| # | Time | Nick | Message |
|---|
| # | 01:11:40 | mck9 has left #evergreen |
| # | 01:45:09 | jamesrf has quit IRC |
| # | 05:16:54 | SeVIIeN has joined #evergreen |
| # | 07:00:12 | SeVIIeN has quit IRC |
| # | 07:40:22 | sfortin has joined #evergreen |
| # | 07:56:14 | mck9 has joined #evergreen |
| # | 08:55:56 | bshum has joined #evergreen |
| # | 09:11:35 | jenny has joined #evergreen |
| # | 10:39:49 | mrpeters-isl has quit IRC |
| # | 10:39:56 | mrpeters-isl has joined #evergreen |
| # | 10:59:06 | dbs has joined #evergreen |
| # | 11:45:40 | r123 has joined #evergreen |
| # | 11:49:21 | bshum | dbs: Question for you, in your description of active vs. passive trigger events (in your first e-mail about it) you mentioned how passive events are handled. But what about active ones? Does something occur to have these events run? |
| # | 11:50:14 | natschil has joined #evergreen |
| # | 11:52:46 | jenny has left #evergreen |
| # | 11:57:31 | dchristens has joined #evergreen |
| # | 12:02:21 | dchristens | Hi folks. Quick question: are there any issues/problems with running the staff client on Windows 7? (I don't have a Windows 7 desktop to test on) |
| # | 12:04:05 | dbs | dchristens: I had to install it outside of Program Files directory, but otherwise it worked for me a few weeks back |
| # | 12:04:22 | dbs | bshum: uhhh |
| # | 12:04:28 | jamesrf has joined #evergreen |
| # | 12:04:48 | dchristens | Thanks, dbs. :-) |
| # | 12:05:20 | bshum | dbs: Heh |
| # | 12:09:36 | bshum | dbs: I only asked my question because I'm looking at our current configuration for notification/action triggers and I think we were under the (apparently) mistaken impression that the active field being true or false meant that it was a live rule or not. Rather than active vs. passive triggers. |
| # | 12:10:52 | dbs | bshum: yeah, that much I can guarantee - it's active vs. passive, not enabled vs. disabled |
| # | 12:11:19 | bshum | dbs: Neat, thanks :) |
| # | 12:11:37 | dbs | I think you still need to run action_trigger_runner.pl --run-pending to fire the active events |
| # | 12:12:03 | dbs | and you need to run action_trigger_runner.pl --process-hooks the generate the passive events, then --run-pending |
| # | 12:12:19 | dbs | but who am I to talk, given that I haven't got that to work _at all_. |
| # | 12:13:13 | dbs | well, one exception: I can use the "Test" widget to generate an overdue notice. Simple! I can just GUI-script something to take a list of overdue barcodes and enter each one individually in the Test interface |
| # | 12:13:29 | bshum | Haha |
| # | 12:14:12 | dbs | bshum++ # thanks for the response on the list, good to know that it's working somewhere for somebody |
| # | 12:14:27 | bshum | dbs: I think it's semi-working for us |
| # | 12:14:42 | bshum | I don't think our active events are working right. |
| # | 12:14:50 | bshum | But it seems like the passive ones are running at least. |
| # | 12:22:07 | dbs | bshum: heh - are you sure that the notices aren't the old-school notices? |
| # | 12:22:59 | dbs | (as generated by generate_circ_notices.pl, would likely be in your crontab) |
| # | 12:24:42 | natschil has quit IRC |
| # | 12:25:48 | natschil has joined #evergreen |
| # | 12:27:44 | natschil has quit IRC |
| # | 12:30:04 | natschil has joined #evergreen |
| # | 12:31:33 | bshum | dbs: hmm, fairly sure |
| # | 12:31:37 | bshum | We do have both running |
| # | 12:31:44 | bshum | But they're generated at different times |
| # | 12:32:51 | bshum | 2 am notices are coming from our action/trigger events, and 7:05 are legacy notices |
| # | 12:33:01 | dbs | bshum: yeah, from re-reading your email it sounds like it would have to be action/trigger for the pre-due notifications |
| # | 12:33:25 | dbs | which also answers one of my questions - negative intervals can work |
| # | 12:33:38 | bshum | Right |
| # | 12:34:06 | bshum | My problem is that we have two pre-due notices and it seems that one of them was marked true and the other one false. |
| # | 12:34:26 | bshum | And I'm not sure if the predues we're seeing are coming from the active trigger or passive triggers |
| # | 12:34:38 | dbs | AWESOME :) |
| # | 12:34:46 | bshum | And if it's the case that it's coming from passive, then it seems that the owning library distinction is meaningless |
| # | 12:34:56 | bshum | Because the owning libraries are completely different |
| # | 12:35:12 | bshum | Going back to the beginning of your e-mail asking how owning libraries operate |
| # | 12:36:35 | dbs | ugh. that would suck if the standard inheritance/override approach didn't work |
| # | 12:37:32 | miker_ | all circulation.due events are passive (the hook is passive) |
| # | 12:38:22 | dbs | answers! manna from heaven! |
| # | 12:39:10 | dbs | ah - so "Active" is actually "Enabled" then? |
| # | 12:39:15 | berick | dbs: if you mean event def at branch A overrides event def at consortium, no, it doesn't do that. |
| # | 12:39:25 | berick | dbs: at the event def level, active == enabled |
| # | 12:39:44 | dbs | "Active" is probably a poor choice of term in the context of this interface :) |
| # | 12:39:56 | berick | probably |
| # | 12:40:55 | dbs | so, given disabled events at the consortial level, then, we could clone and enable events at system levels to get different processing delays |
| # | 12:41:14 | berick | yes, and different templates, etc. |
| # | 12:41:30 | dbs | but if one of your branches wants a different delay, then all of your branches have to have their own events |
| # | 12:41:36 | berick | correct |
| # | 12:41:37 | dbs | (within a given system) |
| # | 12:41:42 | dbs | Okay. It all comes together. |
| # | 12:42:36 | dbs goes to set the checkout.due / hold.available events to "enabled" and give the action_trigger_runner scripts a run again |
| # | 12:43:14 | berick | dbs: a good test technique is to use the ProcessTemplate reactor. the output will be shoved in the db |
| # | 12:49:41 | dbs | yay, sweet success... email has been generated |
| # | 12:50:06 | atz | woot |
| # | 12:50:29 | bshum | Yay |
| # | 12:50:30 | dbs | I'll follow up with an edited version of email with more truthiness |
| # | 12:51:16 | dbs | berick: any objections to my changing "Active" to "Enabled" in the various places where "Active" means "Enabled" and not "Active"-as-opposed-to-passive? |
