Open Source Integrated Library System

Evergreen on IRC

#evergreen Logs for Thursday, December 17th, 2009

< Wednesday, December 16th, 2009Raw Log FileFriday, December 18th, 2009 >
#TimeNickMessage
#00:46:39phase_bb has quit IRC
#02:34:42jamesrf has joined #evergreen
#02:43:37jamesrf has quit IRC
#02:53:25phasefx_ has quit IRC
#04:13:48finnx has quit IRC
#04:14:23moodaepo has quit IRC
#04:14:23jmeeuwen has quit IRC
#04:14:23brendan_ga_ has quit IRC
#04:14:23AbizzalsX has quit IRC
#04:14:23lcassell has quit IRC
#04:14:23jeff has quit IRC
#04:14:23leed has quit IRC
#04:14:23mrpeters-isl has quit IRC
#04:14:23atz has quit IRC
#04:14:23_dkyle_ has quit IRC
#04:14:23_bott_ has quit IRC
#04:14:23pmplett has quit IRC
#04:14:23kbeswick has quit IRC
#04:14:23gmcharlt has quit IRC
#04:14:23rsinger has quit IRC
#04:14:23artunit has quit IRC
#04:14:23gdunbar has quit IRC
#04:17:15pinesol has joined #evergreen
#04:17:41jeff_ has joined #evergreen
#04:18:19rickd has joined #evergreen
#04:19:27chrissharp123 has joined #evergreen
#04:35:41phasefx2 has joined #Evergreen
#04:35:41atheos has joined #Evergreen
#05:33:01levani_el has joined #evergreen
#05:33:03levani_elhello
#05:33:57levani_eli want to generate 008 marc field date
#05:35:51levani_elin template generate only one date
#05:36:09levani_eli want to take information at 005 field
#05:57:32levani_el has quit IRC
#06:24:52jmeeuwen has quit IRC
#06:51:57bradl has joined #evergreen
#07:24:43wlayton has joined #evergreen
#07:32:41wlayton has quit IRC
#07:39:08Meliss has joined #Evergreen
#07:41:01miker_ has joined #evergreen
#07:42:31Melissa has joined #Evergreen
#07:43:07Meliss has quit IRC
#07:45:13Melissa has left #Evergreen
#07:46:59Melissa has joined #Evergreen
#07:49:28phasefx_ has joined #evergreen
#07:51:38phase_bb has joined #evergreen
#07:52:01sylvar has joined #evergreen
#07:57:16mck9 has joined #evergreen
#08:03:36sfortin has joined #evergreen
#08:06:44rickd has quit IRC
#08:07:16rickd has joined #evergreen
#08:28:08wlayton has joined #evergreen
#08:31:00berick has joined #evergreen
#09:11:14jenny has joined #evergreen
#09:15:44senator has joined #Evergreen
#09:27:02rickd has quit IRC
#09:33:32dchristens has joined #evergreen
#09:34:58phasefx3 has joined #evergreen
#09:35:07phasefx3 is now known as phasefx
#09:40:06dbs has joined #evergreen
#09:42:19dbsatz: did you ever work out your opensrf / open-ils build issues? i was wondering if you ever ran autogen.sh (the autotools one, not the evergreen bootstrapping one) in opensrf and evergreen before rerunning configure and make
#09:42:40atzyeah, i finally got it up to speed
#09:44:38atzthe main thing was the libmemcached dependency that was breaking make clean... my update script also threw an unnecessary git-related wrinkle in there
#09:47:18atzthx for the help though
#09:56:23r123 has joined #evergreen
#09:56:56natschil has joined #evergreen
#09:57:35dbwells has joined #Evergreen
#09:59:40mrpeters-islall - are there any flags the patron could set that would prevent their contact info (appearing in the staff client as valid) from printing out on hold reciepts?
#10:00:04mrpeters-island if yes, is there anyway in database to verify that they have "opted out" of having their contact info on those slips
#10:01:24atzi know there is a "hold alias" field recently added, but I'm not sure that overrides *all* contact info
#10:01:31phasefx_the stock hold slip will only print the patron's name (and maybe card number?), and if email notify is enabled on the hold, their email address, and if phone notify is enabled on the hold, whichever phone number they entered for the hold
#10:01:37phasefx_hold alias will completely replace name
#10:02:26mrpeters-islhmm...its being explainted to me that they have perfectly valid phone/email on their slips but they are not printing out with contact info...is it possible to place a hold with no notify?
