2011-02-22T00:04:36 *** Tstallis has quit IRC 2011-02-22T00:35:21 *** bshum has quit IRC 2011-02-22T00:42:19 *** brian_f has quit IRC 2011-02-22T03:20:55 *** mrpeters-isl has quit IRC 2011-02-22T04:12:33 *** rsinger has joined #evergreen 2011-02-22T07:36:11 *** rickd_ has joined #evergreen 2011-02-22T07:38:22 *** sfortin has joined #evergreen 2011-02-22T08:23:13 *** kmlussier has joined #evergreen 2011-02-22T08:24:36 *** rjackson-isl has joined #evergreen 2011-02-22T08:26:04 *** collum has joined #evergreen 2011-02-22T08:26:17 *** rickd_ has quit IRC 2011-02-22T08:38:46 *** AaronZ-PLS has joined #evergreen 2011-02-22T08:40:46 *** rickd_ has joined #evergreen 2011-02-22T08:58:42 *** _bott_ has joined #evergreen 2011-02-22T09:05:08 *** Meliss has joined #evergreen 2011-02-22T09:09:11 * dbs pasted http://paste.lisp.org/display/119895 "hmm. http_translator issues with unicode?" 2011-02-22T09:28:24 *** yboston has joined #evergreen 2011-02-22T09:31:25 *** 77CAAR8OJ has joined #evergreen 2011-02-22T09:31:44 *** 77CAAR8OJ is now known as Dmagick 2011-02-22T09:33:38 dbs: hmm, perhaps the C code is creating invalid XML, would explain the jabber disconnect. wonder if the ejabberd log shows anything 2011-02-22T09:34:07 berick: doesn't look like it 2011-02-22T09:34:30 berick: you see http://svn.open-ils.org/trac/OpenSRF/changeset/2168 ? 2011-02-22T09:36:05 *** jenny has joined #evergreen 2011-02-22T09:37:48 dbs: I think this one is due to mangling by freethaw; upgrading opensrf could resolve it 2011-02-22T09:38:01 *** bshum has joined #evergreen 2011-02-22T09:38:38 gmcharlt: upgrading to OpenSRF trunk now that I've fixed the bug in trunk? 2011-02-22T09:38:47 Unicode fun :) 2011-02-22T09:39:31 dbs: you fixed a different bug, methinks, than the freezethaw issue 2011-02-22T09:39:39 agreed, unicode is fun ;) 2011-02-22T09:40:00 gmcharlt: yes, that's why I said that 2011-02-22T09:40:32 I recognized that it was a different bug. Problem: we don't have an OpenSRF trunk/2.0 official release yet for them to move to 2011-02-22T09:44:08 *** granitize has joined #evergreen 2011-02-22T09:45:18 dbs: Hmm, did we decide to have a dev meeting today at 12 EST? Or delaying again? 2011-02-22T09:45:42 bshum: not sure 2011-02-22T09:49:08 one difference between 1.6.2 and trunk/2.0 is that characters in the basic multilingual plane (like "é") work in 1.6.2; on trunk / 2.0 even those were failing horribly 2011-02-22T09:53:14 gmcharlt: you're right about freezethaw being the problem in 1.6.2, just reproduced it with the 2.0.1 image using characters outside the basic multilingual plane 2011-02-22T10:32:37 Boo, the A/T interface in 2.0 is still framed funny :( 2011-02-22T10:34:48 Can't see the A/T events at the bottom of the screen without doing some tabbing to the lower areas. 2011-02-22T11:02:42 gmcharlt: I don't get why unicode used to work with freezethaw in previous releases but now chokes terribly. Is it a regression in FreezeThaw since 0.43 (2002 release - next release was in 2009), do you think? 2011-02-22T11:04:55 dbs: not sure 2011-02-22T11:05:26 0.44 "Optimize thaw" - hmm 2011-02-22T11:05:47 I'll see if I can roll back to 0.43 2011-02-22T11:24:06 fwiw, rolling back to FreezeThaw 0.43 doesn't seem to make a difference 2011-02-22T11:25:58 *** parsr has joined #evergreen 2011-02-22T11:33:22 My upgrade from 1.6.1 to 2.0 failed with latest changeset to fix FK constraint error (but I got further). I wonder if this was b/c I re-run the upgrade script on database that had already one upgrade script run against it? Before I report I'd like to reload our PG 1.6.1.4 database onto dev server and re-run upgrade script as part of a complete repeat full upgrade steps. 2011-02-22T11:47:29 parsr: hard to tell without seeing your error log 2011-02-22T11:49:22 dbs: FWIW error logs here (temporarily): http://geoscan.ess.nrcan.gc.ca/evergreen/upgrade-error-changeset-19494.txt and here: http://geoscan.ess.nrcan.gc.ca/evergreen/snip-postgresql-8.4-main.log 2011-02-22T11:50:00 parsr: http://geoscan.ess.nrcan.gc.ca/evergreen/upgrade-error-changeset-19494.txt seems clean 2011-02-22T11:51:28 and grp_perm_map_perm_fkey and friends are outside of the transaction (if you look at the start of the upgrade script you'll see that some errors might be expected) 2011-02-22T11:51:31 so... what failed? 2011-02-22T11:54:15 Some error messages are "normal" 2011-02-22T11:55:19 If you believe the script when it says that 2011-02-22T11:55:20 :) 2011-02-22T11:55:39 So, SIP is being weird to us 2011-02-22T11:56:45 The latest pull of the SIP code seems to give us continuous login errors. 2011-02-22T11:57:05 I went back to an older version, but that makes me wonder if maybe I'm doing something wrong with my oils_sip.xml file 2011-02-22T12:00:35 bshum: were you updating from https://github.com/atz/SIPServer ? 2011-02-22T12:00:50 gmcharlt: Yes, pulled a brand new copy via git 2011-02-22T12:01:19 What I saw this morning was a continuous run of this: 2011-02-22T12:02:18 http://paste.lisp.org/display/119898 2011-02-22T12:02:26 Those two lines just repeated endlessly 2011-02-22T12:02:46 bshum: if you're in a position to do some experimentation, try this 2011-02-22T12:02:54 - do the clone 2011-02-22T12:02:58 gmcharlt: Would be happy to try 2011-02-22T12:03:00 - do git checkout 74ac32f0a4ba08fafe 2011-02-22T12:03:03 - bounce SIP server 2011-02-22T12:03:06 - see if it works 2011-02-22T12:03:12 - git checkout master 2011-02-22T12:03:14 - bounce SIP server 2011-02-22T12:03:18 - see if has broken 2011-02-22T12:05:13 dbs: I'm just about ready to re-run on restored 1.6.1.4 db and will report back with more details.. 