| # | 12:51:56 | berick | dbs: no objections, thanks |
| # | 13:09:00 | BobW has joined #evergreen |
| # | 13:16:24 | dbs | Hmm. Can I pass in a variable for an event parameter, such as sender_email = %target_copy.circ_lib.email%, or does the value have to be a hardcoded string? |
| # | 13:18:08 | berick | dbs: it's eval'd. but, you could just do that directly in the template |
| # | 13:18:43 | berick | From: [% target.0.target_copy.circ_lib.email %] # assuming a grouped target list |
| # | 13:18:54 | dbs | That would make more sense, indeed. |
| # | 13:19:00 | mrpeters-isl | all, is there some sort of timeout value where a search will just "give up" |
| # | 13:19:57 | phasefx | no, you get the green knight rider bar of death forever :-/ |
| # | 13:20:07 | dbs | phasefx: that is _so_ 1.2 |
| # | 13:20:14 | dbs | spinny circles |
| # | 13:20:15 | mrpeters-isl | i have a server that simply will not give a keyword result |
| # | 13:20:29 | mrpeters-isl | ive done dbs "cache heating" to no avail |
| # | 13:20:30 | phasefx | ah, death spirals.. improvement |
| # | 13:20:44 | mrpeters-isl | searches by barcode are all good |
| # | 13:20:47 | mrpeters-isl | or tcn, etc. |
| # | 13:21:02 | dbs | sounds like records that haven't been properly ingested. |
| # | 13:21:25 | dbs | did you check metabib.keyword_field_entry to see if there are any entries that match the record for which you're searching? |
| # | 13:23:07 | dbs considers keeping notification emails firmly in the grasp of system administrators - given that reports are still a major stretch for most of our library admin types, action/trigger stuff just seems dangerous |
| # | 13:23:48 | mrpeters-isl | 3398900 rows in metabib.keyword_field_entry, and yes...a match for what i searched |
| # | 13:25:15 | dbs | mrpeters-isl: and you have metarecords? |
| # | 13:25:42 | mrpeters-isl | i would assume so, this is a restore from a good database |
| # | 13:25:56 | mrpeters-isl | \q |
| # | 13:26:03 | mrpeters-isl | oops |
| # | 13:26:47 | dbs | okay - a match for the record that you're searching for in metabib.metarecord would be a good test |
| # | 13:27:23 | dbs | your database server had all the Perl modules, etc, required by the database functions installed when you restored? |
| # | 13:33:38 | mrpeters-isl | yeah, tsearch, all that good stuff |
| # | 13:33:52 | dbs | JSON::XS? |
| # | 13:33:56 | mrpeters-isl | its a strange thing....every once in a while, i'll get results in about 30 seconds |
| # | 13:34:06 | mrpeters-isl | but most of the time, it spins infinently |
| # | 13:34:12 | dbs | mrpeters-isl: have you run "analyze"? |
| # | 13:34:30 | bshum | Does anyone know what the current timeout for the staff client is? If any? And if it's possible to increase the time before the staff client requires you to log in again. |
| # | 13:34:35 | mrpeters-isl | dbs: yep, that was one of my first thoughts...i ran a VACCUM FULL ANALYZE VERBOSE |
| # | 13:34:54 | dbs | bshum: yeah, there are library settings for that |
| # | 13:35:08 | mrpeters-isl | its 600 sec by default, iirc |
| # | 13:35:23 | bshum | dbs: Right in number of seconds. But if we wanted something for a matter of hours, would we multiply out an insane number of seconds? |
| # | 13:35:30 | dbs | bshum: yep |
| # | 13:35:49 | bshum | dbs, mrpeters-isl: Thanks, off to tinker! |
| # | 13:36:28 | dbs | bshum: ours is set to 32000 |
| # | 13:39:29 | dbs | hrm, looks like we need to set VIEW permissions and the like for LSAs for standing penalties / group penalty thresholds |
| # | 13:46:10 | BobW has quit IRC |
| # | 13:46:42 | mrpeters-isl | anyone else have any thoughts on this search deal? |
| # | 13:47:06 | mrpeters-isl | oh, dbs i forgot to answer your question |
| # | 13:48:16 | mrpeters-isl | searching for "william henry harrison" i get the infinite spin, but i do have (at least) 1 entry in metabib.keyworld_field_entry for that search |
| # | 13:48:50 | mrpeters-isl | and, a matching MARC in biblio.record_entry |
| # | 13:50:49 | Callender has joined #evergreen |
| # | 13:51:24 | mrpeters-isl | not sure if this is relevant, but i tried in IE for the heck of it, and i get an error " 'count' is null or not an object on rresult.js" |
| # | 13:52:07 | jeff | i believe that's a symptom of "the xmlhttp request timed out before returning any JSON" |
| # | 13:52:20 | mrpeters-isl | 10-4, just thought i'd throw it out there |
| # | 13:52:26 | jeff | *nod* |
| # | 13:52:27 | mrpeters-isl | i never use IE, so i'd never seen it |
| # | 13:52:46 | jeff | i'm familiar with that one, we used to get it often when we had some settings... over-tweaked. |
| # | 13:52:50 | mrpeters-isl | i guess i could throw some more ram at this thing and try again |
| # | 13:56:30 | Meliss has joined #evergreen |
| # | 13:57:36 | lisppaste3 has joined #evergreen |
| # | 13:58:19 | jamesrf has quit IRC |
| # | 14:03:58 | twirlip has left #evergreen |
| # | 14:21:46 | mrpeters-isl | this is so frustrating. some things give results, others don't |
| # | 14:21:58 | mrpeters-isl | "indy car" gets results, basketball spins forever |
| # | 14:22:24 | mrpeters-isl | "basket ball" gives results |
| # | 14:22:28 | atheos | mrpeters-isl - IMO that's not so much of a problem. the thing has good tastes |
| # | 14:22:35 | mrpeters-isl | ;) |
| # | 14:22:43 | dbs | indy cars are faster than basketballs |
| # | 14:22:46 | mrpeters-isl | true! |
| # | 14:22:53 | mrpeters-isl | its like the search can't handle big words |
| # | 14:23:37 | mrpeters-isl | + |
| # | 14:23:55 | mrpeters-isl | "will har" gets results "william henry" spins |
| # | 14:25:35 | dbs | mrpeters-isl: what locale is your database cluster? |
| # | 14:26:27 | dbs | show LC_COLLATE; -- in psql |
| # | 14:34:04 | bshum | dbs: That locale thing really messes with people. |
| # | 14:35:07 | bshum | Maybe there should be an included step somewhere in the Evergreen installation to change the locale to C regularly from system defined options. |
| # | 14:35:26 | bshum | If C is the most common that is. |
| # | 14:35:27 | dbs | bshum: it's set in the Makefile.install prereqs now |
| # | 14:35:50 | dbs | but it has to be set when PostgreSQL is installed; otherwise, it is subsequent pain |