#10:02:30phasefx_in trunk, the hold slip is completely templatable, but I'm not sure how much patron data is available for it, and there's certainly no dedicated toggle the template could look at for determining whether to use that data or not
#10:02:53mrpeters-islok
#10:03:04phasefx_what specific contact info are they expecting?
#10:03:08mrpeters-islso, next clue is to check the action.hold_request for that item and see if there was a phone or email notify
#10:03:25mrpeters-isl"No contact information given on some hold slips"
#10:03:38atzit could be, for example, that the phone is a different field than the one(s) the template is targeting
#10:03:52mrpeters-islatz: thats a good thought
#10:04:23atzthe template should be made to fail over though all 3 phone fields when possible
#10:05:05atzif it already is designed that way, then look for a possible whitespace data string as a phone value
#10:05:50atzsuch that the template would succeed in retrieving a non-null value... but it still would look like nothing
#10:05:57dbshmm, more install permissions errors at http://testing.esilibrary.com/buildbot/build/?openils=15183
#10:06:32phasefx_in this case, the hold interface itself lets you enter an arbitrary phone number, one that need not be associated with the patron's account at all. It should default with an associated patron phone number, though
#10:07:02phasefx_so action.hold_request.phone_verify contains a phone number
#10:07:14atzah, i forgot about that
#10:09:04phasefx_dbs: so I didn't break the build? :D yay
#10:10:48mrpeters-islwhaddaya know! phone_notify |
#10:10:48mrpeters-islemail_notify | f
#10:11:18mrpeters-islso, they likely didnt enter a phone number to be notified at when the hold came available, plus they opted out of email notification
#10:12:28mrpeters-islthus, a hold capture slip with no way to contact the ptaron
#10:12:45mrpeters-islshould that fall over to provide the day_phone from the patron account if phone notify is null?
#10:12:58bradlthat's why I've been pushing for the ESP module. No one wants to fund that, though. *shrugs*
#10:13:07mrpeters-islESP/
#10:13:10mrpeters-isloh lol
#10:13:12mrpeters-islim dense
#10:13:23phasefxif they patron doesn't want to be contacted by phone, they don't want to be contacted by phone :) they have to go to the effort of disabling it
#10:13:23bradlnah, you just performed a migration :) (congrats)
#10:13:30mrpeters-islthanks!
#10:13:50mrpeters-islphasefx: very true
#10:13:51phasefxnot an ideal situation though. they're still able to check their own My Account section in the OPAC
#10:14:28phasefxcould template away their ability to disable and configure the phone number
#10:14:52phasefxthough I imagine you have patrons in the database without valid phone numbers and/or email addresses
#10:14:59mrpeters-islseems pretty stupid to place a hold, and leave no contact info
#10:15:19phasefxmaybe some coding to warn the patron of this
#10:15:27mrpeters-islbut, i agree...they had to go to the effort of disabling it
#10:16:03phasefxand a link pointing to library hold policies (for how long it'll stay on the shelf, any fees for un-retrieved holds, etc.)
#10:16:16mrpeters-islan idea for sure
#10:26:09lisppaste3 has joined #evergreen
#10:36:47natschil has quit IRC
#11:18:29dbsis anyone opposed to placing a (bucket, target_biblio_record_entry) unique constraint on container.biblio_record_entry_bucket_item?
#11:19:03dbswe've had a lot of people accidentally add multiples, and I can't think of a good rationale for allowing dupes
#11:21:35berickdbs: i fear there may be some assumption of allowing duplicates. I know i've told people before, "sure, you can have dupes!". I recall discussing using dupes as a way to 'rate' items in a bucket, potentially for user ratings in the opac.
#11:21:52bericki don't have anything more concrete, though
#11:23:32dbshuh. it would be a weird implementation, but I guess if you had a "5 star" shared bucket that people could add records to, that would be a way of doing things.
#11:23:56berickor "38 people liked this title"
#11:24:02berick*shrug*
#11:24:26berickhonestly, i don't recall the exact discussion
#11:25:28berickperhaps a "allow dupes" checkbox in the staff client
#11:27:05dbsmebbe. I like good old-fashioned relational constraints :) perhaps I'll just add that constraint here, and open a discussion on the list
#11:27:36dbssomething like rating.biblio_record_entry(id sequence, usr id, record id, rating int, comment text)
#11:28:15dbs(leaving open the possibility of rating.actor_usr(id sequence, rated_usr id, rating_usr id, rating int, comment text) -- fun!)
