2009-03-26T01:52:23 hey all -- 2009-03-26T01:52:58 ok I've imported records via the marc2bre.pl / ingest / pg_loader dance 2009-03-26T01:53:08 and followed instructions from here http://coffeecode.net/archives/184-Unicorn-to-Evergreen-migration-rough-notes.html 2009-03-26T01:53:24 everything is in without kicking any errors 2009-03-26T01:53:39 how come I can't search for anything on the OPAC? 2009-03-26T01:54:13 do I need to identify item_types correctly somewhere first? or some other org set-up? 2009-03-26T01:54:32 do you have holdings? if you hit F5 in the staff client and give it a copy barcode, does it seem correct? 2009-03-26T01:54:51 on my mac - away from windows client 2009-03-26T01:55:21 ah 2009-03-26T01:55:36 I don't have a version of 1.4.0.2 client for mac yet 2009-03-26T01:55:42 do you have a copy barcode? you can search by item/copy barcode in the opac advanced search. 2009-03-26T01:56:00 yes I do have barcodes... I will try 2009-03-26T01:56:53 can i get to your opac, or is it internal-only? 2009-03-26T01:57:10 item barcode worked? 2009-03-26T01:57:23 yup 174.129.217.7 2009-03-26T01:57:36 that ip should get you there 2009-03-26T01:57:47 yep 2009-03-26T01:57:53 have a copy barcode you can share? 2009-03-26T01:57:58 34050100957583 2009-03-26T01:58:14 have more too 2009-03-26T01:58:16 34050100935191 2009-03-26T01:58:22 thanks 2009-03-26T01:58:47 brendan_ga: did you populate metabib.metarecord and metabib.metarecord_source_map ? 2009-03-26T01:59:05 sure is the quick.metabib.sel 2009-03-26T01:59:33 quick_metarecord_map.sql 2009-03-26T02:00:00 is that what you're looking for? 2009-03-26T02:00:04 doesn't seem to be a copy issue, so yeah, metabibs maybe. 2009-03-26T02:00:10 otherwise no. 2009-03-26T02:00:14 search is taking forever. 2009-03-26T02:00:36 yeah the search does nothing but the barcode was pretty quick for me 2009-03-26T02:00:58 could this be something to do with the way I have my org units? 2009-03-26T02:00:59 your "School Libraries" library shows but i can't expand details on it. that's... hrm. maybe it shouldn't "has volumes"? 2009-03-26T02:01:34 i still have my cgi bootstrapping open 2009-03-26T02:01:43 if you want to look... your more than welcome 2009-03-26T02:02:12 ok -- I try that change and see what happens 2009-03-26T02:03:04 probably doesn't make much sense to have "CanHaveVolumes" on your System/System2 org unit types. 2009-03-26T02:03:55 also, you have a "System2" org unit type, but nothing uses it. 2009-03-26T02:04:07 just worth noting, i don't think that's related to the problem. 2009-03-26T02:04:18 (could be, though. i'm stabbing.) 2009-03-26T02:05:01 yeah I originally went with the system 1 and system 2 to have a better display in the OPAC... but i have been messing around with that lately 2009-03-26T02:06:07 yeah, my search went nowhere. 2009-03-26T02:06:12 how much ram do you have? 2009-03-26T02:06:30 8GB 2009-03-26T02:06:52 112,000 items 97000 records 2009-03-26T02:08:01 yeah my searches go now where either 2009-03-26T02:09:00 might try restarting services? 2009-03-26T02:09:31 i've seen that "search goes nowhere" thing, i think lots of things can cause it, i don't remember the exact cause the last time i encountered it. 2009-03-26T02:09:37 actually, before restarting services... 2009-03-26T02:10:20 check /openils/var/log for errors? 2009-03-26T02:10:26 ok -- 2009-03-26T02:10:30 on the way there 2009-03-26T02:10:36 check system to see if it's chewing cpu/swap? 2009-03-26T02:10:49 check dmesg output to ensure that the kernel didn't OOM kill anything? 2009-03-26T02:11:45 in /openils/var/log which log? 2009-03-26T02:13:02 osrfsys.log is a good start. 2009-03-26T02:13:55 on a not-busy test system i sometimes tail all of the files in that dir and try my action again. 