| # | 14:36:05 | bshum | Right, right |
| # | 14:39:52 | vxbush has joined #evergreen |
| # | 14:40:05 | vxbush | hello |
| # | 14:40:27 | dbs | hello |
| # | 14:40:29 | jeff | greetings, vxbush |
| # | 14:40:52 | vxbush | I'm trying not to be rude on the dev listserv, but my emails don't seem to be going through, and I'm seriously stuck in getting 1.6.0.3 to run searches via the opac. |
| # | 14:40:57 | vxbush | Is everyone on holiday today? :D |
| # | 14:41:42 | jeff | vxbush: sometimes there are problems due to some over-zealous spam filters on the lists. Sorry about that! I don't think any of your e-mails have made it through. |
| # | 14:42:05 | dbs | vxbush: I haven't seen any email from you today |
| # | 14:42:06 | vxbush | That would explain it. How odd. Should I try to use the general list instead? I have some updates that might narrow down the problem. |
| # | 14:42:14 | dbs | or here... |
| # | 14:42:17 | jeff | vxbush: but you're here, so ask away! or, if easier, you could try pasting the details from your lost e-mails here: http://paste.lisp.org/new/evergreen |
| # | 14:42:22 | dbs goes to get a coffee |
| # | 14:42:40 | vxbush | Okay, let me get my last message posted via teh paste site...one moment |
| # | 14:44:15 | StephenGWills | sounds like the topic in eg-support-was-dev is performance today? can anybody play? *g* |
| # | 14:45:39 | mrpeters-isl | lc_collate |
| # | 14:45:39 | mrpeters-isl | ------------- |
| # | 14:45:39 | mrpeters-isl | en_US.UTF-8 |
| # | 14:45:39 | mrpeters-isl | (1 row) |
| # | 14:46:12 | lisppaste3 | vxbush pasted "Unable to search in OPAC" at http://paste.lisp.org/display/96838 |
| # | 14:46:16 | bshum | mrpeters-isl: I know your pain with searching now. |
| # | 14:46:23 | mrpeters-isl | :) |
| # | 14:46:32 | vxbush | Okay, post submitted. |
| # | 14:46:59 | mrpeters-isl | vxbush: looks like tsearch2 not enabled |
| # | 14:47:01 | bshum | mrpeters-isl: At least, I've had trouble with searches after upgrading from 8.2 to 8.3 postgres and not having my locales sync'ing right |
| # | 14:47:23 | bshum | mrpeters-isl: Didn't catch the whole story behind yours... *goes to look* |
| # | 14:47:27 | vxbush | Okay, so how do I fix it? I thought I set up tsearch when I tried to move my database up from 1.4.0.4 to 1.6.0.3 |
| # | 14:47:44 | jeff | vxbush: did you also upgrade from postgresql 8.2 to 8.3? |
| # | 14:47:45 | vxbush | "I'm a sysadmin, not a database administrator." |
| # | 14:47:52 | vxbush | jeff: Yes. |
| # | 14:48:13 | mrpeters-isl | DBD::Pg::st execute failed: ERROR: text search configuration "keyword" does not exist |
| # | 14:48:13 | mrpeters-isl | is what's making me think tsearch isnt on |
| # | 14:48:13 | jeff | okay, others here have been through similar -- but i have not. hopefully someone has some insight to share! |
| # | 14:48:51 | vxbush | mrpeters-isl: exactly. Seeing that everywhere. Got a suggestion? |
| # | 14:48:54 | StephenGWills | i am sorry to say that I was flummoxed by the 8.2 -> 8.3 threshold and installed 1.6.0.3 on top of 8.2 |
| # | 14:49:01 | miker_ | mrpeters-isl: you need to run: Open-ILS/src/sql/Pg/000.english.pg83.fts-config.sql from the source tarball |
| # | 14:49:15 | miker_ | vxbush: you too :) |
| # | 14:49:17 | vxbush | miker_: as postgres, I assume. |
| # | 14:49:23 | mrpeters-isl | interesting |
| # | 14:49:25 | miker_ | vxbush: in the database, yes |
| # | 14:49:25 | mrpeters-isl | ok |
| # | 14:49:40 | miker_ | tsearch2 is configured in a completely different way in 8.3 |
| # | 14:50:00 | vxbush | miker_: okay, going to test; will report in a minute |
| # | 14:50:18 | mrpeters-isl | miker_: that gave a bunch of syntax errors? |
| # | 14:50:26 | mrpeters-isl | psql -U evergreen < /home/opensrf/Evergreen-ILS-1.6.0.0/Open-ILS/src/sql/Pg/000.english.pg83.fts-config.sql |
| # | 14:50:59 | miker_ | mrpeters-isl: the db is 8.3? |
| # | 14:51:09 | mrpeters-isl | no, 8.2.12 |
| # | 14:51:16 | miker_ | oh ... sorry, thought you'd upgraded pg |
| # | 14:51:25 | mrpeters-isl | hehe production, yes but this is our migration server |
| # | 14:51:41 | miker_ | so, run the 8.2 version :) |
| # | 14:51:51 | mrpeters-isl | yep, will do |
| # | 14:52:11 | mrpeters-isl | ERROR: duplicate key violates unique constraint "pg_ts_cfg_pkey" |
| # | 14:52:56 | StephenGWills | fts wouldn't happen to be an acronym for Fix TSearch would it? |
| # | 14:53:09 | miker_ | StephenGWills: no, it wouldn't |
| # | 14:54:18 | mrpeters-isl | miker_: is there some table i need to drop, since this was a 8.3 database loaded into 8.2 |
| # | 14:54:23 | mrpeters-isl | i hadn't even given that any thought before |
| # | 14:54:39 | mrpeters-isl | er, wait...no its an 8.2.13 > 8.2.12 |
| # | 14:54:40 | dbs | mrpeters-isl: also - you need to drop the whole damned PG cluster, set your locale to C, and then create a new PG cluster |
| # | 14:54:54 | vxbush | miker_: An update. Ran the script, no errors. That seems like a good sign. Going to play it safe and restart everything and regenerate web pages via autogen. |
| # | 14:55:10 | dbs | mrpeters-isl: Then create a new database, add the contrib files, and then restore the database |
| # | 14:55:18 | mrpeters-isl | ugh, screw it...they can search by barcode then...i dont wanna reload a million bibs |
| # | 14:55:56 | vxbush | General Q: can I add a note to the update instrux (on the web) for 1.4 to 1.6 that you need to run this script? |
| # | 14:56:16 | vxbush | mrpeters-isl: heh |
| # | 14:56:44 | dbs | vxbush: with the condition that it only applies to you if you have also upgraded your database from 8.2 to 8.3 |
| # | 14:57:04 | mrpeters-isl | dbs: even if not on 8.3, still need to dump this? |
| # | 14:57:04 | vxbush | dbs: right. Definitely need to note that it is only required in that one scenario. |
| # | 14:57:08 | dbs | StephenGWills: fts = full text search (not sure if you were kidding) |
| # | 14:57:19 | dbs | mrpeters-isl: for you, your problem is that your locale is not C |
| # | 14:57:31 | mrpeters-isl | yeah, i guess i've never seen where that gets set |
| # | 14:57:49 | StephenGWills | tx dbs, i had so much trouble with 8.2 -> 8.3 that i am not sure if I was kidding either :) |