2011-02-22T12:07:13 bshum++ #2.0 upgrade 2011-02-22T12:08:09 line-endings-- 2011-02-22T12:10:22 Scotland++ 2011-02-22T12:12:13 Nice, go Scotland go 2011-02-22T12:13:28 so, yeah, i guess no dev meeting today 2011-02-22T12:14:08 Should I put one on the calendar for next week? 2011-02-22T12:18:36 I have no problem with that personally 2011-02-22T12:18:49 * tsbere just got into the office after going on-site to troubleshoot networking 2011-02-22T12:18:54 +1 2011-02-22T12:20:56 dang, been lots of drift between OpenSRF trunk and OpenSRF 2.0 2011-02-22T12:22:03 * dbs ironically knows how to merge multiple changesets in git cleanly but does not know how to do the same in svn 2011-02-22T12:22:59 *** rsinger has quit IRC 2011-02-22T12:24:17 dbs: I believe I know the answer to your issue 2011-02-22T12:24:34 SVN has no idea of multiple local changesets that I know of, so the answer is you don't. 2011-02-22T12:24:40 by "cleanly" I mean without either squashing everything into a single commit, or without having to merge and commit each one one by one and manually adding the commit message 2011-02-22T12:24:46 yeah, that's pretty much what I thought 2011-02-22T12:25:05 Or, you use something like git-svn to do it for you. 2011-02-22T12:26:11 Can we move to a DVCS now? :) 2011-02-22T12:26:35 * tsbere already created a non-svn'd clone of the svn repository in git 2011-02-22T12:28:14 tsbere: i think the question is, who hasn't? 2011-02-22T12:29:32 Near as I can tell there are very few complete clones. I see a lot of trunk clones, some partials. Does the one you made somewhat recently auto-pickup branches and tags from svn? 2011-02-22T12:30:22 tsbere: oh, no, misunderstood you. I just cloned evergreen-equinox because I'm primarily interested in trunk 2011-02-22T12:31:00 If you only care about trunk there are piles of clones. If you care about all of it there are fewer. 2011-02-22T12:31:11 the "plan" for that gitorious repo is to mirror the official git repo whenever that happens. or hg, or bzr, or whatever 2011-02-22T12:31:13 damn 2011-02-22T12:31:31 svn.open-ils.org doesn't support "svn mergeinfo" which would have made things a bit more tolerable 2011-02-22T12:38:02 I am having issues with a clean install of EG 2.01, I can use the admin account log in with srfsh but not with the XUL client 2011-02-22T12:38:09 Perhaps the reason is that I committed the faux pas of re-using the recently retired default admin password (open-ils). 2011-02-22T12:38:27 I ran settings-tester.pl and got a fatal error: 2011-02-22T12:38:35 DBI connect('dbname=evergreen;host=localhost;port=5432','postgres',...) failed: FATAL: password authentication failed for user "postgres" FATAL: password authentication failed for user "postgres" 2011-02-22T12:39:31 I wouldn't normally use the postgres user as the main user 2011-02-22T12:39:54 yboston: sounds like opensrf.xml has the wrong database username / password somewhere 2011-02-22T12:40:30 yboston: or that's a red herring from the unused open-ils.qstore database auth section in opensrf.xml and your real problem is that you need to restart apache 2011-02-22T12:41:57 dbs: thanks, I am checking the file 2011-02-22T12:42:21 * tsbere forgot that the file defaults to that 2011-02-22T12:44:12 gmcharlt: Pulled away to poke at some new issue that came up, but I'll give the commands a whirl in a bit. 2011-02-22T12:44:21 bshum: thanks 2011-02-22T12:44:31 bradl: Thanks :D 2011-02-22T12:57:06 dbs: I looked over opensrf.xml and it is showing the password I expected for db user "evergreen" and which is also the same password for user "postgres" 2011-02-22T12:57:20 though I di not remember explicitly settign the password for user "postgres" 2011-02-22T12:58:13 yboston: the "postgres" user likely can't log into PostgreSQL anyways. Have you tried restarting apache? 2011-02-22T12:58:55 dbs: sorry; forgot to mention that I have cold rebooted the VM a couple of times and have restarted opensrf hten apache each time 2011-02-22T13:00:03 yboston: and you've added the SSL certificate for your server (assuming that it was a signed one) in the staff client using the "Add certificate" button? 2011-02-22T13:01:23 dbs: yes, I created a self signed certificate & then on the client I made an exception 2011-02-22T13:01:58 yboston: anything in /openils/var/log/gateway.log or osrfsys.log? 2011-02-22T13:04:00 dbs: here are the errors I got today http://paste.lisp.org/display/119899 2011-02-22T13:04:30 yboston: Your router isn't there? 2011-02-22T13:06:31 tsbere: my apologies, but I am not sure what you mean. I know there is a ejabber router and I did try to set up the ejabber account and password during install 2011-02-22T13:06:47 The router process is not connected to jabber 2011-02-22T13:07:02 bshum++ 2011-02-22T13:07:08 Bibliomation++ 2011-02-22T13:07:18 yboston: osrf_ctl.sh -a start_router (or -la if you are using localhost block) 2011-02-22T13:07:31 See what that tells you 2011-02-22T13:07:56 tsbere: I will try that 2011-02-22T13:09:43 *** pmplett_ has joined #evergreen 2011-02-22T13:11:15 tsbere: "OpenSRF Router already started : 1387 1388" 2011-02-22T13:11:50 yboston: Try stop_all, then, and see what you are told. (you will want everything down anyway) 2011-02-22T13:14:27 tsbere: nothing unusual was shown http://paste.lisp.org/display/119900 2011-02-22T13:15:05 Then start_perl and start_c (I believe) to get everything back up, then restart apache, and see if it works? 