2009-03-26T02:14:03 grep ERR sometimes helps, too 2009-03-26T02:35:34 *** B_Bonner has quit IRC 2009-03-26T02:39:18 looks like i'm getting a search timed out 2009-03-26T02:39:24 was search fro 70secs 2009-03-26T02:39:34 sorry -- was searching for 70 secs 2009-03-26T02:42:33 Dan has a post on coffeecode for how to "prime" the database for searching. That might be helpfull 2009-03-26T02:45:36 ok will take a look at that 2009-03-26T03:01:27 the biggest difference that i see right away is that my relpages are 1 whereas dan's where big numbers 2009-03-26T03:01:53 40281 | metabib_author_field_entry_index_vector_idx | 1 2009-03-26T03:02:03 relfilenode | relname | relpages 2009-03-26T03:20:57 open-ils.storage.biblio.multiclass.staged.search_fts.atomic 2009-03-26T03:20:59 failes 2009-03-26T03:21:03 failed 2009-03-26T03:22:11 sorry wrong thing there 2009-03-26T03:51:55 good night thanks for the help -- still stuck - and going to bed will ask again in the morning 2009-03-26T06:00:00 *** BenO has joined #openils-evergreen 2009-03-26T08:03:23 *** phase_bb has quit IRC 2009-03-26T08:18:15 *** kgs has joined #openils-evergreen 2009-03-26T09:53:52 *** dbs has joined #openils-evergreen 2009-03-26T10:02:17 dbs: did those indexes and some analyze get your Validate going a little better? 2009-03-26T10:02:59 miker_: I don't know if you had noticed; I had already committed tag and subfield indexes as separate indexes that brought query times down from 23000 ms to 500 ms 2009-03-26T10:03:02 hrm... looking at the commit, it was just the one combined index 2009-03-26T10:03:12 goood 2009-03-26T10:03:38 but tried out your combined index, and that brought the query down to 300ms, so I committed that instead 2009-03-26T10:03:39 yes, "good" is an understatement :) 2009-03-26T10:03:39 :) 2009-03-26T10:03:43 yay for indexes 2009-03-26T10:03:54 aye 2009-03-26T10:04:07 explain_analyze++ 2009-03-26T10:04:40 interestingly, the planner is using the tpo index as well 2009-03-26T10:05:39 yeah ... bitmap-and'ed indexes are fun 2009-03-26T10:20:37 *** phasefx2 has quit IRC 2009-03-26T10:20:37 *** kgs has quit IRC 2009-03-26T10:20:37 *** rsinger has quit IRC 2009-03-26T10:20:37 *** eby has quit IRC 2009-03-26T10:20:37 *** lisppaste6 has quit IRC 2009-03-26T10:20:37 *** atheos has quit IRC 2009-03-26T10:20:37 *** berick has quit IRC 2009-03-26T10:20:37 *** jeff has quit IRC 2009-03-26T10:20:37 *** brendan_ga has quit IRC 2009-03-26T10:20:37 *** mrpeters-isl has quit IRC 2009-03-26T10:20:37 *** BenO has quit IRC 2009-03-26T10:20:38 *** greg-g has quit IRC 2009-03-26T10:20:38 *** jeff_ has quit IRC 2009-03-26T10:20:38 *** lmcfarland has quit IRC 2009-03-26T10:20:38 *** phasefx has quit IRC 2009-03-26T10:21:25 *** BenO has joined #openils-evergreen 2009-03-26T10:21:25 *** greg-g has joined #openils-evergreen 2009-03-26T10:21:25 *** jeff_ has joined #openils-evergreen 2009-03-26T10:21:25 *** phasefx has joined #openils-evergreen 2009-03-26T10:21:25 *** lmcfarland has joined #openils-evergreen 2009-03-26T10:24:08 *** phasefx2 has joined #openils-evergreen 2009-03-26T10:24:47 *** kgs has joined #openils-evergreen 2009-03-26T10:24:47 *** rsinger has joined #openils-evergreen 2009-03-26T10:24:47 *** eby has joined #openils-evergreen 2009-03-26T10:24:47 *** lisppaste6 has joined #openils-evergreen 2009-03-26T10:24:47 *** atheos has joined #openils-evergreen 2009-03-26T10:24:47 *** berick has joined #openils-evergreen 2009-03-26T10:30:10 *** jeff has joined #openils-evergreen 2009-03-26T10:30:57 miker_: is that Google Preview stuff acceptable to you now? If so, could you close ticket http://svn.open-ils.org/trac/ILS/ticket/50 (and perhaps I could backport it to rel_1_4 - obviously not rel_1_4_0) 2009-03-26T10:35:49 *** brendan_ga has joined #openils-evergreen 