yep. There's plenty of data there. 2009-12-03T09:50:31 I pulled the first 100 rows, just to check. 2009-12-03T09:50:31 But there's over 900,000 rows 2009-12-03T09:50:51 You've run ANALYZE? 2009-12-03T09:51:12 Yes, I've run analyze as well. 2009-12-03T09:51:53 I'm wondering if maybe there's a mis-step in my relocation 2009-12-03T09:51:56 if you create a new record, can you search for that? 2009-12-03T09:52:00 Hmm 2009-12-03T09:52:11 I'll give that a try. 2009-12-03T09:54:07 also, what does "show LC_CTYPE;" say at the psql prompt when you connect to the new database? 2009-12-03T09:54:44 * dbs flashes back to http://evergreen-ils.org/irc_logs/evergreen/2009-11/%23evergreen.27-Fri-2009.log - maybe we could build a troubleshooter 2009-12-03T09:56:11 The show LC_CTYPE command returned "en_US.UTF-8" 2009-12-03T09:57:18 okay. so you're probably missing a text_pattern_ops index or two, then 2009-12-03T09:57:45 I bet your original database has "C" instead of "en_US.UTF-8" 2009-12-03T09:58:19 Hmm 2009-12-03T09:58:26 I could check the original database 2009-12-03T09:58:40 that would be good 2009-12-03T09:59:01 Did you set up the original database? 2009-12-03T09:59:18 Heh, yes. 2009-12-03T09:59:32 Aha 2009-12-03T09:59:35 Yes, it is C 2009-12-03T09:59:48 For the original database. 2009-12-03T10:00:26 That's the likely problem, then. In 1.6 we added a bunch of text_pattern_ops indexes to help out the non-C locale database search speed, but the recommended approach is to start with a database cluster init'ed to C 2009-12-03T10:01:01 The link that I posted a few minutes back has more of a conversation with branflakes and miker_ that describes what's going on 2009-12-03T10:02:45 Oh - I lied, in 1.6/trunk we didn't actually create text_pattern_ops indices on the metabib.*fe tables 2009-12-03T10:04:20 just one on metabib.real_full_rec, and one on authority.full_rec - hmm 2009-12-03T10:06:41 if you're feeling spunky, you could create text_pattern_ops indices like: 2009-12-03T10:07:44 actually - probably wouldn't work due to the row size maximums. hmm 2009-12-03T10:09:59 at least, not for metabib.keyword_field_entry 2009-12-03T10:11:52 *** brendan_ga has quit IRC 2009-12-03T10:12:06 and then... tpo really only comes into play for LIKE or regexes, which probably doesn't come into play here 2009-12-03T10:14:41 Does this mean that I should drop the current database and start it over with the correct locale settings? 2009-12-03T10:14:56 bshum: you would have to create a new database cluster, too 2009-12-03T10:15:14 dbs: Okay. 2009-12-03T10:15:24 dbs: I'll give that a shot. 2009-12-03T10:15:34 first, though, it would be interesting to see if a new record could be retrieved 2009-12-03T10:15:40 Right 2009-12-03T10:18:12 staged search does use LIKE and ILIKE, so maybe the text_pattern_ops indices would be useful after all - still, branflakes reported that search was still far slower with tpo on a non-C locale than with locale C 2009-12-03T10:23:15 miker_ is free to beat me about the head and shoulders for any mangling I 2009-12-03T10:23:22 'm doing of this subject 2009-12-03T10:23:39 *** eguest309 has joined #Evergreen 2009-12-03T10:23:55 Well, it's definitely a very helpful direction. 2009-12-03T10:26:44 dbs: those tpo indexes are not useful -- they're (essentially) applied after the fts search to "disable" stemming on quoted words and phrases 2009-12-03T10:27:14 dbs: only left-anchored LIKE is helped by a tpo index ... and we don't do that in staged search 2009-12-03T10:27:29 * miker__ finishes the beatings ... for now 2009-12-03T10:27:30 ;) 2009-12-03T10:27:55 * dbs feels better 2009-12-03T10:28:06 thanks miker_ 2009-12-03T10:29:09 any other ideas on why bshum's database restore would result in non-functional search? 