2011-02-22T13:15:56 uh 2011-02-22T13:16:02 you missed start_router 2011-02-22T13:16:09 (after "stop_all") 2011-02-22T13:16:22 His paste included a start_router 2011-02-22T13:17:39 tsbere: oh, sorry - didn't read the paste 2011-02-22T13:19:14 tsbere: it still not working. I don't mind doing another clean install today of EG 2.0 2011-02-22T13:19:38 yboston: really shouldn't be necessary 2011-02-22T13:19:39 Do the logs still spit out router not on the network errors? 2011-02-22T13:20:29 *** elene has joined #evergreen 2011-02-22T13:21:02 berick / gmcharlt: (as the only other recent committers to OpenSRF) I now have everything except for the buildbot example backported from trunk to rel_2_0 2011-02-22T13:21:30 *** jenny has quit IRC 2011-02-22T13:22:40 Any suggestions on things to check for a cataloger login that comes up with a completely borked OPAC view when logged in? 2011-02-22T13:22:48 It's a single login 2011-02-22T13:23:02 Doesn't seem to affect all the other cataloger logins 2011-02-22T13:23:17 I looked in actor.usr_setting and didn't see any additional entries. 2011-02-22T13:23:26 * dbs watches http://testing.evergreen-ils.org/buildbot/waterfall build away 2011-02-22T13:23:49 dear dbs and colleagues, could someone point me how I  can convert db-seed.po file to db-seed sql? 2011-02-22T13:23:59 tsbere: I pasted the latest log "err" entries as an annotation http://paste.lisp.org/display/119900 2011-02-22T13:24:27 i have tried dbs's script db-seed-i18n.py but it seems to expect a POT file rather than po file 2011-02-22T13:24:57 yboston: Doesn't look like any more showed up, then 2011-02-22T13:25:06 Does srfsh still work? 2011-02-22T13:25:54 elene: that should accept PO files for input, as long as you have the polib prereq installed 2011-02-22T13:26:36 thanks dan, yes, i have polib prereq installed, but it generates an empty output file 2011-02-22T13:26:54 so is there any documentation out in the wild on setting up in-db circulation and holds policies? I've looked at the DIG docs and have done cursory searches of the wiki 2011-02-22T13:27:27 (in 1.6.1.X if possible) 2011-02-22T13:27:45 csharp: I had a draft document kicking around on the wiki. The 1.6 labeled portions may be of help. 2011-02-22T13:27:56 tsbere: thanks - I'll search again 2011-02-22T13:28:08 csharp: Or you can click on this link. http://open-ils.org/dokuwiki/doku.php?id=evergreen-admin:customization:indb-circ 2011-02-22T13:28:16 *** jamesrf has quit IRC 2011-02-22T13:28:26 this is the syntax i am using: ./db-seed-i18n.py -s build_i18n_po_db.seed_db.seed-ka.pot -l ka-GE -o test.sql 2011-02-22T13:28:31 tsbere: much appreciated :-) 2011-02-22T13:28:33 elene: "python scripts/db-seed-i18n.py -l hy-AM -o /tmp/foo.sql -s po/db.seed/hy-AM.po" works for me with the Evergreen-2.0 source 2011-02-22T13:28:44 tsbere: I was successful in loging into to srfsh and was able to do the math test 2011-02-22T13:28:51 thanks, i will then try ur syntax 2011-02-22T13:28:55 csharp: Doesn't cover holds, just circ, but better than nothing 2011-02-22T13:29:36 elene: maybe compare build/i18n/po/hy-AM.po to build_i18n_po_db.seed_db.seed-ka.pot to see if there are any obvious formatting differences? 2011-02-22T13:29:37 yboston: By "logging into" do you mean "starting and then doing the math test" or "starting, running the login command, then doing the math test"? 2011-02-22T13:30:11 tsbere: definitely - and it may enlighten holds (assuming consistency in design) 2011-02-22T13:30:23 tsbere: I used login admin xxxxx; then I did the math test 2011-02-22T13:30:42 *** sfortin has quit IRC 2011-02-22T13:32:40 *** jamesrf has joined #evergreen 2011-02-22T13:32:43 hmm. "opensrf.math" is a public service, "open-ils.settings" is a private service. maybe a problem with the private router or opensrf XMPP credentials? 2011-02-22T13:36:36 dbs++ 2011-02-22T13:36:53 # delayed karma bumps FTW! 2011-02-22T13:39:40 dbs: running the "1.6.1-2.0-upgrade-db.sql" now - last thing reported back in my upgrade shell is "CREATE RULE" and top command has been showing postgres at 100% CPU for last 23 minutes. Does this make sense for this particular upgrade script (i.e. time wise)? 2011-02-22T13:41:49 dbs: I looked back at my history and I went over what passwords I used for createing the router and opensrf ejabber accounts; and it matches what is in opensrf_core.xml 2011-02-22T13:44:28 tsbere: I see "Lookups are done by way of a call to action.item_user_circ_test or action.item_user_renew_test" - how would one do this? via srfsh? 2011-02-22T13:44:42 yboston: any passwords with special characters? 2011-02-22T13:44:47 Oh, nvm, broken copy templates, sigh :( 2011-02-22T13:44:47 parsr: sounds reasonable 2011-02-22T13:44:52 csharp: Checkout does it for you normally. 2011-02-22T13:45:07 Beyond that, those are DB functions 2011-02-22T13:47:42 tsbere: okay - I see them now - thanks 2011-02-22T13:48:58 dbs: all passwords created during install are alphanumeric, except I set the admin EG password to the deprecated value of 'open-ils' 2011-02-22T13:49:44 dbs: that's what I wanted to hear. I expect the subsequent in-db re-indexing to take a long time but wasn't expecting the database update script to take so long since they're usually < 5 min. long from my experience. No errors in my postgresql-8.4-main.log yet (other than those performance warnings about "checkpoint_segments" etc.) -- Ok, so maybe I'm in business now. 2011-02-22T13:49:54 *** artunit has quit IRC 2011-02-22T13:50:57 *** mrpeters-isl has joined #evergreen 2011-02-22T13:51:24 *** rsinger has joined #evergreen 2011-02-22T13:51:40 moodaepo/bshum: evertyhing look ok with those A/T printed notices? 2011-02-22T13:52:26 mrpeters-isl: Haven't gotten to poke at it yet. Troubleshooting miscellaneous issues. 