2009-03-26T10:35:49 *** mrpeters-isl has joined #openils-evergreen 2009-03-26T10:36:31 dbs: can we turn it on or off with an org unit setting or config file switch? in the constants file is a good first step, though 2009-03-26T10:36:44 *** mrpeters-isl has quit IRC 2009-03-26T10:36:44 *** brendan_ga has quit IRC 2009-03-26T10:36:44 miker_: yes, a config file switch is set up 2009-03-26T10:37:08 cool... then if it's not stomping the other boxen, yeah 2009-03-26T10:37:09 dbs++ 2009-03-26T10:37:34 var googleBookPreview = true; in rdetail.js 2009-03-26T10:38:08 you're going to want that in common/js/config.js aren't you? :) 2009-03-26T10:38:45 heh ... well, we need a place to put all that stuff ... but wherever for now 2009-03-26T10:39:02 does that control the result list level links too now? 2009-03-26T10:39:05 or result_common.js, where the other one is 2009-03-26T10:39:27 heh, no, the config for the result list level links is in result_common.js, not config.js :) 2009-03-26T10:40:12 herding cats! 2009-03-26T10:40:15 oh, I meant is it one switch ... seems like it'd be good, but I don't feel strongly either way 2009-03-26T10:41:06 *** atheos has left #OpenILS-Evergreen 2009-03-26T10:41:20 *** brendan_ga has joined #openils-evergreen 2009-03-26T10:41:20 *** mrpeters-isl has joined #openils-evergreen 2009-03-26T10:43:48 nope - they are separate switches, which is desirable (I think) 2009-03-26T10:44:00 *** atheos has joined #OpenILS-Evergreen 2009-03-26T10:48:05 *** mrpeters-isl has quit IRC 2009-03-26T10:48:05 *** brendan_ga has quit IRC 2009-03-26T10:48:34 *** phase_bb has joined #openils-evergreen 2009-03-26T10:49:02 does this look crazy for everyone else? http://dev.gapines.org/opac/en-US/extras/slimpac/start.html 2009-03-26T10:49:21 hah, SSI error in the dropdown 2009-03-26T10:49:30 select field is a little wide 2009-03-26T10:53:11 and a little truncated :) 2009-03-26T10:55:18 *** mrpeters-isl has joined #openils-evergreen 2009-03-26T10:55:24 appears to be an invalid xml token, giving xmlent heartburn 2009-03-26T10:55:45 oh, well, at least i know it's not just my browser now ;) 2009-03-26T11:00:38 berick: any of your org units named "&&&" or "%%%" ? 2009-03-26T11:01:22 In my dream world, we replace all of XMLENT with server-side templating like Genshi or Mako. 2009-03-26T11:01:53 TT 2009-03-26T11:02:30 dbs: yep, Library for the Blind & Physically Handicapped 2009-03-26T11:02:37 I still haven't wrapped my head around how to do good i18n with TT. :( 2009-03-26T11:02:47 bastages! 2009-03-26T11:03:15 exactly the people that would be trying to access the slimpac 2009-03-26T11:03:28 *** eby has quit IRC 2009-03-26T11:03:30 so... I need to XML-escape that when generating the org tree, then 2009-03-26T11:04:21 aha 2009-03-26T11:04:32 dbs: re:TT, I found a way to make it work, but IIR, it required compiling .mo files and not just relying on the .po files . (I have no idea if what I just said made sense) 2009-03-26T11:05:03 and this is where the answer to my commit message musing "Hmm, maybe we should add ampersize() to entityize() globally" is a resounding "yes" 2009-03-26T11:05:15 berick: that makes sense 2009-03-26T11:05:53 pocompile 2009-03-26T11:06:46 or msgfmt 2009-03-26T11:07:10 berick: lemme give a little test for that problem here, the fix should be simple 2009-03-26T11:09:33 can someone tell me how many params open-ils.actor.patron.update takes? it's it param1:authtoken param2:userid param3:actor_object ? 2009-03-26T11:18:57 or, if it's easier to answer, how bad can I screw up a production system with open-ils.actor.patron.update? :) 2009-03-26T11:20:04 atheos: parm1:auth, param2:user-object 2009-03-26T11:20:20 all you can really do is screw up that patron's account 2009-03-26T11:20:45 very good, I'll commence. 