2009-12-03T10:29:39 sounds like metabib.metarecord_source_map or metabib.rec_descriptor could be empty 2009-12-03T10:30:47 could be, although I don't know why that part of the restore would fail... 2009-12-03T10:31:09 neither do I ... 2009-12-03T10:31:49 I have seen restores fail in weird ways when the partition for the database cluster filled up 2009-12-03T10:33:19 so it's probably worth bshum's time to count the entries in those tables, just in case 2009-12-03T10:35:16 The count for metarecord_source_map returned to me 96489 entries 2009-12-03T10:35:46 metabib.rec_descriptor came back with 96488 2009-12-03T10:35:52 sounds reasonable, then 2009-12-03T10:39:40 *** brendan_ga has joined #evergreen 2009-12-03T10:42:01 We tried adding a new record to our weird system and it couldn't find it. 2009-12-03T10:45:00 heh, okay 2009-12-03T10:47:38 Where does the initdb --locale=C need to be setup or done? 2009-12-03T10:47:53 If we were using a packaged version of Postgres 2009-12-03T10:48:00 Or do we have to set it up manually for it to work? 2009-12-03T10:54:49 You would still have to set it up manually 2009-12-03T10:55:27 the package uses the locale of the installing user, by default, when it creates the initial cluster 2009-12-03T10:55:36 *** jenny has quit IRC 2009-12-03T10:56:00 Gotcha 2009-12-03T10:56:03 so you could do something like: /usr/lib/postgresql/8.3/bin/initdb -d /data/postgresql_c_locale --locale C 2009-12-03T10:56:32 (assuming you're running as the 'postgres' user, and that 'postgres' has write perms to /data/postgresql_c_locale 2009-12-03T10:56:39 http://www.postgresql.org/docs/8.3/interactive/creating-cluster.html 2009-12-03T10:56:45 Thanks. 2009-12-03T10:57:01 it would be great to document this whole process 2009-12-03T10:57:26 *** jamesrf-gone is now known as jamesrf 2009-12-03T10:59:09 I guess I should take some nice notes then. 2009-12-03T11:16:15 bshum: actually, we have some notes thanks to r123 at http://biblio.laurentian.ca/tickets/conifer/wiki/sysadminPostgreSQLConfiguration#CreatingADataCluster 2009-12-03T11:16:23 r123++ 2009-12-03T11:16:51 Ah, nice. 2009-12-03T11:22:20 *** mrpeters-isl has quit IRC 2009-12-03T11:26:58 There's a note in opensrf.xml on 1.6 that says if you run a multi-brick installation that the vandelay import directory needs to be an NFS rw share, but it's not clear whether each individual brick gets its own, or whether it's one NFS share for the whole cluster. 2009-12-03T11:27:24 That is, if I have two bricks of three machines each (an OpenSRF router, and two drones), do I need two NFS shares for vandelay, or just one? 2009-12-03T11:28:53 branflakes: needs to be 1 global nfs share 2009-12-03T11:29:17 thanks berick, that was our interpretation of the note here, but we wanted to make sure :) 2009-12-03T11:29:21 np 2009-12-03T11:35:23 *** branflakes has quit IRC 2009-12-03T11:35:58 *** branflakes has joined #Evergreen 2009-12-03T11:44:12 *** mrpeters-isl has joined #evergreen 2009-12-03T11:50:39 *** mrpeters-isl has quit IRC 2009-12-03T11:50:42 *** mrpeters-isl has joined #evergreen 2009-12-03T11:55:32 dbs: btw, preventing holds on items already checked out... will be implementing that (w/ permission control) soon. will make sure everyone has the perm by default for backwards compat 2009-12-03T12:01:19 *** _dkyle_ has joined #evergreen 2009-12-03T12:04:42 berick: oh, very good to know! berick__ 2009-12-03T12:04:45 berick: oh, very good to know! berick++ 