2011-02-22T13:52:31 ok 2011-02-22T13:52:41 SIP hates me :) 2011-02-22T13:53:00 whats up? 2011-02-22T13:53:21 Oh I have to test some stuff for gmcharlt 2011-02-22T13:53:35 * gmcharlt cracks the whip ;) 2011-02-22T13:53:37 The latest pull of the SIP code resulted in continuous failure for me in the logs. 2011-02-22T13:53:48 I'm sorry, this one patron was vexing me 2011-02-22T13:54:43 It ended up being a missing quote in that user's copy templates :( 2011-02-22T13:55:21 gmcharlt: I'm not seeing any errors with that new checkout 2011-02-22T13:55:25 But... 2011-02-22T13:55:33 It hangs when attempting to connect 2011-02-22T13:58:04 is SIPServer.pm running? 2011-02-22T13:58:41 Seems to be. There are two running processes when I do ps aux | grep SIP 2011-02-22T14:04:22 bshum: did you ever figure out the borked OPAC view issue? 2011-02-22T14:04:44 (sorry - absorbed here in various things meself) 2011-02-22T14:04:53 *** dbs has quit IRC 2011-02-22T14:05:04 bshum: try bouncing it again 2011-02-22T14:05:36 csharp: Yeah, it was a missing quote in that user's copy template user setting. 2011-02-22T14:05:44 gmcharlt: Did that and it seems to be running much happier now. 2011-02-22T14:05:44 ah 2011-02-22T14:05:59 bshum: cool, time to see if you can break it :) 2011-02-22T14:06:02 csharp: It seems that it pulls all the user settings (not just OPAC ones) when viewing that information. 2011-02-22T14:06:12 gmcharlt: Right!, checking out master and giving it a whirl 2011-02-22T14:06:13 hmm 2011-02-22T14:06:38 bshum: btw, the specific commit that I suspect is the culprit is d797881f3f8e 2011-02-22T14:07:08 gmcharlt: I don't really know how to use git to inspect that sort of information. But sounds quite important ;) 2011-02-22T14:09:07 *** jenny has joined #evergreen 2011-02-22T14:09:51 bshum: you mean you don't memorize random SHA1 hash values? tsk, tsk 2011-02-22T14:10:05 more seriously, https://github.com/atz/SIPServer/commits/master gives a nice browsable view of the changes 2011-02-22T14:10:32 what's the SIP2 device or service in question, by the way? 2011-02-22T14:11:33 Overdrive :) 2011-02-22T14:11:46 But I'm just testing using telnet host 6001 2011-02-22T14:11:51 To see what it's doing 2011-02-22T14:11:53 postgresql_docs++ # an example to aspire too 2011-02-22T14:11:59 s/too/to/ 2011-02-22T14:16:38 Hmm 2011-02-22T14:17:49 Doesn't seem to be complaining anymore 2011-02-22T14:17:54 After switching to master 2011-02-22T14:18:17 Maybe I started/stopped it wrong? 2011-02-22T14:24:13 Yep, I have no explanation. 2011-02-22T14:24:14 So weird. 2011-02-22T14:30:24 Really we have the schedule out as an XLS file? I'll have to complain to afterl later! 2011-02-22T14:35:07 bshum: weird indeed, but beats a regression 2011-02-22T14:35:26 gmcharlt: I'm wondering if maybe I had a bad SIP process running around 2011-02-22T14:37:06 I did end up killing one 2011-02-22T14:37:09 That wouldn't go away 2011-02-22T14:37:57 Guess I'll watch it for awhile and see how it behaves (or not) 2011-02-22T14:38:04 Thanks! gmcharlt++ 2011-02-22T14:38:14 bshum: what do you use for server monitoring? 2011-02-22T14:38:19 Nagios or anything? 2011-02-22T14:38:43 mrpeters-isl: That's on my to-do list between now and May 2011-02-22T14:38:49 reason i ask - we had a library when we first started using SIP that had a buggy piece of software that was opening hundreds of connections per day 2011-02-22T14:38:58 and never closing them 2011-02-22T14:39:08 mrpeters-isl: Went for barebones startup with the multi-server config 2011-02-22T14:39:16 so, you want to monitor to make sure you have a reasonable number of perl SIP processes running at any given day 2011-02-22T14:39:24 its also, best practice, IIRC - to restart SIP nightly 2011-02-22T14:39:46 to clear out any stale processes since your capped to a certain number of processes via the config xml 2011-02-22T14:40:08 just a couple things i ran into, that might be helpful 2011-02-22T14:42:36 mrpeters-isl: Thanks for the tips, they do sound helpful. 2011-02-22T14:49:16 *** parsr has quit IRC 2011-02-22T14:54:35 *** jenny has quit IRC 2011-02-22T15:07:49 *** parsr has joined #evergreen 2011-02-22T15:09:59 UPGRADE: my 1.6.1-2.0-upgrade-db.sql ran succesfully (using latest from 2.0.2); however in the "2.0.0-2.0.1-upgrade-db.sql" I get this error: psql:Open-ILS/src/sql/Pg/2.0.0-2.0.1-upgrade-db.sql:337: ERROR: relation "serial.unit" does not exist ROLLBACK --- Any thoughts -- normal? 2011-02-22T15:10:10 *** dbs has joined #evergreen 2011-02-22T15:10:11 *** dbs has joined #evergreen 2011-02-22T15:12:27 *** kmlussier has quit IRC 2011-02-22T15:18:35 parsr: sounds like your 1.6.1 -> 2.0 upgrade did not actually succeed, then 2011-02-22T15:19:11 dbs: so ... in-db unapi (aka unapi2 topic branch) ... you have thoughts on immediate integration? 2011-02-22T15:19:57 fwiw, and not to put words in his mouth, I believe berick is behind merging now-ish 2011-02-22T15:21:21 "CREATE TABLE serial.unit ..." is one of the steps in the 1.6.1->2.0 upgrade 2011-02-22T15:21:32 eeevil: haven't had a chance to look at it yet, was trying to figure out the Armenian regression 2011-02-22T15:23:13 dbs: it worked :) turns out I was using an old version of the script from http://bzr.coffeecode.net/ and also I used po files exported from Launchpad and thus all the extra comments were stripped. 2011-02-22T15:24:14 maybe we could add it to the agenda for the next dev meeting? 