#11:28:30dbsopen/reopen
#11:29:49dbsthanks, berick - as always
#11:32:01dbshuh, I wonder if some of these dupes are due to back-button usage. I can't imagine how anyone would add 57(!) dupes to a single bucket
#11:38:00berick"57 people liked this title"
#11:38:02berick ducks ;)
#11:38:21berick"1 person liked it 57 times"
#11:39:31dbs"you students really should read this book. I mean it. I really mean it!"
#11:39:34phasefxor for a patron bookbag, I read it twice, or something. But I likes a toggle :)
#11:39:48berickphasefx: good point, actually..
#11:39:59berickwe use buckets to model opt-in checkout history
#11:40:23agJohn has joined #evergreen
#11:42:22agJohnIs there a way to set up the OPAC for two different libraries in a consortium so that one library sees/searches one type of subject heading (based on indicators) and another library a different type? (I don't need to know details, just if it could be done, at this point)?
#11:44:04dbsagJohn: sees, yes, using skins and bibtemplate
#11:44:22agJohnCool, thanks, dbs.
#11:44:34dbsagJohn: searches, a bit harder; need a skin and a separate index I guess
#11:44:43agJohnMakes sense.
#11:45:08dbsrelated subjects might be a tiny bit wonky
#11:45:24agJohnI wasn't sure the indexing was sophisticated enough to take into account the indicators. Yeah, I figured it could get complicated.
#11:48:12dbsagJohn: this a MeSH vs. LCSH sort of thing?
#11:50:05agJohndbs: I wasn't clear on why they wanted to do this. I sort of had the impression is was subtler than that: snooty college library doesn't trust the cavalier public library's cataloging kind of thing. There's no member of the group that would be using MeSH, so far as I know, so I'm not entirely sure. I did not pursue the details because I didn't have a good answer when they wanted to know if...
#11:50:06agJohn...it was possible or not.
#11:51:07agJohnSeems like everyone I'm talking to recently wants to share the server(s) and DB, but have everything be treated as if they were separate....
#11:51:30dbshah. lots of work involved in that.
#11:51:55r123 has quit IRC
#11:51:56agJohnSeems kinda nutty to me; I suggested they consider just having separate DBs w/ VMs for the EG instances.
#11:57:04brendan_ga_ has quit IRC
#12:05:21godzilla8nj has joined #Evergreen
#12:06:56godzilla8njInstalling Evergreen (well, at this point still installing OpneSRF.) As opensrf user run osrf_ctl.sh -l -a start_all returns "Can't locate OpenSRF/Utils/Logger.pm". Help?
#12:11:43brendan_ga has joined #evergreen
#12:11:59brendan_ga has quit IRC
#12:15:07godzilla8nj has quit IRC
#12:33:36gmcharltoops, I hope he doesn't take out the lack of an answer on Tokyo
#12:36:24sboyettehistory shows, again and again, &c. &c.
#12:46:53jamesrf has joined #evergreen
#12:49:39jamesrf has left #evergreen
#13:36:44dbs@later tell godzilla8nj Check your PERL5LIB environment variable to ensure it contains /openils/lib/perl5
#13:36:44pinesoldbs: The operation succeeded.
#13:42:20sfortin has quit IRC
#14:04:37eby__ has joined #evergreen
#14:18:11jenny1 has joined #evergreen
#14:23:41phasefx__ has joined #evergreen
#14:23:58phasefx__ is now known as phasefx3
#14:33:08jenny has quit IRC
#14:57:37rickd_ has joined #evergreen
#15:00:18senator has quit IRC
#15:10:50moodaepodbs: I'm setting up JMeter test plans for EG and was wondering if I could run them by your test concat opac?
#15:11:51dbsmoodaepo: fine with me. don't expect greatness :)
#15:12:01moodaepoOnce the plans are designed decently, I'll post it up on the wiki for benchmarking EG systems
#15:12:22moodaepodbs: Thnx
#15:13:14dbsmoodaepo++
#15:22:37moodaepodbs: might want to check if I /.ed your test server : )
#15:28:46dbsmoodaepo: I have a terrible connection at the moment, so it's all the same to me :)
#15:30:15rickd_ has quit IRC
#15:32:09dbsmoodaepo: it's alive. just slow
#15:39:45moodaepodbs: I think I might have helped it slow down, dumped 2000 users on it at the same time but doing basic gets.