2009-12-03T12:18:59 *** branflakes has left #Evergreen 2009-12-03T12:25:43 *** twirlip is now known as raketemensch 2009-12-03T12:28:13 *** jenny has joined #evergreen 2009-12-03T12:41:00 dbs: It seems that using the command pg_dropcluster --stop 8.3 main will remove the main cluster that comes with the autoinstallation from using apt-get install postgresql-8.3, etc. Then you can build a new cluster using pg_createcluster --locale=C 8.3 main to replace it? 2009-12-03T12:41:25 Going to try now to rebuild the database using this new cluster. 2009-12-03T12:41:52 bshum: it sounds likely that you could do that, I was assuming that you would want to keep the old cluster / database around for comparison 2009-12-03T12:42:17 dbs: I have a few machines to work with. 2009-12-03T12:42:25 excellent! 2009-12-03T12:42:33 dbs: So I'm doing this on a different clean install. 2009-12-03T12:42:44 that's clearly the best approach 2009-12-03T12:42:48 bshum++ 2009-12-03T12:42:59 dbs: Would setting the locale=c on all future EG installs be the recommended option? 2009-12-03T12:43:22 dbs: Or is this only because of the leftovers from 1.4? 2009-12-03T12:43:24 Yes, that's the current (unstated) recommendation 2009-12-03T12:43:34 dbs: Okay, cool. 2009-12-03T13:11:12 *** eguest309 has quit IRC 2009-12-03T13:19:04 *** eguest309 has joined #Evergreen 2009-12-03T13:26:23 *** rickd has joined #evergreen 2009-12-03T13:36:20 *** raketemensch is now known as twirlip 2009-12-03T13:41:22 *** eby_ has joined #evergreen 2009-12-03T13:55:34 *** eguest309 has left #Evergreen 2009-12-03T13:57:49 *** eguest309_ has joined #Evergreen 2009-12-03T13:58:54 *** eguest309_ has left #Evergreen 2009-12-03T13:59:27 *** pmplett has joined #evergreen 2009-12-03T14:13:13 Why would the "Database Source Browser" when creating a report template be empty.? It just does not list any sources. 2009-12-03T14:16:08 moodaepo: sounds like a familiar problem, can't put my finger on it though. 2009-12-03T14:17:54 dbs Interesting, I've been searching the webs and can't see anyone else with a similar issue : ( 2009-12-03T14:18:22 moodaepo: I've seen it myself sometimes. did you create the symlink in /openils/var/web/xul? 2009-12-03T14:19:41 I'll check, this is a server I didn't setup. 2009-12-03T14:22:00 dbs: Yup I see it there, it's got the old "1_4_0_4" the newer "1_6_0_0" (RC1) and the symlink "server" pointing to the 1_6_0_0 2009-12-03T14:22:47 pointing to 1_6_0_0/server ? or just 1_6_0_0 2009-12-03T14:23:45 1_6_0_0/server 2009-12-03T14:25:25 does http://hostname/reports/fm_IDL.xml resolve? 2009-12-03T14:25:27 dbs: while I have your attention, you'd posted to the mailing list about a shared report templates repository sometime ago, any feedback on that? I didn't see any responses. 2009-12-03T14:25:36 moodaepo: yep, and nope 2009-12-03T14:27:13 dbs: Yup it resolves. 2009-12-03T14:28:10 then my guess is a javascript problem. hmm 2009-12-03T14:28:51 dbs: I was thinking of posting the basic ones I have setup to say a wikipage as a start. The entries in reporter.template.data should suffice (?) 2009-12-03T14:29:49 *** senator has quit IRC 2009-12-03T14:31:28 name and description might be helpful, too :) 2009-12-03T14:32:27 Ah fine! 2009-12-03T14:40:06 Well, I managed to reload our database, but there's still something wrong with it. 2009-12-03T14:40:43 We still can't find any of our items with keyword (or any) search. 2009-12-03T14:40:54 Though everything seems to be "there" when looking at the tables. 2009-12-03T14:41:03 The locale setting is definitely set to C this time though. 2009-12-03T14:41:07 So small victories? 