2011-02-22T15:24:14 * dbs will peek then 2011-02-22T15:24:37 dbs: I'd be happy with a code review, but the high-level version is that it will deprecate (much of) the feed generation of O::A::SuperCat and consolidate all xml tranformation of bib records and holdings xml generation into one place ... it's also the basis for some big speed gains 2011-02-22T15:25:03 and a pile of other stuff gets easier once it's in (both maintenance and features) 2011-02-22T15:25:18 dbs: my pg log of upgrade here: http://geoscan.ess.nrcan.gc.ca/evergreen/postgresql-8.4-main.log -- "current transaction is aborted, commands ignored until end of transaction block" 2011-02-22T15:25:32 (more looking for a "don't do that" than a tacit approval) 2011-02-22T15:26:03 dbs: aslo, great work on the armenian regression! 2011-02-22T15:26:23 "the armenian regression" sounds like a spy novel 2011-02-22T15:26:29 gah! /me needs to kill those old bzr clones 2011-02-22T15:37:36 eeevil: so... two patches if I'm reading git format-patch correctly? 2011-02-22T15:37:41 setting the stage for SuperCat feed deprecation, but not actually touching O:A:SuperCat yet, yes? 2011-02-22T15:37:42 elene: btw, you would be interested in this - I'm going to put together an updated OpenSRF 2.0 alpha release so that we can support non-ASCII content; OpenSRF 1.6.2 appears to only support ASCII + the basic multilingual plane at the moment 2011-02-22T15:38:04 parsr: so your problem was at "CREATE TEXT SEARCH CONFIGURATION identifier ( COPY = title );" it looks like? 2011-02-22T15:38:06 that seems unlikely to exist in a stock 1.6.1 database - this was a completely clean database that you loaded your 1.6.1 dump into? 2011-02-22T15:39:26 dbs: nope not clean - this was our production system dumped and re-loaded into dev server for upgrade test; could it be related to some of our custom indexes we have setup for us?? 2011-02-22T15:39:58 dbs: or what do you mean by "clean" 2011-02-22T15:41:49 *** Meliss has quit IRC 2011-02-22T15:41:50 *** _dkyle_1 has quit IRC 2011-02-22T15:41:50 *** shadowspar has quit IRC 2011-02-22T15:42:04 *** jenny1 has joined #evergreen 2011-02-22T15:44:47 parsr: quite possibly - might be a job for your support co. to check out 2011-02-22T15:45:15 agreed -- just opening ticket now. Thanks for the pointers. Getting closer... 2011-02-22T15:48:37 dbs: sorry, yes. stored procs and a new tag URI parser module (and new non-arbitrary tag URI format). integration comes later 2011-02-22T15:48:47 *** collum has left #evergreen 2011-02-22T15:50:37 eeevil: nothing horrifying leaps out at me, I say go for it :) 2011-02-22T15:50:53 eeevil: although "teste queries" makes me a bit uncomfortable 2011-02-22T15:51:51 dbs: that's gr8 news. will it work with EG 2.0.1? do you have opensrf messages for l10n? 2011-02-22T15:53:37 dbs: I'll squash that :) (ow?) 2011-02-22T15:54:42 eeevil: ouch! 2011-02-22T16:00:04 *** Meliss has joined #evergreen 2011-02-22T16:00:04 *** _dkyle_1 has joined #evergreen 2011-02-22T16:00:04 *** shadowspar has joined #evergreen 2011-02-22T16:00:11 elene: yes, OpenSRF 2.0 will work with EG 2.0.x - it's what I run during development and everything seems okay, but of course there might be some other bugs that we need to deal with 2011-02-22T16:01:12 dbs: that't great. i will keep an eye on svn for the po files 2011-02-22T16:01:42 elene: oh, and there are no translatable messages that I'm aware of in OpenSRF 2011-02-22T16:02:14 *** Meliss has quit IRC 2011-02-22T16:02:16 dbs: i see, so it's merely for transporting doublebyte chars, right? 2011-02-22T16:04:31 elene: OpenSRF is all infrastructure, nothing that users actually see or touch, thus no messages 2011-02-22T16:04:36 hmm. http://vagrantup.com/ looks like a possible way of freeing up some time (automated generation of VirtualBox images) 2011-02-22T16:05:19 *** dbwells has quit IRC 2011-02-22T16:05:19 *** wjr has quit IRC 2011-02-22T16:05:19 *** kbeswick has quit IRC 2011-02-22T16:05:19 *** jamesrf has quit IRC 2011-02-22T16:05:19 *** AaronZ-PLS has quit IRC 2011-02-22T16:05:20 *** bencahill has quit IRC 2011-02-22T16:05:20 *** jenny1 has quit IRC 2011-02-22T16:05:20 *** mrpeters-isl has quit IRC 2011-02-22T16:05:21 *** bshum has quit IRC 2011-02-22T16:05:25 *** pmplett_ has quit IRC 2011-02-22T16:05:26 *** brendan2 has quit IRC 2011-02-22T16:05:26 *** finnapz has quit IRC 2011-02-22T16:05:26 *** tsbere has quit IRC 2011-02-22T16:05:27 *** pmplett has quit IRC 2011-02-22T16:05:28 *** leed has quit IRC 2011-02-22T16:05:29 *** eby has quit IRC 2011-02-22T16:05:30 *** mjgiarlo has quit IRC 2011-02-22T16:05:30 *** rangi has quit IRC 2011-02-22T16:05:30 *** eeevil has quit IRC 2011-02-22T16:05:30 *** jeff has quit IRC 2011-02-22T16:05:30 *** wjr_ has joined #evergreen 2011-02-22T16:05:50 *** dbwells has joined #evergreen 2011-02-22T16:06:07 dbs: I think I found very few messages at http://svn.open-ils.org/trac/ILS/browser/branches/rel_2_1/Open-ILS/src/templates/strings/ 2011-02-22T16:06:25 *** kbeswick has joined #evergreen 2011-02-22T16:06:25 *** jenny1 has joined #evergreen 2011-02-22T16:06:25 *** mrpeters-isl has joined #evergreen 2011-02-22T16:06:25 *** jamesrf has joined #evergreen 2011-02-22T16:06:25 *** bshum has joined #evergreen 2011-02-22T16:06:25 *** AaronZ-PLS has joined #evergreen 2011-02-22T16:06:25 *** bencahill has joined #evergreen 2011-02-22T16:06:27 elene: that's Evergreen, not Opensrf 2011-02-22T16:06:42 OpenSRF is under http://svn.open-ils.org/trac/OpenSRF 2011-02-22T16:06:52 dbs: right :) 2011-02-22T16:06:58 *** leed has joined #evergreen 2011-02-22T16:06:58 *** eby has joined #evergreen 2011-02-22T16:06:58 *** mjgiarlo has joined #evergreen 2011-02-22T16:06:58 *** rangi has joined #evergreen 2011-02-22T16:06:58 *** eeevil has joined #evergreen 2011-02-22T16:06:58 *** jeff has joined #evergreen 