2009-03-26T11:21:21 so param1:auth (of a staff user) param2: user-object 2009-03-26T11:21:37 not the auth, for the user-object in question, right? 2009-03-26T11:21:51 correct, staff auth. 2009-03-26T11:22:18 and the evergreen extracts the userid of the user-object, from within the param2. I think I follow, thanks for the help 2009-03-26T11:22:24 the/then 2009-03-26T11:22:38 Hrm. An org_unit named "Less < More" will kill things too. 2009-03-26T11:22:49 I'll commit this fix for now and ponder deeper thoughts. 2009-03-26T11:23:39 berick, tell me, does this make sense to you? Can't call method "isnew" on unblessed reference at /openils/lib/perl5/OpenILS/Application/Actor.pm line 245. 2009-03-26T11:25:06 nevermind, I see an issue. hopefully, THE issue. 2009-03-26T11:30:40 berick: c12686 should fix that particular problem for you 2009-03-26T11:30:59 * berick updates dev 2009-03-26T11:33:22 dbs++ 2009-03-26T11:36:43 hrm... language dropdown is putting armenian first (for obvious reasons) ... perhaps the generation of the dropdown stuff should put "this" language first always? 2009-03-26T11:38:40 miker_: yes, it should. just haven't got around to that. 2009-03-26T11:39:23 (and, each language should be presented in its own language, rather than in the currently chosen language) 2009-03-26T11:40:27 ... and, I keep on wanting to hit "Select Language" to launch the search. which is pretty funny, when you get an armenian page back instead 2009-03-26T11:41:48 :) 2009-03-26T11:42:28 dbs: see the global indexes stuff on pg-hackers? 2009-03-26T11:42:33 * miker_ doing a little dance right now 2009-03-26T11:42:46 miker_: btw, have you tried searching for something with an apostrophe in slimpac? it's not good 2009-03-26T11:42:54 no - missed that! 2009-03-26T11:42:59 * dbs goes to read pg-hackers 2009-03-26T11:43:29 I haven't ... that was fixed long ago, but then the search methods changed ... probably needs some lovin' in supercat 2009-03-26T11:43:59 I briefly tried to track it down, but am easily distracted. by pg-hackers! 2009-03-26T11:45:36 weird, I don't see srinath's original post, just greg and tom's "your lawyers suck" replies 2009-03-26T11:45:52 heh 2009-03-26T11:46:04 I'll forward 2009-03-26T11:46:11 ah, probably because google groups thinks that srinath's post comes from the future (2:16 pm today) 2009-03-26T11:46:32 ok, you found it? 2009-03-26T11:46:51 forwarded anyway 2009-03-26T11:48:21 looks like supercat is munging all searches to type=keyword, too 2009-03-26T11:48:49 a search on "pickle" in series returns the same 20 results as "pickle" in author 2009-03-26T11:49:35 berick thanks for your help. I think it works now (I got a permissions error, and yup, my test account can't update staff passwords) Permission Denied [ilsevent] => 5000 [ilsperm] => group_application.user.staff.admin.global_admin [ilspermloc] => 4 2009-03-26T11:49:53 I'll make my test user an admin, and I think I've got it 2009-03-26T11:50:54 miker_++ # thanks for the forward, that does look pretty damned cool 2009-03-26T11:57:44 *** _bott_ has joined #OpenILS-Evergreen 2009-03-26T12:00:05 atheos: yr welcome, glad I could help 2009-03-26T12:03:56 a user should be able to update (some of) their own fields, including password, correct berick? 2009-03-26T12:08:51 *** brendan_ga has joined #OpenILS-Evergreen 2009-03-26T12:09:51 miker_: open-ils.actor.user.[password|username|email].update 2009-03-26T12:09:52 hey all 2009-03-26T12:10:09 I've got my records and item holdings into my database... I can search via the barcode no problem 2009-03-26T12:10:39 but when I search via keyword or author or something - nothing returns 2009-03-26T12:10:42 brendan_ga: yeah, saw the post 2009-03-26T12:10:52 I've checked the dsmeg 2009-03-26T12:10:59 dmesg 2009-03-26T12:11:14 and not seeing anything die 2009-03-26T12:11:22 but I see it time out in the logs 2009-03-26T12:11:36 do you have anything in metabib.title_field_entry? 