2009-12-03T14:45:20 *** dcage has quit IRC 2009-12-03T14:46:33 bshum this is basically replicating your server right? 2009-12-03T14:48:09 and then attempting an upgrade 2009-12-03T14:48:15 Replicating it first. But the end goal is to attempt an upgrade, yp. 2009-12-03T14:48:17 *yep 2009-12-03T14:48:33 I don't suppose you are doing virtual machines are you? 2009-12-03T14:49:15 Not currently, I've just been using a rotation of different server machines that have been designated for testing. 2009-12-03T14:49:48 Maybe in the future. 2009-12-03T14:49:55 I upgraded by just replicating the virtual machine which hosted 1.4 and changed the ip on it and did the upgrade 2009-12-03T14:50:40 how are you reloading the db? 2009-12-03T14:50:42 My problem is that I'm not sure why we can't replicate our 1.4 outright. It's definitely been a puzzling couple of days. 2009-12-03T14:51:15 I've been using pg_dumpall or something like that to pull out everything, but I'm beginning to think that maybe I should try breaking it up into smaller parts. 2009-12-03T14:51:31 ok let's try another way then 2009-12-03T14:52:01 And I've used... let's see, psql -f filename to pull back everything 2009-12-03T14:53:32 one sec I'm trying to find my upgrade notes 2009-12-03T14:53:43 *** miker__ is now known as miker_ 2009-12-03T14:56:21 ok I did the same thing but in one line as per miker's suggestion "pg_dumpall -p 5432 | psql -d evergreen -p 5433 --single-transaction" 2009-12-03T14:56:39 I was upgrading from 8.2 to 8.3 so the ports are different 2009-12-03T14:57:02 *** branflakes has joined #Evergreen 2009-12-03T14:57:48 Right. What does that --single-transaction option do? 2009-12-03T14:58:51 miker_: moodaepo: if all the contrib modules are in place, theoretically, you'll just need to add the 000.english.pg83.fts-config.sql before loading the output of pg_dumpall ... but, I'd suggest running the load inside a transaction (-1 or --single-transaction on the psql command line) 2009-12-03T15:00:00 moodaepo: right ... also, I'd suggest using pg_dump against just the one db and loading that after the contribs are in place 2009-12-03T15:00:12 bshum: --^ 2009-12-03T15:00:13 contrib modules == postgresql contrib modules 2009-12-03T15:00:16 right 2009-12-03T15:00:42 tsearch2 (compat module in 8.3), tablefunc, etc 2009-12-03T15:04:09 So, that 000.english.pg83.fts-config.sql that you mentioned, this is for configuring the database initially then I take it? 2009-12-03T15:04:29 After loading all the contrib modules, that is. 2009-12-03T15:07:29 bshum: were you moving from 8.2 to 8.3? 2009-12-03T15:08:36 dbs: I attempted that but got thwarted by my current host of issues, so I last attempted to stick with 8.2. Though I suppose moving to 8.3 would be ideal? 2009-12-03T15:09:25 bshum: I was just checking to see if you actually needed to load 000.english.pg83.fts-config.sql first - which you wouldn't if you're just going from 8.2 to 8.2 2009-12-03T15:10:08 *** natschil has joined #evergreen 2009-12-03T15:10:08 dbs: But if I were to move from 8.2 to 8.3, then I would need to load that then? 2009-12-03T15:10:22 dbs: Along with I'm sure probably other things to consider. 2009-12-03T15:10:33 yep 2009-12-03T15:10:43 moodaepo: It's not theoretical. BC Sitka [jamesrf and I] successfully upgraded from 1.2.4.0 to 1.6.0.0, and from Postgres 8.2 to 8.3 last night using that exact procedure. 2009-12-03T15:12:18 branflakes++ 2009-12-03T15:12:25 jamesrf++ 2009-12-03T15:13:14 miker's advice. 