#15:39:56dbshah
#15:40:16moodaepoIt handled 200 decently.
#15:40:17dbsI was just going to log on to find out where the death is, apache seems find
#15:40:41moodaepoHmm memory?
#15:41:48dbsmaybe. config is a 16GB 4-core box for apache + opensrf, and a 16GB 4-core box for the DB. there's probably a ton of postgresql processes churning away.
#15:42:43dbsyeah, 10 concurrent SELECTs
#15:43:33dbsall running search queries
#15:44:15moodaepoSo the test servers are two servers aye, I would have thought 16GB each would be able to handle that, no?
#15:46:15dbsdatabase is the bottleneck; if it was a read-only replica I could configure it to kill long-running queries, but it's the master
#16:20:13Melissa has left #Evergreen
#16:25:55rickd_ has joined #evergreen
#16:29:01dbwellsHello all. I am trying to work with adjusting some permissions and not having much luck. Specifically, I would like to allow a particular user group to register workstations.
#16:29:31dbwellsI have added the REGISTER_WORKSTATION permission to the group, and set it to the top unit type of the OU hierarchy.
#16:30:16dbwellsWhat parts of the system, if any, should be restarted to reflect permission changes such as this?
#16:30:35phasefx_one gotcha we fixed recently, the menu tries to default to staff's home_ou, which, if not in the list, will cause an error. Doesn't look like you're hitting that though :)
#16:30:36rickd_ has quit IRC
#16:31:00phasefx_as far as I know, you shouldn't have to restart anything for that
#16:31:09rickd_ has joined #evergreen
#16:32:48dbsdbwells: you mean you gave the grp_perm_map a depth of 0?
#16:32:59rickd_ has joined #evergreen
#16:33:01dbwellsphasefx_: thanks for the information.
#16:34:02dbwellsdbs: yes, that is what I mean.
#16:35:25dbwellsdbs: grp_perm_map row looks like this: 161 | 10 | 84 | 0 | f
#16:36:56dbsdbwells: I would be tempted to try clearing the staff client cache, restarting opensrf.settings, and open-ils.auth - not having poked at the particular code in question in recent memory
#16:37:43dbwellsdbs: tried all that I think, will try again.
#16:42:33pmplett is now known as pmp_afk
#16:42:48dbwellsstill no luck :(
#16:43:45dbsdbwells: what are you seeing?
#16:44:14dbwellsdbs: "You do not have permission to register a workstation."
#16:44:47dbsoh. does the user have a work_ou where they're trying to register?
#16:46:25dbspermission.usr_work_ou_map if I recall correctly
#16:47:12phasefxin srfsh, request open-ils.actor open-ils.actor.user.has_work_perm_at "d0786abfd208fc14b0c6d4c94af60c16" "REGISTER_WORKSTATION" (replace the session key)
#16:47:31phasefxthat'll show you at what orgs you effectively have that permission
#16:48:15dbs throws a dbwells in to ensure we have his attention :)
#16:48:38dbwellsdbs: thanks, just started catching up
#16:50:13dbwellsdbs: looks like you're right, there is no entry in work_ou_map
#16:50:44dbsthat would do it :) I suspect we've got zero documentation on that
#16:51:38phasefx_we should probably have the patron editor check to see if the profile group chosen has the STAFF_LOGIN permission, and either offer work org selection then and there, or at least warn about it
#16:52:13dbwellsdbs: I feel as if I am finding some important piece of the puzzle here. How is that table supposed to be populated?
#16:52:45dbsphasefx: I'd rather move towards a usr belonging to multiple perm groups and profile only being used to determine borrowing privileges. But I want ponies.
#16:53:32dbsdbwells: permission.grp_perm_map defines what levels of the hierarchy a given group has a given permission, and where it's grantable
#16:54:03phasefx_ would like a multiple perm group interface. "For that last part, may as well move away from Profile group altogether and have a Patron Type field which runs orthogonal to permission groups
#16:54:06dbspermission.usr_work_ou_map defines what OUs a given user can deploy their powers
#16:54:46phasefx_dbwells: User Perm Editor lets you set the Work Locations for a staff member
#16:55:07dbsactor.usr.home_ou only defines the user's home library for borrowing purposes, they don't get to apply their staff permissions to their home_ou
#16:56:28dbsphasefx: yeah, what you said (re multiple perm group interface)
#16:56:45dbsbecause that's what we've implemented, with SQL being our interface :)
#16:58:39phasefx_you guys are using multiple perms now? coool
#16:58:43phasefx_perm groups, rather
#16:59:24dbsphasefx: have been since day one
#16:59:26dbwellsdbs: phasefx_: thanks guys, that did the trick. I blindly assumed the home library would be the work location as well.