2011-02-22T16:07:07 elene: oh, unless you're talking about strings that have not yet been added to the PO infrastructure - which is what those are 2011-02-22T16:07:41 *** pmplett_ has joined #evergreen 2011-02-22T16:07:41 *** brendan2 has joined #evergreen 2011-02-22T16:07:41 *** finnapz has joined #evergreen 2011-02-22T16:07:41 *** tsbere has joined #evergreen 2011-02-22T16:07:41 *** pmplett has joined #evergreen 2011-02-22T16:18:21 *** parsr has quit IRC 2011-02-22T16:21:19 *** rjackson-isl has quit IRC 2011-02-22T16:36:08 grabbing 0489 2011-02-22T16:49:04 *** dbs has quit IRC 2011-02-22T16:49:04 *** _bott_ has quit IRC 2011-02-22T16:49:06 *** _dkyle_1 has quit IRC 2011-02-22T16:49:07 *** shadowspar has quit IRC 2011-02-22T16:50:02 *** shadowspar has joined #evergreen 2011-02-22T16:50:02 *** _dkyle_1 has joined #evergreen 2011-02-22T16:50:02 *** dbs has joined #evergreen 2011-02-22T16:50:02 *** _bott_ has joined #evergreen 2011-02-22T17:06:26 *** vickir has joined #evergreen 2011-02-22T17:09:40 We are having a problem with registering workstations. Permission on staff is set to allow workstation registration. Circulators are a group under staff. When circulators try to register a workstation we get the message that they do not have permission. Does anyone know how to fix this? 2011-02-22T17:15:43 vickir: Let me check our circ staff settings, one sec 2011-02-22T17:16:44 my thanks to dbs and tsbere; I figured out what was wrong. I was using a saved copy of opensrf_core.xml with the correct passwords, but the version I used when installing OSRF 2011-02-22T17:18:09 vickir: moodaepo: Maybe they need an additional permission. Does the error say anything about which is missing? "REGISTER_WORKSTATION" is one 2011-02-22T17:18:18 But maybe it's "VIEW_ORG_UNITS" 2011-02-22T17:18:21 Or some other such 2011-02-22T17:18:29 To generate the dropdown selector 2011-02-22T17:18:52 The permission error usually highlights the missing permission in red 2011-02-22T17:18:57 If I remember correctly 2011-02-22T17:20:10 Just a one line error message "you don't have permission to register a workstation" with a yellow triangle. Only admin can register 2011-02-22T17:22:08 *** dave-esi has quit IRC 2011-02-22T17:22:29 vickir: What about group_application.user.staff.admin.lib_manager 2011-02-22T17:22:29 In our setup only the branch manager can register and that user has that assigned 2011-02-22T17:23:24 *** yboston has quit IRC 2011-02-22T17:23:37 we're checking 2011-02-22T17:24:06 moodaepo: group_application.user.x are permissions for adding, editing, etc. of users associated with that permission. 2011-02-22T17:24:41 I don't think they're related to workstation 2011-02-22T17:25:07 bshum: I didn't check the documentation : ) 2011-02-22T17:25:29 moodaepo: We use that permission alot to control specific staff groups from registering specific patron groups. 2011-02-22T17:25:37 Split the public side from school side 2011-02-22T17:26:26 Maybe "VIEW_ORG_SETTINGS" 2011-02-22T17:26:35 We have that assigned to our staff group at the system level 2011-02-22T17:27:08 No, wait, that's also probably unrelated... 2011-02-22T17:27:15 Bah sorry vickir 2011-02-22T17:27:26 This is a good one to tail log files if that's possible, the logs clearly say why it was denied (more or less) 2011-02-22T17:28:04 Don't apologize - all clues are appreciated. We're checking the log files. Thank you. 2011-02-22T17:31:43 The errorlog tells us to see errorlog for more details. Still chasing that rabbit. 2011-02-22T17:33:07 vickir: Could you paste the error log section here > http://paste.lisp.org/new/evergreen 2011-02-22T17:34:48 we're trying to figure out which log we want to look in - obviously the one that told us to see the errorlog for details was not the right one. We'll paste as soon as we find the right log. Thanks 2011-02-22T17:36:48 vickir: I would look in osrfsys.log and the gateway.log 2011-02-22T17:40:05 We ran the command and watched to see what it spewed out. I'll go see if I can find the logs you mentioned. 2011-02-22T17:40:47 Looking in /var/log - I don't find osrfsys.log or gateway.log 2011-02-22T17:40:59 /openils/var/log 2011-02-22T17:41:24 vickir: ^^ 2011-02-22T17:42:06 Thank you. I don't see anything in those logs that definitely fits with the error but I'm looking. 2011-02-22T17:44:58 I just sent the lines from gateway.log 2011-02-22T17:51:12 vickir: Oh lisp paste bot might be down, could you paste the link? 2011-02-22T17:51:57 *** bshum has quit IRC 2011-02-22T17:52:36 http://paste.lisp.org/new/evergreen is the link I pasted the info into - is that what you need? 2011-02-22T17:54:51 vickir: do your staff have "working locations" assigned to their accounts (in their user accounts > other > user permission editor)? 2011-02-22T17:55:15 looking 2011-02-22T17:56:11 one of our challenges is that our staff needs to be able to move from branch to branch so no one can have a permanently assigned workstation 2011-02-22T17:56:51 vickir: After you pasted you should have gotten a slightly different link. 2011-02-22T17:56:53 the staff member who was trying to register when we got this error is assigned to org unit DER 2011-02-22T17:57:17 we have ours set up to be assigned to all locations; workstations may be used by any staff assigned to the working location. 