2009-12-03T15:13:17 miker++ 2009-12-03T15:13:22 Errr, miker_++ 2009-12-03T15:13:47 moodaepo: can you check web/reports/oils_rpt_common.xhtml to see if it's pulling in dojo.js or openils_dojo.js? 2009-12-03T15:14:01 Umm, UPDATE karma SET value = value + 1 WHERE user LIKE '%miker%'; 2009-12-03T15:14:31 the only thing I can see in http://biblio.laurentian.ca/tickets/conifer/wiki/sysadminPostgreSQLConfiguration#RestoringFromADumpFile that might be of interest is the "checkpoint_segments" suggestion that r123 documented 2009-12-03T15:15:00 but I would have expected bshum to report that errors were thrown during the restore if that was an issue 2009-12-03T15:15:07 *** jenny1 has joined #evergreen 2009-12-03T15:15:08 *** branflakes is now known as branflakes-afk 2009-12-03T15:15:58 Yeah, I haven't hit on any weird errors during my restores yet. 2009-12-03T15:17:14 *** natschil has quit IRC 2009-12-03T15:17:18 *** natschil has joined #evergreen 2009-12-03T15:18:50 branflakes I did the same a few weeks ago so yea it does work : ) 2009-12-03T15:19:32 and we've done it several times (albeit 8.3 to 8.3) per http://biblio.laurentian.ca/tickets/conifer/wiki/sysadminPostgreSQLConfiguration 2009-12-03T15:20:01 and yes miker_++ 2009-12-03T15:22:08 dbs dojo.js 2009-12-03T15:22:28 moodaepo: dang. 2009-12-03T15:22:41 I know! 2009-12-03T15:24:28 moodaepo: any javascript console errors to report? 2009-12-03T15:24:53 *** senator has joined #Evergreen 2009-12-03T15:26:18 moodaepo pasted "PG 8.2 to 8.3 upgrade on the same server" at http://paste.lisp.org/display/91537 2009-12-03T15:26:38 branflakes-afk / jamesrf: congrats! would you guys mind writing up your full checklist for the community? 2009-12-03T15:27:00 *** natschil has quit IRC 2009-12-03T15:28:29 *** natschil has joined #evergreen 2009-12-03T15:32:06 *** jenny has quit IRC 2009-12-03T15:34:42 The value for "Holds: Hard boundary" - is that an org_unit depth? Or a # of proximity levels to travel? (e.g. to prevent items from transiting outside a given org_unit at depth 2 in the hierarchy, do I enter "2" or "0")? 2009-12-03T15:35:11 dbs: depth 2009-12-03T15:35:43 2 == "branch" (by default) == never look elsewhere for copies 2009-12-03T15:36:11 dbs: it's essentially setting the selection_depth field on holds 2009-12-03T15:36:13 mmm, so that would still be able to transit to other libraries at the same depth? 2009-12-03T15:36:53 the hold would only consider copies at exactly pickup lib 2009-12-03T15:36:57 http://www.open-ils.org/dokuwiki/doku.php?id=feature_list_1_2_2 is the only place I've found with any docs on the setting 2009-12-03T15:37:28 okay, that sounds about right then. part 2 of the request is "prevent copies from transiting - pickup lib must always == circ_lib" 2009-12-03T15:37:56 that would do it, yeah 2009-12-03T15:38:09 (at least for on campus libraries, people gots feet, they should use'em) 2009-12-03T15:38:19 miker_++ # racking up the karma 2009-12-03T15:38:20 dbs: think of boundary as a ceiling or umbrella. the size of the umbrella is determined by the depth 2009-12-03T15:38:40 frank_zappa: I think you stole berick's nick 2009-12-03T15:39:06 * berick plays off-kilter sax riff 2009-12-03T15:39:48 * dbs wonders if there are any catholic schools near eg 2010 2009-12-03T15:40:00 *** stevenyvr has joined #evergreen 2009-12-03T15:40:03 *** alxp has quit IRC 2009-12-03T15:41:08 * moodaepo holy perv : ) 2009-12-03T15:43:16 moodaepo pasted ""Database Source Browser" when creating a report template does not list any sources" at http://paste.lisp.org/display/91541 2009-12-03T15:44:49 dbs: I did see a couple of errors which I don't in my 1.6 install (not RC but the final release), I pasted them ^ along with notes of what we've checked earlier 2009-12-03T15:57:57 moodaepo: the source issue was definitely in the RC 2009-12-03T15:58:38 miker_ Thnx! 2009-12-03T15:59:01 and fixed before the release 2009-12-03T15:59:12 ah, hard boundaries won't affect the pickup org selector I guess 2009-12-03T15:59:28 dbs++ /for helping me try troubleshoot the dang issue! 