2009-03-26T12:11:46 will take a look 2009-03-26T12:12:00 select count(*) from metabib.title_field_entry; would do. 2009-03-26T12:12:20 also, not sure if you mentioned what version you had installed. 2009-03-26T12:12:50 yes I do have entries in the metabib.title_field_entry 2009-03-26T12:12:56 1.4.0.2 stable 2009-03-26T12:13:01 of EG 2009-03-26T12:13:38 I believe I've been following your instructions... 2009-03-26T12:14:00 I don't think I have a swap -- so I'm looking into that 2009-03-26T12:15:15 a swap? 2009-03-26T12:15:19 I've gotta go get food 2009-03-26T12:16:28 swap partion 2009-03-26T12:17:33 I was following along the lines of looking for a swap partition for memory usage.. 2009-03-26T12:19:22 dbs for the select count(*) from metabib.title_field_entry - I have 114,000 entries 1 row 2009-03-26T12:21:25 how about metabib.metarecord? 2009-03-26T12:24:00 yup 88,000 1 row 2009-03-26T12:27:42 *** BenO has quit IRC 2009-03-26T12:31:00 *** eby has joined #OpenILS-Evergreen 2009-03-26T12:56:36 brendan_ga: sounds good. did you make any changes to actor.org_unit_type or actor.org_unit? 2009-03-26T12:57:09 yes - but before I imported items to the database... 2009-03-26T12:58:34 I have consortium, system, system2, branch1, branch2... 2009-03-26T12:58:51 doesn't matter when you made the changes, if the org_unit_type depths don't match the position of the org_units in the hierarchy. 2009-03-26T12:59:11 everything but consortium can have users and copies 2009-03-26T13:00:31 that's an answer to a different question :) 2009-03-26T13:04:03 brendan_ga pasted "org_units - BG" at http://paste.lisp.org/display/77634 2009-03-26T13:07:54 and org_unit_types? 2009-03-26T13:07:57 brendan_ga annotated #77634 "org_units_types-BG" at http://paste.lisp.org/display/77634#1 2009-03-26T13:10:05 you have no org_unit_types with depth 1 2009-03-26T13:10:05 *** Meliss has joined #openils-evergreen 2009-03-26T13:11:12 thus, the hierarchy is broken. 2009-03-26T13:13:21 *** sarabee_ has quit IRC 2009-03-26T13:13:26 ok -- will I be able to change this now? 2009-03-26T13:14:18 update actor.org_unit_type set depth = depth - 1 where depth > 0; 2009-03-26T13:19:07 you'll probably have to restart opensrf and run autogen again after that 2009-03-26T13:23:36 trying some searches 2009-03-26T13:30:24 no luck -- still spins away on the keyword and author, title searches 2009-03-26T13:40:42 phasefx: I'm going to backport your "staff client: default to right org unit in deep hierarchy" patch to 1.4.1 / 1.4.0; it's working well for me on 1.4.0 and definitely closes a bug. 2009-03-26T13:40:50 http://svn.open-ils.org/trac/ILS/changeset/12482 2009-03-26T13:43:33 *** sarabee has joined #openils-evergreen 2009-03-26T13:50:41 berick (or any else) using open-ils.actor.patron.update to update user account, by editing patron object extracted from open-ils.actor.user.fleshed.retrieve (works great). I'm passing along all of the comments inside of json. Would it be ok for me to omit those comments? 2009-03-26T13:51:20 atheos: like "/* S */" or something like that? 2009-03-26T13:51:25 yes 2009-03-26T13:51:37 I know they serve a function, just not sure if it applies here 2009-03-26T13:51:52 I don't think that's okay. in the OpenSRF 0.9 days, those signified class names / payloads 2009-03-26T13:52:32 'course, it might work :) 2009-03-26T13:52:53 hmm, ok. my php json functions barf on those comments. I'm doing some very nasty stuff in order to update the password, and feed it back into open-ils.actor.patron.update 2009-03-26T13:53:48 yeah, that comment method predated JSON standardization - ergo the change in OpenSRF 1.0. 