2011-02-22T17:57:20 the window closed itself 2011-02-22T17:58:14 vickir: Hah, might have re-paste 2011-02-22T17:58:20 K 2011-02-22T17:59:46 vickir: Is the org unit only assigned to the user account? Under "other > user permission editor" you can choose to give staff/users permission to work from workstations registered to various orgs by selecting the boxes next to each org unit. Assigned orgs in the user account are separate from assigned working locations 2011-02-22T18:00:35 *** pmplett_ is now known as pmpafk 2011-02-22T18:01:09 I repasted and clicked 'submit'. Browser window is blank and the url is http://paste.lisp.org/submit 2011-02-22T18:02:24 Do you see a link on the page 2011-02-22T18:02:58 Never mind 2011-02-22T18:03:11 vickir: the service seems to have gone down : ) 2011-02-22T18:04:14 k - we're looking at the various permissions again. Being a single planting leaves one feeling rather alone-ish. Thank you all for the suggestions. 2011-02-22T18:04:23 Also vickir you need to search for the user account and then go to the user profile 2011-02-22T18:04:38 that's where jenny1 wants you to check for the assigned workstations. 2011-02-22T18:05:54 the user profile is set to circulators. circulators should be inheriting permissions from their group which is staff. staff specifically has permission to register workstations. 2011-02-22T18:06:09 vickir: If you still are in the mood to paste > http://pastebin.com/ 2011-02-22T18:07:10 Yea in the user profile can you click on the dropdown "Other" 2011-02-22T18:08:02 Pasted and sent 2011-02-22T18:08:41 fwiw, you can see the paste URLs at http://paste.lisp.org 2011-02-22T18:08:58 vickir: What was the resulting page url? It should similar to http://pastebin.com/pKcZ2Q9A 2011-02-22T18:09:01 and I think jenny1 nailed it - the staff probably need to have working locations assigned 2011-02-22T18:09:06 dbs: It seemed to be down 2011-02-22T18:09:20 http://pastebin.com/TAUPikfs 2011-02-22T18:10:46 moodaepo: I know it seems to be down, but when you submit it generally submits. their redirect that normally shows you the result just doesn't seem to be working. http://paste.lisp.org shows attempts by vickir from 15 minutes ago 2011-02-22T18:11:29 dbs: Do'h I hadn't noticed that, thanks. 2011-02-22T18:11:50 In the profile, the staff user has multiple pages and we don't see any dropdowns other than the one for the group which is set to circulators 2011-02-22T18:11:50 moodaepo: s'ok, I only know from bitter experience :) 2011-02-22T18:13:46 vickir: no, search for the staff member in the "Search for patron" UI, retrieve the staff member, and go into "Other -> User permission editor" 2011-02-22T18:13:53 we just set that user to be the system administrator and he still cannot register a workstation. The only user we have who can register a workstation is the user admin. If we can't figure out how to fix this we'll have to drive around to all the branches to register the workstations. 2011-02-22T18:14:28 k looking in search paron - thanks 2011-02-22T18:15:05 if this isn't documented in the official documentation, it really should be. 2011-02-22T18:16:12 we did not have a working location assigned to the staff member. If we assign him to the highest level, will he be able to register at the locations under that one? 2011-02-22T18:17:13 vickir: Yes I believe so 2011-02-22T18:17:14 http://docs.evergreen-ils.org/1.6/draft/html/permissions.html sort of covers this 2011-02-22T18:17:18 moodaepo: vickir: no 2011-02-22T18:17:21 ok, can anyone tell us where this gets stored in the database so we can assign it on the way in on dump day? 2011-02-22T18:17:39 dbs: so each location needs to be assigned? 2011-02-22T18:17:55 vickir: permission.usr_work_ou_map 2011-02-22T18:17:56 moodaepo, 2011-02-22T18:18:12 moodaepo: yes, I believe this is a case where hierarchy does not apply 2011-02-22T18:18:22 thank you for the link - I don't think we had that one. 2011-02-22T18:18:30 jenny1 setup our stuff and I am hazarding a guess 2011-02-22T18:19:06 So vickir as dbs says you will need to add all the locations for each staff member since you said they move around 2011-02-22T18:19:17 we just found where it goes in the database so I think we are good to go 2011-02-22T18:20:11 thank you - we think we have it now. 2011-02-22T18:20:52 vickir: Just curious which library are you working with on this? 2011-02-22T18:25:39 We're Beauregard Parish Library in DeRidder, Louisiana. If (and it's beginning to feel like a big IF) we manage to pull this off I think we'll be the first library in Louisiana to be up on Evergreen. 2011-02-22T18:26:41 *** vickir has left #evergreen 2011-02-22T18:29:57 * dbs wonders - "just found where it goes in the database" via my entry from two minutes before? 2011-02-22T18:32:02 *** jenny has joined #evergreen 2011-02-22T18:33:27 *** jenny1 has quit IRC 2011-02-22T18:55:57 I am looking for a way to change OPAC default interface language 2011-02-22T18:56:10 and skin as well 2011-02-22T19:06:17 elene: we use HTTP redirects to do that in the apache configuration 2011-02-22T19:07:09 dbs: thank you. does this mean I have to look in apache.conf and find the relevant section there? 2011-02-22T19:07:31 elene: in eg_vhost.conf, for example, we have RewriteCond %{HTTP_HOST} ^ldcr.concat\.ca 2011-02-22T19:07:31 RewriteRule ^/?$ http://ldcr.concat.ca/opac/en-CA/skin/default/xml/index.xml?l=132 [R=permanent,L] 2011-02-22T19:08:13 so it's easy enough to replace 'en-CA' with 'ka-GE' and 'default' with 'myskin' 2011-02-22T19:08:56 dbs: great. so this RewriteRule goes to apache.conf file, right? 2011-02-22T19:09:42 elene: we have it in /etc/apache2/eg_vhost.conf 2011-02-22T19:10:28 dbs: oh yes, how stupid of me. you already mentioned it. thank you so much :) 2011-02-22T19:17:05 no problem! 