2009-12-03T15:59:30 dbs: nope 2009-12-03T15:59:46 s/try/try to/ 2009-12-03T15:59:47 moodaepo: so you were running against the 1.6 RC? 2009-12-03T16:00:22 dbs yup I mentioned that in the line about the symlink : ) 2009-12-03T16:00:52 moodaepo: argh. I missed that (RC1). 2009-12-03T16:00:56 sorry 2009-12-03T16:03:22 hmm. this'll get tricky, then, I guess I'll have to limit the pickup selector to the owning_lib or circ_lib of the callnums/copies in question 2009-12-03T16:03:44 * dbs wonders why on earth anyone is still working with a 1.6RC1 test instance 2009-12-03T16:04:04 * dbs meanwhile is dealing with a 1.6RC0.9 production instance :) 2009-12-03T16:06:23 okay; the prototype for holdsBuildOrgSelector is (node) but it gets called with holdsBuildOrgSelector(null,0). maybe my brain isn't working, but that doesn't look right. 2009-12-03T16:10:58 dbs: yr brain is working. it's not right, but it works as needed anyway. 2009-12-03T16:11:23 heh, yay 2009-12-03T16:11:27 looks like there was an expectation of depth for the second param, but since the only time that expectation happens it's using depth 0, it's a no-op 2009-12-03T16:16:21 *** natschil has quit IRC 2009-12-03T16:21:37 *** stevenyvr has quit IRC 2009-12-03T16:24:16 I guess the way to do it would be to poke at open-ils.search.biblio.copy_location_counts.summary.retrieve and limit to the org_units in the payload as part of the holdsBuildOrgSelector - at least for our skin. 2009-12-03T16:29:38 bshum: oooh 2009-12-03T16:30:33 I completely forgot that the database server needs to have at least one Perl module installed - JSON::XS - before the database restore 2009-12-03T16:32:04 Hmm 2009-12-03T16:32:56 dbs: That would need to be done at the beginning somewhere I suppose then. 2009-12-03T16:32:59 bshum: yessir 2009-12-03T16:33:05 *wipes out his database (again) 2009-12-03T16:33:35 dbs: Are there some written steps for that process somewhere that I can find? 2009-12-03T16:34:25 dbs: do we have that documented somewhere? 2009-12-03T16:34:35 bshum: see http://biblio.laurentian.ca/tickets/conifer/wiki/sysadminPostgreSQLConfiguration#References (yes, that dang thing once again) right near the bottom 2009-12-03T16:34:45 "Install JSON::XS" 2009-12-03T16:35:43 dbs: Nice, thanks for checking on this for me. 2009-12-03T16:37:40 Anyone Egers going to the code4lib conf? 2009-12-03T16:37:48 s/Anyone/Any/ 2009-12-03T16:39:05 dbs: Hmm, the command for installing that module is giving me an error: "Unrecognized character \xE2 in column 21 at -e line." 2009-12-03T16:39:09 you'll need MARC::Record and MARC::File::XML too 2009-12-03T16:39:35 bshum: get rid of the ? 2009-12-03T16:39:43 dbs: But I switched into using the cpan command then installed JSON::XS directly. 2009-12-03T16:39:51 ugh 2009-12-03T16:40:15 yeah, I was just going to say "cpan JSON::XS" 2009-12-03T16:40:58 dshum what if you do "sudo make -f Open-ILS/src/extras/Makefile.install XXX" first, wouldn't that install the requirements...dbs? 2009-12-03T16:41:02 Shouldn't MARC::Record and MARC::File::XML be picked up during the EG install process? 