2009-03-26T13:54:02 *** eby_ has joined #OpenILS-Evergreen 2009-03-26T13:54:14 yea, I knew that. I guess I'll just have to do the uglyness until we are on 1.4 2009-03-26T13:54:59 atheos++ # awesome to have people like you, jeff, and bott hacking away on EG 2009-03-26T13:55:14 hmm, well I might test it without the comments. worse I can do (hopefully) is screw up my own account 2009-03-26T13:55:15 brendan_ga: I dunno, man, I've run out of ideas 2009-03-26T13:55:30 ha, hopefully I don't mess anything up along the way 2009-03-26T13:55:41 dbs -- thanks so -- much I'll keep working along and see what I can find... 2009-03-26T13:55:57 I think i should be able to see something in the logs.. eventually 2009-03-26T13:56:18 brendan_ga: yeah, you might want to turn statement logging on in postgresql and follow the trail 2009-03-26T13:56:34 thanks dbs 2009-03-26T13:57:07 *** rsinger_ has joined #OpenILS-Evergreen 2009-03-26T14:02:50 *** eby- has joined #OpenILS-Evergreen 2009-03-26T14:08:33 brendan_ga annotated #77634 "error from postgres logs" at http://paste.lisp.org/display/77634#2 2009-03-26T14:09:25 I'm also seeing all my records have format electronic / multimedia... so I'll need to change that too... 2009-03-26T14:09:55 atheos: how much access do you have to your evergreen server? how committed are you to doing this in php? 2009-03-26T14:10:23 atheos: i ask, because creating a new opensrf call for staff to reset patron passwords might be easier than what you're trying to do. :) 2009-03-26T14:10:47 jeff It may be, but I've already got it working. I just have to polish it up 2009-03-26T14:10:58 atheos: or, having your php call a perl script which made the call to the full-on open-ils.actor.patron.update 2009-03-26T14:11:01 brendan_ga: not to worry about, that's just clark-kent.pl saying those tables don't exist. doesn't affect searching. 2009-03-26T14:11:17 ok 2009-03-26T14:11:23 atheos: oh, right. just caught that you had it working now. wasn't done reading scroll. yay! 2009-03-26T14:12:22 *** rsinger has quit IRC 2009-03-26T14:12:42 jeff, it's ugly though. but, I'm just trying to get something functional, so that I can move onto the next project. I'd like to do this in the OPAC, but I'm limited on time, and I don't want to support this for more than our own library at the moment. you probably know, but our system is shared. 2009-03-26T14:13:07 * jeff nods 2009-03-26T14:13:58 screw that other library! :) 2009-03-26T14:14:04 *** jamesrf has joined #openils-evergreen 2009-03-26T14:14:34 atheos pasted "my ugly password update function in php" at http://paste.lisp.org/display/77637 2009-03-26T14:16:28 *** eby has quit IRC 2009-03-26T14:19:23 *** eby_ has quit IRC 2009-03-26T14:23:51 *** natschil has joined #openils-evergreen 2009-03-26T14:24:51 *** rsinger_ is now known as rsinger 2009-03-26T14:29:17 Are there any standard things that I should have set in EG before I imported records... I don't know like item form map or something? 2009-03-26T14:29:31 all I did was the org_units 2009-03-26T14:31:35 brendan_ga: as long as your shelving locations, statuses in use, org units and copies are all marked opac-visible you should see results. Does the staff client search return results? 2009-03-26T14:32:10 only for "harry potter" nothing else 2009-03-26T14:33:32 really bad cache maybe? I would be tempted to shut down apache, opensrf, and memcached; then start up in reverse order 2009-03-26T14:34:06 all my results are even by the item barcode search are showing as electronic resources 2009-03-26T14:34:16 sorry bad grammer 2009-03-26T14:34:25 here's a better attempt. 2009-03-26T14:36:29 all results show a format of software, multimedia 2009-03-26T14:36:46 ok -- will stop and start in order 2009-03-26T14:38:02 ok maybe close out I'm going to restart everything.... 