| # | 14:58:02 | dbs | mrpeters-isl: in testing conditions, with a handful of bibs, it wouldn't manifest as a problem |
| # | 14:59:44 | mrpeters-isl | yeah, with a couple or 3 million, it sure does :) |
| # | 15:01:17 | natschil has quit IRC |
| # | 15:01:47 | dbs | locale gets inherited from the user's environment when the PG cluster is created, which on Debian happens automatically when you install the PostgreSQL server |
| # | 15:02:06 | dbs | you can specify the locale if you create the cluster yourself |
| # | 15:02:12 | mrpeters-isl | ok, i was going to say that i don't recall ever setting it |
| # | 15:04:53 | dbs wonders how vxbush is doing |
| # | 15:06:08 | StephenGWills | do we have timing/metric tools to help find bottlenecks in an installation at this point? |
| # | 15:07:09 | bshum | mrpeters-isl: "pg_dropcluster --stop 8.3 main" and "pg_createcluster --locale=C 8.3 main" is how I've been doing it on our servers with the packaged version of postgres 8.3 |
| # | 15:07:31 | mrpeters-isl | bsuhm, yeah, i don't even have 8.3 installed |
| # | 15:07:44 | vxbush | dbs: sorry, I'm being super hyper careful; I'm restarting *everything* and having to help another customer at the same time |
| # | 15:08:14 | vxbush | Oh, frabjous day, I think it worked |
| # | 15:08:19 | dbs | miker_++ |
| # | 15:08:37 | vxbush | going to test the staff clients now |
| # | 15:09:01 | miker_ | dbs: eh? |
| # | 15:09:05 | miker_ | OH |
| # | 15:09:08 | miker_ | er, oh |
| # | 15:09:13 | miker_ | great! :) |
| # | 15:09:42 | bshum | mrpeters-isl: Well, it should work with 8.2 in place of 8.3 too. I think. |
| # | 15:09:56 | dbs | s/8.3/8.2/ of course |
| # | 15:10:04 | bshum | Not sure how the wrapper scripts handle themselves in previous versions. |
| # | 15:10:23 | mrpeters-isl | its worth a shot, im just nervous to tinker with it since it DOES work with barcode searches |
| # | 15:10:31 | mrpeters-isl | and its only a server for testing |
| # | 15:11:00 | vxbush | Update: interesting. The Mac client search works fine. The Windows client does not. If I enter a search term, it simply goes back to the regular search screen. |
| # | 15:11:12 | mrpeters-isl | sounds liek you're timed out |
| # | 15:11:23 | mrpeters-isl | idle more than 600 seconds? |
| # | 15:11:53 | bshum | dbs: You mentioned that you had set 32000 seconds for your timeout, what does the value for that look like? Do we need anything special there, like the word "seconds" or single quotes ' or some other funny thing? |
| # | 15:11:59 | bshum | *staff client timeout |
| # | 15:12:03 | phasefx | the staff client doesn't handle session timeouts with embedded interfaces like the OPAC very gracefully |
| # | 15:12:06 | dbs | bshum: just plain old 32000 |
| # | 15:12:09 | mrpeters-isl | bshum: just the # |
| # | 15:12:12 | vxbush | Okay, stopping and restarting widnows client |
| # | 15:12:23 | bshum | Cool, going to give it a shot with 28800 then |
| # | 15:12:28 | phasefx | if you pull up a xul native interface like Checkout, it'll prompt you for credentials to re-auth |
| # | 15:12:29 | dbs | vxbush: clear cache, too |
| # | 15:12:47 | dbs | in case some of the horrible old JS is cached |
| # | 15:13:07 | vxbush | dbs: will do |
| # | 15:13:40 | vxbush | Ah, windows client is good. Yay, fantastic, magnifico. |
| # | 15:13:54 | vxbush | I have potentially found a bug in the mac client |
| # | 15:13:58 | vxbush | unrelated to searches |
| # | 15:14:14 | dbs | seems likely |
| # | 15:14:18 | vxbush | Oh, no, nevermind; looks like that might have been a cache issue, too. |
| # | 15:14:56 | vxbush | Okay, thanks to all. One last question, for the documentation: is there a preferred time to run the 000.english.pg83.fts-config.sql script during the update? Should it be done at the very end, after all the upgrade scripts? |
| # | 15:15:22 | vxbush | (And who do I have to bribe to get off the spam list on the dev listserv?) |
| # | 15:17:41 | phasefx | you may be greylisted |
| # | 15:17:44 | dbs | vxbush: I haven't figured out how to get off the spamlist myself :) |
| # | 15:18:36 | vxbush | phasefx: Perhaps. I honestly did not see 000.english.pg83.fts-config.sql listed anywhere as required to update from postgres 8.2 to 8.3. That would have been nice. |
| # | 15:18:48 | phasefx | I think anything I post always gets SPAM put into the subject heading, friggin fraggin |
| # | 15:19:08 | phasefx | s/heading/header/, libraries on the brain |
| # | 15:19:56 | vxbush | Rats; I can't change the upgrade instrux in the wiki. No account. |
| # | 15:22:16 | phasefx | vxbush: pm me a preferred email address and username; I'll hook you up |
| # | 15:24:59 | dbwells | Is there a simple explanation concerning when 'locale=C' is needed? We have the same LC_COLLATE value as mrpeters-isl, and while things are working fine for us, I am still concerned. |
| # | 15:37:57 | brendan_bywater has joined #evergreen |
| # | 15:39:18 | bshum | I thought it had something to do with how the encoding for the database was done and what your initial setup looked like. |
| # | 15:39:34 | bshum | When we installed 1.4.0.4 for testing, we followed the old instructions |
| # | 15:39:42 | bshum | And they set it up with locale=C there |
| # | 15:39:43 | brendan_bywater has quit IRC |
| # | 15:39:58 | bshum | So when we migrated, we had to keep it locale=C or ended up being broken. |
| # | 15:40:39 | dbwells | bshum: thanks |
| # | 15:42:12 | pmplett has joined #evergreen |
| # | 15:42:23 | StephenGWills | dbs++ on the slides for OLA 2010, would like to have heard the talk that went with them. |
| # | 15:43:48 | dbwells | So locale only matters when importing from one DB server to another? Does that mean we will need to customize the Makefile.install (if/when the time comes to move) to not use C but to use our current value of en_US.UTF-8? |
| # | 15:43:54 | dbs | dbwells: no |
| # | 15:44:05 | dbs | dbwells: locale affects the speed of full-text search |
| # | 15:44:27 | dbwells | dbs: ah |
| # | 15:44:46 | dbs | dbwells: we dumped our en_CA.UTF8 database and loaded it into a new C database in February. No explosions. Faster searches & transactions. |
| # | 15:45:25 | dbs | dbwells: see bshum's instructions from above: |
| # | 15:45:37 | bshum | But back it up first :) |