2009-12-03T16:41:39 bshum: yes - but if you're running the database on a separate server, it won't 2009-12-03T16:41:51 dbs: Ah, I see. 2009-12-03T16:42:21 I don't know what your setup is like, so I'm just getting as much info out there as possible 2009-12-03T16:42:54 Gotcha. That'll be important for the future I think. Currently, we're just running everything off the same server. 2009-12-03T16:43:05 Well we don't need to install Eg just run the pre-reqs install : ) 2009-12-03T16:43:18 Someday I'll get the fun of learning how to get everything working on separate boxes. And in clusters. 2009-12-03T16:43:52 bshum yea I'm looking forward to it too! 2009-12-03T16:44:41 *** r123 has quit IRC 2009-12-03T16:46:17 moodaepo: it's documented now: http://www.open-ils.org/dokuwiki/doku.php?id=server:1.6.0:install 2009-12-03T16:46:43 bshum: ah okay. so not an issue in a single-server install then 2009-12-03T16:47:04 dbs++ /can there be enough karma in the world to shower over you sir : ) 2009-12-03T16:47:24 @later tell dbs Add those deps to the install_pgsql_server_debs_* Makefile.install targets 2009-12-03T16:47:24 dbs: The operation succeeded. 2009-12-03T16:48:12 *** dbs has quit IRC 2009-12-03T16:48:44 is there anything that would prevent patron_penalty.js from running in 1.6? 2009-12-03T16:48:53 it doesn't seem to be running at all 2009-12-03T16:54:43 quit 2009-12-03T16:54:48 *** bshum has quit IRC 2009-12-03T17:03:21 is berick still around? 2009-12-03T17:07:19 branflakes-afk: yo 2009-12-03T17:07:52 yo berick 2009-12-03T17:07:52 You were the vandelay expert earlier, IIRC, so I decided to bug you again 2009-12-03T17:08:37 Wait, never mind. I figured out the mystery. 2009-12-03T17:08:49 * berick bows, says goodnight 2009-12-03T17:08:51 ;) 2009-12-03T17:09:25 It looked like vandelay was uploading a file to both bricks by itself, but it was because the file was uploaded before we took a copy of the brick that received it, and put it on the other one :) 2009-12-03T17:10:02 What permissions does the share need? rw, and owned by? opensrf? 2009-12-03T17:10:16 branflakes-afk: that's it 2009-12-03T17:10:48 Great, thx muchly. 2009-12-03T17:10:48 *** branflakes-afk has left #Evergreen 2009-12-03T17:11:07 berick do you know about the patron_penalty.js? 2009-12-03T17:11:27 jamesrf: yes. 1 sec 2009-12-03T17:13:03 jamesrf: see #6 up from the bottom of http://open-ils.org/dokuwiki/doku.php?id=release_notes_1_6_0_0_rc1&s[]=release&s[]=notes 2009-12-03T17:13:45 in short, patron_penalty.js is deprecated in 1.6 in favor of in-db penalty configs 2009-12-03T17:13:55 so i can't use it at all? 2009-12-03T17:14:17 * berick verifies 2009-12-03T17:14:25 * jamesrf facepalms 2009-12-03T17:15:14 jamesrf: are you doing stuff in there that cannot be mapped into the DB penalties? 2009-12-03T17:15:20 no it can 2009-12-03T17:15:29 but i have to transcribe 25 libraries from the js files 2009-12-03T17:16:34 do you know is there a bug in the UI for that? it doesn't seem to list any org units 2009-12-03T17:16:46 and i have work locations set 2009-12-03T17:17:33 maybe if i fix that then our libraries can do it themselves 2009-12-03T17:17:38 jamesrf: that sounds like a permission issue. ADMIN_GROUP_PENALTY_THRESHOLD for write ops / VIEW_GROUP_PENALTY_THRESHOLD for read 2009-12-03T17:17:47 jamesrf: yes, delegating is the goal 2009-12-03T17:18:25 i can't guarantee those perms were added to the 1.6 seed data, so may need to be inserted 2009-12-03T17:18:46 *** eby_ has quit IRC 2009-12-03T17:20:18 ah thanks that's what i needed 2009-12-03T17:26:20 jamesrf you will see another UI issue after/while you enter the circ matrix rules in the client 2009-12-03T17:26:54 It doesn't seem to list them all properly for us 2009-12-03T17:28:54 *** jenny1 has left #evergreen 2009-12-03T17:33:05 it's probably another missing perm 2009-12-03T17:33:11 there's a few missing like ADMIN_SURVEY 2009-12-03T17:33:21 i don't know about the circ matrix because we aren't using that yet 2009-12-03T17:36:04 So where do you enter the patron penalties? Isn't that in Circulation Policies? 