2009-03-26T14:38:32 sorry wrong window! 2009-03-26T14:43:55 <_bott_> is there still an ingest problem in 1.4.0.3? 2009-03-26T14:44:12 _bott_: which ingest problem is that? 2009-03-26T14:45:06 <_bott_> I did a z39 import, got the biblio.record_entry fine, but no metabib entries. Did a MARC edit, re-saved, still no entries. I see the ingest call, but nada 2009-03-26T14:46:09 _bott_: that should be covered, yes. 1.4.0.4 will be cut tonight and (wonder of wonders!) exposed on the downloads page an wiki 2009-03-26T14:46:09 <_bott_> I remember dbs mentioning something along these lines a week or so ago, but didn't note any follow up 2009-03-26T14:46:24 s/an/and 2009-03-26T14:46:39 <_bott_> I'll check the commits, see if I can fix this temporarily. thanks! 2009-03-26T14:46:44 _bott_: fwiw, we're still running into problems on our test server, but miker_ doesn't believe me :) 2009-03-26T14:47:01 s/miker_ doesn't believe me/I haven't produced a test case/ 2009-03-26T14:47:09 dbs: I believe something's happening :) 2009-03-26T14:47:25 <_bott_> It's odd, I see the ingest call, but it doesn't seem to do anything, but also doesn't complain 2009-03-26T14:47:49 we're getting ingest, but edits to the marc record aren't being saved; and I've turned off statement logging and can't turn it back on until the quiet night 2009-03-26T14:48:15 and of course, the problem goes away after you restart the services. for a while :) 2009-03-26T14:48:22 <_bott_> You mean the marc isn't getting saved? 2009-03-26T14:48:43 hah! foiled by our own app deployment. restarted my test vm, 1.2.4.0 staff client was re-installed. 2009-03-26T14:49:15 right, if I edit a MARC record and save it, the staff client says "success!" but the bre.marc field is the same as it ever was. 2009-03-26T14:49:33 <_bott_> I haven't been able to reproduce that since the $no_ingest addition, unfortunately, now I really am getting "No ingest" 2009-03-26T14:50:07 same thing on staff client - can get results on search "harry potter" but nothing else. 2009-03-26T14:59:01 hello, just wondering ... how does evergreen encrypt passwords? 2009-03-26T15:01:57 natschil: one-way md5 hash in database 2009-03-26T15:03:00 thanks 2009-03-26T15:03:01 NEW.passwd = MD5( NEW.passwd ); 2009-03-26T15:07:22 dbs: btw, have you heard anything about that race condition bug in 1.4.0.2? I haven't been able to get those logfiles yet, sorry, so I was hoping someone else might've had the same bug.... 2009-03-26T15:28:09 where are we pulling the item_format from ? 2009-03-26T15:29:39 brendan_ga: from the MARC after being transformed to MODS (if you mean on the result list display) 2009-03-26T15:30:14 ok from the 008 field? 2009-03-26T15:34:09 brendan_ga: are you referring to the type of material icons in particular? 2009-03-26T15:36:31 brendan_ga: if so, those come from a combination of type (ldr/06) and bib level (ldr/07) based on the table at the very bottom of this page: http://www.oclc.org/bibformats/en/fixedfield/type.shtm 2009-03-26T15:38:56 ok -- i don't have a 007 field 2009-03-26T15:38:57 if you mean the Electronic Resources in the Formats dropdown in the search interfaces, that is a combination of record type (ldr/06) and item form (008/[23|29]) 2009-03-26T15:39:02 I have them in 008 2009-03-26T15:39:33 sorry wrong thing -- I need to look at my LDR and not the 008 2009-03-26T15:39:37 one sec 2009-03-26T15:40:41 by ldr/06 I mean leader position 6. sorry if that was unclear 2009-03-26T15:41:28 ok for one of my examples I have the LDR position 7 as "a" but for book -- but showing a computer resource for icon 2009-03-26T15:42:07 another interesting thing too -- I found and checked-in an item "zug the bug" 2009-03-26T15:42:13 what's the next position? 