2011-02-22T19:29:43 *** elene has quit IRC 2011-02-22T20:21:51 * dbs peers at dojo's attempt to load NLS content from /js/dojo/openils/opac/nls/fr-ca/opac.js 2011-02-22T20:22:11 I thought it was supposed to pull from fr/ca/opac.js. hmm 2011-02-22T20:24:09 http://dojotoolkit.org/reference-guide/dojo/requireLocalization.html says fr-ca. argh. 2011-02-22T20:30:29 dang old bugs.launchpad.net/evergreen doesn't have up-to-date milestones. grumble grumble 2011-02-22T20:31:23 dbs: I know, without a doubt, I've seen en/us/blah.js and /not/ en-us/blah.js 404s since forever ... not sure what's going on there 2011-02-22T20:31:47 eeevil: well, thanks for making me feel like I'm not totally crazy 2011-02-22T20:32:23 chromium dev tools tells me en-ca/opac.js for https://laurentian.concat.ca/opac/en-CA/skin/lul/xml/myopac.xml?l=105&d=1 2011-02-22T20:33:19 I wonder if the semantics of that changed between 1.1 (what we were originally shipping) and 1.3? 2011-02-22T20:33:51 hmm. mebbe 2011-02-22T20:33:52 I think the "since V1.0" just refers to the existence of dojo.requireLocalization 2011-02-22T20:34:02 biab 2011-02-22T20:38:35 * dbs runs "make install_all_locales" with a slight tweak of build/i18n/Makefile 2011-02-22T20:57:59 *** tater-laptop has joined #evergreen 2011-02-22T20:58:14 *** vickir has joined #evergreen 2011-02-22T20:59:31 *** vickir has left #evergreen 2011-02-22T21:18:52 eeevil: so any thoughts on SRU for authority records? It's simplistic, but it can be augmented if/when the underlying auth search support gets richer 2011-02-22T21:19:09 dbs: I'm for it 2011-02-22T21:19:41 right-o 2011-02-22T21:19:53 also, r19506 was my first dcommit ... whee 2011-02-22T21:24:40 dcommit stuff looks like a special kind of pain 2011-02-22T21:24:49 it ... was 2011-02-22T21:25:08 monstly just figuring out the merge/rebase dance 2011-02-22T21:25:19 and waiting for git svn fetch to finish 2011-02-22T21:26:06 * dbs looking at http://stackoverflow.com/posts/4238528/revisions as an example workflow 2011-02-22T21:26:49 make sure that you branch tracking branches, though! 2011-02-22T21:27:22 if you want to dcommit into anything but trunk 2011-02-22T21:27:43 * dbs will probably just format-patch and apply it all as a single patch via ye olde svn for now 2011-02-22T21:29:24 hrm... that's more complicated than what I did, fwiw 2011-02-22T21:31:08 you think so? 2011-02-22T21:32:00 yep ... I'll wiki-fy the basics of what I've got going ... I'm keeping the svn stuff separate to reduce conceptual impedence mismatch 2011-02-22T21:34:12 though the --no-ff looks like a nice replacement for rebase -i 2011-02-22T21:34:24 if I'm understanding that correctly 2011-02-22T21:46:24 eeevil: my dumbass workflow = 'git diff master dbs/enhance_sru > /tmp/sru_auth.patch; cd /path/to/svn/trunk; patch -p1 < /tmp/sru_auth.patch; svn ci' 2011-02-22T21:47:12 dbs: I came this >< close to that too, but I'm /really/ trying to give git a chance ;) 2011-02-22T21:48:13 eeevil: heh, I don't think having SVN in the mix is a fair test for git - I think that adds an extra layer of wtfOMG!?! 2011-02-22T21:54:01 dbs: well, with my workflow it's completely outside git except for the last dcommit bit, which I'm /only/ doing to trunk. then I merge in svn between branches 2011-02-22T21:54:49 *** rickd_ has quit IRC 2011-02-22T21:55:04 aye, I'll be doing that too. "svn merge -c ### svn://foo/trunk; svn log -c ### svn://foo/trunk; svn ci" (copying and pasting the log) is pretty straightforward 2011-02-22T21:55:37 straightforward git-to-git merges seem a heck of a lot easier than svn-to-git, for sure 2011-02-22T21:56:25 git-to-git is still much more complicated than svn-to-svn, though 2011-02-22T21:58:36 I dunno. Merging OpenSRF trunk to rel_2_0 was painful - having "mergeinfo" available would have helped, but still seems to be a ton of manual effort if you want to avoid flattening umpteen commits into a single commit 2011-02-22T21:59:26 Okay. sru_auth has been merged 2011-02-22T22:00:34 yeah... with enough svn log shell pipes you could probably reuse the commit message without copy/paste ... but I usually squash backports anyway, if there's more than one commit 2011-02-22T22:05:10 one thing to pick your brain about; I ended up using open-ils.cstore.direct.authority.record_entry.retrieve to pull each record individually, because the retrieve_record_objects() method in O:A:SuperCat returns records in arbitrary db order, not list order 2011-02-22T22:06:32 that's not a big deal for bibs where relevancy ranking is a bit fuzzy anyway, but for auth records where alphabetical browse is our only interface, it's pretty noticeable 2011-02-22T22:07:30 so... cstore direct generates a ton of INFO messages - anything more consequential to worry about with that approach? 2011-02-22T22:09:43 it's a lot of back-and-forth ... are there log messages that concern you? 2011-02-22T22:10:43 eeevil: no. the 10 calls instead of 1 call bothered me, but I'm not sure there's a better solution 2011-02-22T22:10:59 dbs: unapi2 ;) 2011-02-22T22:11:21 hey, sru_auth could be a good testcase 2011-02-22T22:12:24 (srsly ... adding unapi.are() and are-feed-builder stored procs and calling json_query on the feed builder would be rad) 2011-02-22T22:13:17 sounds like a good todo 2011-02-22T22:26:08 *** tater-laptop has quit IRC 2011-02-22T22:33:09 back to the insulation mines. I've been one hour away from being done for days, it seems... 2011-02-22T22:43:15 *** dbs has quit IRC 2011-02-22T23:02:42 *** tsbere has quit IRC 2011-02-22T23:02:47 *** tsbere has joined #evergreen 2011-02-22T23:03:37 *** pmpafk has quit IRC 2011-02-22T23:07:38 *** jamesrf has quit IRC 2011-02-22T23:29:04 *** pmpafk has joined #evergreen 2011-02-22T23:35:41 *** pmpafk has quit IRC