#16:59:47dbsdbwells: it's a common misconception, and it's not documented to my knowledge
#17:00:07dbsso don't feel bad!
#17:00:19dbsinstead, feel angry :)
#17:00:26dbwellsdbs: :D
#17:01:20dbwellsalmost angry enough to write some documentation! Just angry enough to write a note to myself to write some documentation post-migration.
#17:01:27dbshah :)
#17:04:19phasefx_so, I implement stuff in my bazaar branch, and then bring it into subversion when I'm ready. Do I commit all as one changeset, or do I cherry-pick things like utility functions that could be useful elsewhere?
#17:04:35phasefx_and commit those separately
#17:04:58phasefx_big commit makes it easier to backport without forgetting anything
#17:05:11phasefx_but makes it harder to cherry pick later if needed
#17:05:58phasefx_and makes it easier to overlook smaller changes when scrutinizing commit logs
#17:07:04phasefx_just something that's been bugging me :)
#17:08:26dbsphasefx: you could be using multiple bzr branches if you wanted to be able to keep it all separate easily-ish, yet merging them all together in one integration & testing branch
#17:09:14dbssmall changesets make it easier to follow changes - yeah
#17:09:42dbsI meant use the separate branches to generate the small changesets
#17:10:15gmcharltand avoid the dreaded changset that says does one thing but also does 2 other random things
#17:10:58phasefx_okay, so I have a bzr-svn checkout of svn trunk, and I branch that into "test". I'm using "test" currently for a largish feature. I find a typo or some small incidental thing I could tweak. I should immediately branch "test" into "small-fix", and make the change there, and then what?
#17:11:07dbsindeed!
#17:12:03sboyetteor switch branches to mail, make the small fix there, adn merge from main to "test" (warning this statement is git-flavored, but i have found git adn bzr to be generally like each other)
#17:12:08sboyettes/mail/main/
#17:16:02dbsphasefx: I would branch the bzr-svn checkout to "small-fix", make the fix there, and commit that to svn. then bzr-svn and test get the fix.
#17:16:58dbsi try to never modify bzr-svn checkout because it takes so damned long to get another clean one if I screw it up
#17:17:12eby__ has quit IRC
#17:17:46phasefx_so much work for a small fix.. right now I have some overhead in that the bzr-svn checkout is on a real server, the test branch is on that same real server, and a checkout of the test branch is on virtual Evergreen server, where I can actually test things
#17:18:30phasefx_I've stopped commiting from the bzr-svn checkout, and merge test into it just to get a diff to patch svn trunk with. Then I revert the merge in the bzr-svn checkout
#17:36:22phasefx has quit IRC
#17:51:52dchristens has quit IRC
#17:52:35phasefx has joined #evergreen
#18:00:38jenny1 has left #evergreen
#19:20:31jamesrf has joined #evergreen
#19:21:19wlayton has quit IRC
#19:28:22pmp_afk is now known as pmpett
#19:30:49jamesrf has quit IRC
#20:04:30jamesrf has joined #evergreen
#20:07:08pmpett is now known as pmplett
#21:08:24brendan_ga has joined #evergreen
#21:20:45dbs has quit IRC
#21:40:42brendan_ga has quit IRC
#21:41:37brendan_ga has joined #evergreen
#21:43:40bradl has quit IRC
#21:43:52miker_ has quit IRC
#21:43:55berick has quit IRC
#21:44:03phasefx_ has quit IRC
#21:44:18sylvar has quit IRC
#21:51:39brendan_ga has quit IRC
#21:52:40brendan_ga has joined #evergreen
#22:01:03miker_ has joined #evergreen
#22:11:33miker_ has quit IRC
#22:28:26rsinger has quit IRC
#22:29:11rsinger has joined #evergreen
< Wednesday, December 16th, 2009Raw Log FileFriday, December 18th, 2009 >