| # | 15:45:43 | dbs | bshum> mrpeters-isl: "pg_dropcluster --stop 8.3 main" and "pg_createcluster --locale=C 8.3 main" is how I've been doing it on our servers with the packaged version of postgres 8.3 |
| # | 15:45:52 | mrpeters-isl | dbs: did you document that process at all? |
| # | 15:45:53 | dbs | bshum: good point :) |
| # | 15:45:58 | dbs | mrpeters-isl: what process? |
| # | 15:46:04 | mrpeters-isl | dbwells: we dumped our en_CA.UTF8 database and loaded it into a new C database in February. No explosions. Faster searches & transactions. |
| # | 15:46:22 | dbs | mrpeters-isl: nothing to document, it's a regular dump & restore |
| # | 15:46:29 | mrpeters-isl | ok |
| # | 15:46:41 | dbwells | dbs: That makes more sense. Seemed a bit strange that locale would stick around somehow for dump/restore, but I am no expert there. |
| # | 15:46:44 | dbs | only difference being that the db cluster was inited in C |
| # | 15:47:09 | mrpeters-isl | yeah, i guess im not clear on how to do that....i dont have pg_dropcluster anywhere...is that an 8.3 feature? |
| # | 15:47:25 | bshum | It depends on how the postgresql was installed |
| # | 15:49:26 | vxbush | phasefx: did you see my message? |
| # | 15:49:42 | jeff | phasefx: yes, bizarre bazaar metadata |
| # | 15:50:37 | dbs | mrpeters-isl: it's probably a debian wrapper |
| # | 15:50:44 | phasefx | jeff: a newer version of subversion would let bzr-svn do that differently, and I think trac could be configured to just not display it regardless |
| # | 15:50:47 | bshum | mrpeters-isl: pg_dropcluster is a script that comes with the packaged version of postgresql (at least that's the case in 8.3). If memory serves, it's supposed to handle dropping the cluster, etc. Then the pg_createcluster wraps up the initialization process for a new database. |
| # | 15:51:01 | phasefx will likely switch to git-svn at some point, as it is |
| # | 15:51:13 | mrpeters-isl | ok, thanks |
| # | 15:51:43 | mrpeters-isl | now to totally switch gears, could one make a custom macro for reciepts with some custom sql? |
| # | 15:51:45 | bshum | mrpeters-isl: There's a bunch of commands that is part of the process I think, and those commands just form a wrapper for the rest. |
| # | 15:51:46 | phasefx | jeff: a lot of merge data had accumulated in phasefx-sandbox because I stopped merging it into the bzr-svn trunk checkout :) |
| # | 15:51:54 | dbwells | dbs: Any major drawbacks to the switch? If we eventually add something like an author browse, is the simpler sorting logic in locale=C going to cause problems for diacritics, etc? |
| # | 15:52:25 | dbs | dbwells: all I can say is we're a bilingual library and we switched to locale=C |
| # | 15:52:25 | jeff | mrpeters-isl: macro for receipts would likely be js not sql, but what did you have in mind? |
| # | 15:52:37 | mrpeters-isl | a total of all outstanding bils for the user |
| # | 15:52:55 | mrpeters-isl | *bills |
| # | 15:53:17 | dbwells | dbs: :) I guess one could always add more ram and switch back if needed. |
| # | 15:53:17 | mrpeters-isl | so, "you currently owe $56.00 in fines" at the bottom of your reciept |
| # | 15:55:30 | phasefx | vxbush: you should be good to go |
| # | 15:57:58 | phasefx | mrpeters-isl: http://svn.open-ils.org/trac/ILS/changeset/15261 |
| # | 15:58:19 | jeff | nice. |
| # | 15:59:12 | vxbush | phasefx: Thanks, sir! |
| # | 15:59:43 | phasefx | mrpeters-isl: I'm not sure what other changes are needed with that though; doubt it would be fun to port it into 1.6 |
| # | 16:00:18 | jeff | phasefx: looks pretty simple to port at first glance -- anything specific you're worrying about? |
| # | 16:00:37 | phasefx | jeff: the field in that payload from the checkout event |
| # | 16:01:38 | bshum | mrpeters-isl: Out of curiosity, you said you guys are using postgres 8.2.12, right? Did you install it via the cheat sheet method of using stow? |
| # | 16:01:41 | jeff | oh, the most_recent_balance_owed? |
| # | 16:02:24 | jeff | oh, nevermind. i see. |
| # | 16:03:20 | phasefx | right |
| # | 16:03:39 | dbs | phasefx: then we should probably just switch to git period - if you, atz, senator, and berick are all using it |
| # | 16:04:16 | phasefx | dbs: no objections from me. Could vote on it during a dev meeting, and then task some folks with making it happen :) |
| # | 16:04:32 | jeff | the folked tasked with switching you to git? |
| # | 16:04:34 | dbs | dev meetings? We have those? And action items that get acted on? :) |
| # | 16:04:36 | jeff | oh. again, nevermind. |
| # | 16:05:01 | Meliss has quit IRC |
| # | 16:05:11 | dbs considers just how many action items have his name beside them, waiting forlornly |
| # | 16:05:12 | sfortin has quit IRC |
| # | 16:05:25 | phasefx seconds loading dbs up with more action items |
| # | 16:06:25 | jeff | mrpeters-isl: http://svn.open-ils.org/trac/ILS/changeset/15245 should be the other changeset required to make that %balance_owed% macro work. |
| # | 16:06:46 | phasefx | jeff++ |
| # | 16:07:14 | jeff | You have borrowed $26.90 worth of materials. Thank you for using the library! |
| # | 16:07:19 | StephenGWills | where is the timeout we increase to try and solve staff clients hanging? |
| # | 16:07:39 | jeff | oh, don't increment me yet. retrieve_money_user_summary doesn't exist in 1.6 either. |
| # | 16:07:42 | bshum | StephenGWills: If you're using 1.6 I believe it's part of the Library Settings Editor |
| # | 16:07:46 | StephenGWills | app_settings? |
| # | 16:07:55 | phasefx | StephenGWills: but a staff client hanging probably isn't due to any timeout settings |
| # | 16:08:02 | jeff | oh. nevermind... it's there, it's auto-generated. :P |
| # | 16:08:27 | phasefx | jeff: the editor threw me for a loop the first time I encountered it |
| # | 16:08:29 | bshum | oh right, hanging. Hanging sounds bad. |
| # | 16:09:27 | StephenGWills | catalogers are reporting pulling in recs via z39.50 and having the client hang while resolving the result, when they look later the record is there. |
| # | 16:09:43 | StephenGWills | I thought that might mean the js is timing out and going away? |
| # | 16:11:08 | vxbush | My Evergreen installation is a much happier camper now. However, I have a completely different question I'd like to pose, something that has been an issue for a while. Some of the MARC records seem to have extra colons and spaces in the title and author fields. |