2009-12-03T17:36:55 the penalties are in admin -> local admin -> group penalty threshold 2009-12-03T17:37:25 you can set the max overdues or max items out or max fines 2009-12-03T17:38:11 aah ok 2009-12-03T17:38:50 it used to be in javascript but i actually like this better because our sites can eidt themselves 2009-12-03T18:24:47 *** atz has quit IRC 2009-12-03T18:25:28 *** atz has joined #evergreen 2009-12-03T18:47:14 jamesrf pasted "according to firebug, this is causing the problem when POSTed to our server. it is generated by group threshold UI" at http://paste.lisp.org/display/91559 2009-12-03T18:59:57 berick: since patron_penalty.js is deprecated, i would suggest someone remove it from opensrf.xml 2009-12-03T19:01:08 it is quite misleading 2009-12-03T19:02:54 jamesrf pasted "patch to remove deprecated patron penalty script" at http://paste.lisp.org/display/91562 2009-12-03T19:50:32 *** [1]atz has joined #evergreen 2009-12-03T19:58:01 *** atz has quit IRC 2009-12-03T19:58:01 *** [1]atz is now known as atz 2009-12-03T20:29:57 jamesrf: chances are you'll need to add open-ils.pcrud to your list of public services in opensrf_core.xml 2009-12-03T20:35:47 hmm ok thanks i'll have to give that a whirl after hours 2009-12-03T20:37:28 hm you mean the 'gateway' bit? 2009-12-03T20:39:17 no, under //config/opensrf/routers/router[domain[text()='public.localhost']]/services (adjust domain name to taste) 2009-12-03T20:39:33 well, maybe in ther gateway bit too... 2009-12-03T20:39:48 yes, in gateway/services too 2009-12-03T20:40:24 but the main one of conern is the router/services one 2009-12-03T20:41:45 yeah it's in the router/services 2009-12-03T20:41:56 but 2009-12-03T20:42:05 maybe it needs to be in the activeapps on the lead box of the brick? 2009-12-03T20:45:33 no, anywhere in the brick is fine 2009-12-03T20:57:30 *** brendan_ga has quit IRC 2009-12-03T20:58:39 *** dbs has joined #evergreen 2009-12-03T21:05:56 *** ChanServ sets mode: +o dbs 2009-12-03T21:06:00 *** dbs changes topic to "Welcome to the #Evergreen library system channel! | We are publicly logged. | Large pastes at http://paste.lisp.org/new/evergreen" 2009-12-03T21:06:06 *** ChanServ sets mode: -o dbs 2009-12-03T21:20:39 maker_++ 2009-12-03T21:20:43 er 2009-12-03T21:20:46 miker_++ 2009-12-03T21:20:52 maker_-- 2009-12-03T21:35:18 argh. argh. 2009-12-03T21:35:39 there's a nice little sentence tucked away at the end of the po2moz help 2009-12-03T21:35:47 "The file extension of the original Mozilla file is required to tell the Toolkit how to do the conversion. Therefore, a file like foo.dtd must be named foo.dtd.po in order to po2moz to recognise it as a DTD file. " 2009-12-03T21:36:03 *** brendan_ga has joined #evergreen 2009-12-03T21:36:11 which isn't GNU gettext-layout compliant. argh. 2009-12-03T21:36:18 i18n-- 2009-12-03T21:38:32 * dbs ponders a horrible hack of renaming the foo.dtd/ll-LL.po to foo.dtd/ll-LL-foo.dtd.po as part of the make step 2009-12-03T21:39:00 bah, what the hell 2009-12-03T22:09:36 there we go. hackity hack hack. 2009-12-03T22:15:09 *** lisppaste3 has quit IRC 2009-12-03T22:33:26 *** bshum has joined #evergreen 2009-12-03T23:25:54 *** pmplett has quit IRC 2009-12-03T23:50:02 *** dbs has quit IRC 2009-12-03T23:56:20 *** phasefx3 has quit IRC 2009-12-03T23:59:03 *** bshum has quit IRC