| # | 16:11:23 | vxbush | If I do a search for those books, they don't show up in searches. |
| # | 16:11:30 | vxbush | But if I do a shelf browser view, i can see the books. |
| # | 16:11:46 | vxbush | As I am not a database administrator, what's the easiest way to go through and clean this up to get rid of excess spaces and colons? |
| # | 16:11:48 | phasefx | StephenGWills: you generally have 60 seconds before the http library being used times out on a given request |
| # | 16:13:32 | StephenGWills | is there a commandline cleint like srfsh running on windows I can use to test requests from staff clients in the field w/o the Staff Client overhead? |
| # | 16:13:46 | StephenGWills | client* |
| # | 16:13:52 | vxbush | StephenGWills: a good idea, methinks |
| # | 16:14:14 | StephenGWills | i'll bring my linux laptop :) |
| # | 16:15:12 | vxbush | StephenGWills: what about a stand-alone unix-style shell that has osrfsh installed? Not sure if it's doable, but if you could find something that would run off a USB drive, that could be an easy solution. If possible. |
| # | 16:17:01 | vxbush | StephenGWills: an example of cygwin on a USB drive: http://sites.google.com/site/devinsezer/Home/software/portable-cygwin |
| # | 16:17:58 | StephenGWills sips a margarita and says.... too much work! j/k I do not think it is a network issue, turnaround is 230ms however. |
| # | 16:18:31 | StephenGWills | the fact that it is inconsistant is why i thought it might be timeout |
| # | 16:19:02 | StephenGWills | anyway... Dr's appt trumps it. will care again tomorrow, night all |
| # | 16:19:18 | StephenGWills has left #evergreen |
| # | 16:20:53 | jamesrf has joined #evergreen |
| # | 16:21:54 | bshum | pg_dropcluster and pg_createcluster are apparently scripts that come with the postgresql-common package. Not sure if that's helpful or not. |
| # | 16:22:57 | dbs | pg_createcluster is probably just a wrapper for initdb; pg_dropcluster is probably just a wrapper for "rm -fr /path/to/database/cluster" :) |
| # | 16:23:31 | bshum | Pretty much actually :) |
| # | 16:23:42 | bshum | I was going to say that next, that you would just delete your cluster's data files |
| # | 16:23:52 | bshum | And then do a new initdb with the locale option of C |
| # | 16:23:57 | dbs | bshum: with start and stop commands I guess :) |
| # | 16:24:38 | bshum | dbs: heh, it was just easier for me to write down two lines to make the locale switch every time I installed postgres. |
| # | 16:26:19 | dbs | yeah, just confirmed that the various install_pgsql_server_debs_* targets in rel_1_6_0 are prefixed with "LANG=C" |
| # | 16:27:04 | bshum | Excellent. |
| # | 16:28:01 | bshum | And on that fun note, have a good night everybody. Have fun with action/trigger stuff dbs, I'll be poking around more on that tomorrow. |
| # | 16:29:09 | bshum has quit IRC |
| # | 16:34:18 | vxbush | perhaps I should ask my question tomorrow? :D |
| # | 16:36:15 | phasefx | vxbush: a link to a problem record in your OPAC might help. If your indicators are off in a record's 245, it can produce incorrect colons during a behind the scenes MARC to MODS transform |
| # | 16:37:45 | phasefx | also, for cases where records aren't showing up when they should, it may be useful to see if editing and saving the record (even with no changes) makes a difference. And with such a test, remember to add some garbage to the search (eg. harry potter -asdfkjasdflkasfj) since search results are cached for a while, even if the results were broken |
| # | 16:38:27 | dbs | vxbush: specific examples would certainly help. I've poked around in those waters before. |
| # | 16:39:31 | vxbush | phasefx: sure. Here's an example (I think), that uses an extra slash. I'm pretty sure this record was imported via Z39.50: |
| # | 16:39:32 | vxbush | http://commons.ctlt.ilstu.edu/opac/en-US/skin/ctlt/xml/rdetail.xml?r=315&t=art&tp=keyword&d=0&hc=4&rt=keyword |
| # | 16:40:27 | vxbush | (sigh; they didn't clean out the 920 record, silly humans) |
| # | 16:41:27 | phasefx | mm, that page is taking forever to load from here |
| # | 16:42:05 | phasefx | and a reload fixes it |
| # | 16:42:31 | vxbush | Whew; I was worried the university had forgotten to leave a hole in the firewall for that site. |
| # | 16:42:50 | phasefx | vxbush: okay, so there's a trailing slash in the 245a, and you want to get rid of that? |
| # | 16:43:10 | vxbush | phasefx: Yes. But some records have slashes, some have spaces, some have colons. |
| # | 16:43:26 | phasefx | I thought catalogers desperately want such things to stay in their records :) |
| # | 16:43:29 | vxbush | And it's not just 245a. It's also 100a |
| # | 16:43:32 | dbs | vxbush: that's pretty common cataloguing practice |
| # | 16:44:01 | phasefx | harkens back to ye old card catalog days when the values from these subfields would just be concatenated together and printed |
| # | 16:44:12 | vxbush | dbs: That's great, but then the search functions don't work. If I try to search on "culture" in the title, I don't ever find this book. |
| # | 16:44:30 | dbs | vxbush: that's crazy talk |
| # | 16:44:32 | vxbush | phasefx: oh, you would remind me of that; |
| # | 16:44:43 | phasefx | if a slash caused that, removing the slash wouldn't be the solution |
| # | 16:44:44 | vxbush | my mother was a librarian. Well do I remember card catalogs. |
| # | 16:44:46 | jeff | the "extra colons and spaces" may be a red herring, you might actually be seeing a problem at ingest that's making these not-searchable. |
| # | 16:44:55 | vxbush | So a re-ingest is needed? |
| # | 16:45:06 | vxbush | I thought I recalled a script for that....right? |
| # | 16:45:27 | jeff | there is. others here could advise you on how to see if it's an ingest-related problem before you try to fix it like one. |
| # | 16:46:42 | vxbush | I'm willing to try anything. My database is backed up. :D |
| # | 16:46:46 | dbs | vxbush: there are no copies attached to the record? |
| # | 16:46:58 | jeff | my rdetail page hasn't loaded yet. |
| # | 16:47:06 | vxbush | dbs: yes, there is one copy available out of one. |
| # | 16:47:34 | vxbush | Odd how it is so slow; it's zippy fast here. |
| # | 16:47:44 | phasefx | vxbush: I see no copies in that opac view, in which case, that record would only be searchable in the staff client, all else being equal |
| # | 16:47:57 | dbs | weird - I see copies in Firefox, not in Chromium |
| # | 16:48:09 | vxbush | phasefx: but I see a copy available in the opac. It's there. |
| # | 16:48:11 | phasefx | I'm using Firefox 3.6 and no copies |
| # | 16:48:33 | dbs | there is also no org tree |
| # | 16:48:34 | vxbush | oh, crap; what just went wrong? This is what I was seeing before, when copies wouldn't appear. |
| # | 16:48:45 | vxbush | But I ran the script you guys suggested above. |
| # | 16:48:50 | vxbush | Why would it go back? |
| # | 16:49:04 | vxbush | [hits head on wall] |
| # | 16:50:16 | dbs | and now... no copies in the locatin |
| # | 16:50:31 | dbs | is this a multi-server setup? |
| # | 16:50:42 | phasefx | something screwy with the org tree.. I see 3 org units in OrgTree.js, but the Choose a library menu doesn't work |
| # | 16:50:57 | vxbush | Right. So why is that happening? |
| # | 16:51:01 | vxbush | grumble grumble |
| # | 16:51:29 | vxbush | dbs: no, no multi-server setup. Just a single machine. |
| # | 16:51:49 | jeff | i suspect something is in the way of us and your web server, possibly dropping/blocking sessions. |
| # | 16:52:04 | vxbush | jeff: must be. Doesn't make sense, otherwise. |
| # | 16:52:12 | jeff | doesn't seem to be dns related. |
| # | 16:52:18 | dbs | http://commons.ctlt.ilstu.edu/osrf-gateway-v1?locale=en-US&method=open-ils.search.biblio.copy_location_counts.summary.retrieve¶m=315¶m=1¶m=0&service=open-ils.search |
| # | 16:52:25 | phasefx tries https to no avail |
| # | 16:52:29 | dbs | returns a copy location count |
| # | 16:52:30 | jeff | my first guess was the hostname returning multiple IPs, or nameservers returning different IPs |
| # | 16:52:59 | dbs | bad ou depth match with ou_type? |
| # | 16:53:02 | vxbush | I'm seeing about ten idle processes |
| # | 16:53:14 | vxbush | dbs: Um, English, please. |
| # | 16:53:34 | vxbush | jeff: no, just one IP address, hardcoded. No DHCP issues. |
| # | 16:53:50 | jeff | just now is when i was able to load the rdetail page and javascript for the first time. |
| # | 16:54:51 | jeff | {"payload":[[["5","LB1051.B736 1996","Commons",{"0":1}]]],"status":200} |
| # | 16:55:33 | vxbush | Perhaps I should redo the ou stuff. Back when I was first setting this up, I wanted to give us the option to expand to a second location, so I set it up as university, then department, then the local library. |
| # | 16:55:36 | vxbush | But that doesn't seem to have worked. |
| # | 16:55:46 | dbs | ISU is the parent of CTLT, which is the parent of RC? |
| # | 16:56:02 | vxbush | dbs: right. RC is the name of our little library. |
| # | 16:56:39 | dbs tries to decode http://commons.ctlt.ilstu.edu/opac/common/js/en-US/OrgTree.js |
| # | 16:56:55 | vxbush | dbs: I see: var _l = [[2,2,1,"Center for Teaching, Learning & Technology",0,"CTLT"],[1,1,null,"Illinois State University",0,"ISU"],[5,3,2,"Resource Commons",1,"RC"]]; /* Org Units */ var globalOrgTypes = [new aout([null,0,0,0,1,"University","University",null]), new aout([null,0,0,1,2,"Department","Local Department",1]), new aout([null,1,1,2,3,"Library","This Specialized Library",2])]; /* OU Types */ |
| # | 16:57:12 | dbs | right, that's what I'm seeing. Just trying to ensure that the depths match up |
| # | 16:58:01 | dbs | ou_type is second parm, according to org_utils.js |
| # | 17:00:27 | dbs | Things seem to match up, if I'm reading the aout definition in fm_IDL.xml and the _l parsing in org_utils.js correctly |
| # | 17:00:46 | vxbush | Heh; who just tried to log into admin? |
| # | 17:01:00 | dbs | not I |
| # | 17:01:08 | jeff | there is no open-ils.search.biblio.record.copy_count call being made that i can see. |
| # | 17:01:13 | BobW has joined #evergreen |
| # | 17:03:06 | vxbush | dbs: okay, that's good news. Now, here's a question: I only want the RC library to show up, so right now I have ISU and CTLT set to be OPAC invisible; I don't want people to think our library can be used to search the general ISU library. |
| # | 17:03:19 | vxbush | Am I hitting a bug? |
| # | 17:03:24 | dbs | oh |
| # | 17:03:37 | dbs | children of invisible parents don't get seen? |
| # | 17:03:58 | vxbush | dbs: Maybe that's it. I could make the two invisible, but that will cause problems for me. |
| # | 17:04:12 | vxbush | s/invisible/visible (end of day) |
| # | 17:04:32 | dbs | I would do that first and see if it solves the problem, then change the names to avoid confusion :) |
| # | 17:04:52 | vxbush | dbs: wait. As I'm tired, I want to make sure I undersatnd. Change the names of .... what? |
| # | 17:05:10 | vxbush | But I don't see how this will fix the copy problem. |
| # | 17:05:15 | dbs | From "ISU" to whatever, to prevent people from thinking they can search everywhere |
| # | 17:05:34 | vxbush | Oh, I wonder if a postgres vacuum did something to trick evergreen.....but that seems very unlikely. |
| # | 17:05:45 | phasefx | I'm just pasting this in here for irc log goodness: http://localhost/osrf-gateway-v1?locale=en-US&service=open-ils.actor&method=open-ils.actor.org_tree.retrieve&format=xml |
| # | 17:06:11 | phasefx | for catching when OrgTree.js doesn't match what the database thinks |
| # | 17:06:53 | vxbush | Can I take the site down for a minute and restart? I'd like to do that before I try to fix the OU names. |
| # | 17:06:57 | dbs | yeah - the 0 after the _l names are "opac.visible", should have caught that |
| # | 17:06:59 | vxbush | Don't want to trip anyone |
| # | 17:07:06 | dbs | doesn't bother me! |
| # | 17:07:15 | vxbush | going down |
| # | 17:10:32 | vxbush | [twiddles thumbs] |
| # | 17:13:23 | vxbush | okay, server is back up, and copies seem to be showing up again |
| # | 17:14:19 | dbs | alas, not here |
| # | 17:14:36 | phasefx | no copies for me |
| # | 17:15:18 | dbs | vxbush: time to change the opac visibility on the parent libraries and rerun autogen |
| # | 17:15:32 | dbs | you won't have to restart |
| # | 17:15:42 | vxbush | Right; now I have had both off. Shall I turn both on, to begin? Seems prudent... |
| # | 17:16:03 | dbs | yes. |
| # | 17:16:33 | vxbush | running autogen |
| # | 17:16:36 | vxbush | done |
| # | 17:16:53 | dbs | copies! |
| # | 17:16:55 | dbs | huzzah |
| # | 17:16:58 | jeff | hrm. |
| # | 17:17:14 | phasefx | copies here |
| # | 17:17:42 | vxbush | Right, but those showed up before I changed the OU and ran autogen. |
| # | 17:17:44 | phasefx | vxbush: some libraries are hacking their opac skins to hide parent libraries; it's probably not fun |
| # | 17:18:04 | phasefx | some = at least one |
| # | 17:18:38 | vxbush | This library isn't related to the campus library, so there's a real need to make our library look independent. |
| # | 17:19:01 | vxbush | But the OU doesn't look like it would accept a one-unit tree. |
| # | 17:19:06 | vxbush | At least, not happily. |
| # | 17:19:22 | phasefx | could make the top-level org very very generic |
| # | 17:19:51 | vxbush | Like make it "location" rather than "university". Or, I could use a building name instead. That might actually work better. |
| # | 17:19:59 | vxbush | It would make it much clearer. |
| # | 17:20:04 | vxbush | Okay. This is great and loverly. |
| # | 17:20:05 | vxbush | BUT |
| # | 17:20:27 | vxbush | I'm still worried about my copies disappearing again. And then there's still the issue of searches on authors and titles not showing up. |
| # | 17:20:35 | vxbush | But I'm willing to revisit this tomorrow. |
| # | 17:20:39 | vxbush | It's almost the end of my workday. |
| # | 17:20:41 | dbs | sounds like a plan |
| # | 17:20:44 | phasefx | 1) Consortium 2a) University hierarchy 2b) unrelated library |
| # | 17:21:00 | vxbush | Actually, I'm thinking: |
| # | 17:21:07 | vxbush | 1) Building 2) department 3) library |
| # | 17:21:30 | vxbush | Then it's very clear we're not hooked up to the campus library. |
| # | 17:21:37 | vxbush | It's a cheat, and I freely admit it. |
| # | 17:22:22 | jeff | dbs: why does your opac rdetail.xml call open-ils.search.biblio.record.copy_count? others don't seem to... are you doing something unique, or am i just seeing different behavior at different depths? |
| # | 17:23:04 | dbs | jeff: IIRC, it's because I hide the "Place hold" button if any copies are available |
| # | 17:23:14 | jeff | ah! :-) |
| # | 17:23:23 | phasefx | vxbush: was always intended for higher-up org units to be abstract (whether geographical, political, functional, etc) |
| # | 17:24:27 | dbs | vxbush: http://svn.open-ils.org/trac/ILS-Contrib/browser/conifer/branches/rel_1_6_0/tools/daily-scripts/reingest_uningested.pl might be of assistance tomorrow |
| # | 17:25:02 | dbs | run it without the --reingest flag, and it will list the record IDs of records that haven't been ingested |
| # | 17:25:17 | vxbush | dbs: beautiful |
| # | 17:25:21 | vxbush | stored |
| # | 17:25:39 | dbs | it's far from foolproof, but it might be a good start |
| # | 17:28:17 | jeff | anyone played with courtesy/pre-overdue e-mails that vary based on item circ mod or circ due_time, such that you don't send "item due in 3 days" reminders the second day on an item with a four day duration? |
| # | 17:28:31 | vxbush | I'm pretty sure I ran that before I did the update to 1.6, but I'll play with it. |
| # | 17:28:33 | vxbush | Okay, logging off |
| # | 17:28:39 | r123 has quit IRC |
| # | 17:28:44 | jeff | i can make it happen, but i'm wondering if anyone else has attempted similar, and what their approach was. |
| # | 17:28:47 | vxbush | thanks, everyone. Seriously. I was going to end up with severe brain damage. |
| # | 17:29:00 | vxbush | I'll log in tomorrow |
| # | 17:29:24 | vxbush has left #evergreen |
| # | 17:31:30 | dbs | jeff: not there yet, going to try and do it with the action / trigger framework |
| # | 17:31:54 | dchristens has left #evergreen |
| # | 17:51:44 | BobW has quit IRC |
| # | 17:55:05 | dbs has quit IRC |
| # | 18:17:36 | jamesrf has quit IRC |
| # | 18:32:46 | jamesrf has joined #evergreen |
| # | 18:36:27 | jamesrf has quit IRC |
| # | 18:37:46 | jamesrf has joined #evergreen |
| # | 18:44:47 | jamesrf has quit IRC |
| # | 18:44:53 | jamesrf_ has joined #evergreen |
| # | 18:45:11 | jamesrf_ has left #evergreen |
| # | 19:33:35 | jamesrf_ has joined #evergreen |
| # | 19:58:15 | miker_ | jeff: IIRC, there's code to handle that in trunk, but not 1.6 |
| # | 20:05:50 | jamesrf_ has quit IRC |
| # | 20:29:37 | Callender has quit IRC |
| # | 20:29:37 | mrpeters-isl has quit IRC |
| # | 20:29:37 | jeff__ has quit IRC |
| # | 20:29:37 | _bott_ has quit IRC |
| # | 20:29:37 | Dmagick-mac has quit IRC |
| # | 20:29:38 | lisppaste3 has quit IRC |
| # | 20:29:38 | jeff has quit IRC |
| # | 20:29:38 | csharp_GPLS has quit IRC |
| # | 20:29:38 | mck9 has quit IRC |
| # | 20:29:38 | emrikol^ has quit IRC |
| # | 20:29:39 | eby has quit IRC |
| # | 20:29:39 | Dmagick has quit IRC |
| # | 20:29:40 | phasefx_ has quit IRC |
| # | 20:29:40 | berick has quit IRC |
| # | 20:30:54 | Dmagick has joined #evergreen |
| # | 20:32:18 | Callender has joined #evergreen |
| # | 20:32:18 | mrpeters-isl has joined #evergreen |
| # | 20:32:18 | mck9 has joined #evergreen |
| # | 20:32:18 | emrikol^ has joined #evergreen |
| # | 20:32:18 | eby has joined #evergreen |
| # | 20:32:18 | jeff__ has joined #evergreen |
| # | 20:32:18 | _bott_ has joined #evergreen |
| # | 20:32:40 | jeff has joined #evergreen |
| # | 20:32:41 | berick has joined #evergreen |
| # | 20:32:46 | chrissharp123 has joined #evergreen |
| # | 20:32:51 | Dmagick-mac has joined #evergreen |
| # | 20:33:00 | phasefx_ has joined #evergreen |
| # | 20:45:10 | dbs has joined #evergreen |
| # | 21:26:52 | dbs | miker_: thanks for the email "from the archives" - I swear that I've never seen that before |
| # | 21:32:56 | dbs | Definitely helps describe the passive vs. active hooks distinctions |
| # | 21:38:26 | miker_ | you haven't ... it was between berick while I was designing a/t and working on the initial implementation |
| # | 21:38:33 | miker_ | haven't seen it, I mean |
| # | 21:52:51 | dbs | ahh! I thought I had missed it on the list or something :) |
| # | 22:04:47 | phasefx_ has quit IRC |
| # | 22:04:53 | Dmagick-mac has quit IRC |
| # | 22:05:06 | phasefx_ has joined #evergreen |
| # | 22:06:31 | Dmagick-mac has joined #evergreen |
| # | 23:56:27 | dbs has quit IRC |
April 10-13