2009-03-26T15:42:17 now I can find that in my OPAC but before I couldn't 2009-03-26T15:42:49 sorry, previous position 2009-03-26T15:43:49 ldr/7 is bib level (monographic component part) 2009-03-26T15:44:13 could ldr/6 be m? 2009-03-26T15:45:04 brendan_ga: paste your opac url for this record -- it may help. :) 2009-03-26T15:45:15 jeff++ 2009-03-26T15:45:21 on the phone 2009-03-26T15:49:31 LDR 01955am a2200445a 44500 2009-03-26T15:49:37 i think something's shifted. 2009-03-26T15:50:18 LDR 00574nam 22001815 4500 is an example from my test box 2009-03-26T15:51:01 http://174.129.217.7/opac/en-US/skin/default/xml/rdetail.xml?r=13856&l=4&d=2 2009-03-26T15:51:26 It's showing all of the LDR's as m 2009-03-26T15:51:29 on all the records... 2009-03-26T15:51:41 doesn't match my original 2009-03-26T15:52:08 @ 01352cam a22004214a 4500 2009-03-26T15:52:31 off by one on the leaders... something forgot they were zero indexed, perhaps? 2009-03-26T15:52:33 vs. EG one - LDR 01265am a2200409a 44500 2009-03-26T15:55:14 *** jamesrf has quit IRC 2009-03-26T15:55:52 ok so how can I go about fixing this.... 2009-03-26T15:56:02 That's pretty crazy. 2009-03-26T15:56:07 here's another example of a difference 2009-03-26T15:56:11 EG = LDR 01570am a2200325a 44500 2009-03-26T15:56:19 brendan_ga: where were your bibs exported from? 2009-03-26T15:56:32 From Horizon 2009-03-26T15:56:33 and did you convert them to XML before processing with marc2bre? 2009-03-26T15:56:45 no 2009-03-26T15:56:48 brendan_ga: and can you go through the various import staging files to see where the LDR was mangled, at what step? 2009-03-26T15:56:57 jeff+ 2009-03-26T15:56:59 jeff++ 2009-03-26T15:57:15 sure will start taking a look now... 2009-03-26T15:59:48 somewhere you're losing your LDR/05, which is shifting 06 and 07 over, and breaking all manner of things. :) 2009-03-26T16:01:13 (so your ldr/07 m for monograph is being moved over to ldr/06 where m means computer file) 2009-03-26T16:02:19 yes -- I've looked at my bre file and it seems to have happened right then 2009-03-26T16:02:27 the "somehow" is the interesting thing 2009-03-26T16:02:38 brendan_ga: are the records MARC8 coming out of Horizon? or UTF8? 2009-03-26T16:03:02 MARC8 - using MARCedit i transfer them to UTF8 2009-03-26T16:03:38 oh, interesting. And they're okay coming out of MARCedit? 2009-03-26T16:03:56 I've had tangles with it in the past, although more from the UTF8 -> MARC8 direction. 2009-03-26T16:04:07 going back to compare them all right now -- be a moment 2009-03-26T16:05:50 yup it's getting mangled from .mrc to .bre 2009-03-26T16:06:04 looks good in .mrc 2009-03-26T16:07:59 nope it's mangled from MARCedit... 2009-03-26T16:08:39 yaz-marcdump -f marc8 -t utf8 -i marc -o marcxml horizon.mrc > horizon_utf8.xml 2009-03-26T16:08:45 bre is coping the info correctly... I'm losing it somewhere from horizon to running the first step... 2009-03-26T16:11:35 oh hey, i don't CARE that supercat's marcxml output doesn't set 901|c like WWW:Exporter does, because i already have the record id! 2009-03-26T16:11:44 (it's the little realizations) 2009-03-26T16:14:13 heh 2009-03-26T16:14:32 brendan_ga: seriously, try yaz-marcdump. much nicer. doesn't mangle data (much) 2009-03-26T16:14:51 ok -- I'm off to another part of the building to analize my out put... 2009-03-26T16:15:04 dbs I will use the yaz-marcdump... thanks 2009-03-26T16:15:29 might as well use it to convert to marcxml at the same time and save a processing step in marc2bre 2009-03-26T16:16:08 yup -- ok I'll be back in a few ... 2009-03-26T16:16:29 need to reset the database... 2009-03-26T16:19:06 have to shut down the entire park to do that. don't forget to turn the perimeter fences back on... 2009-03-26T16:28:36 hrm. staff client MARC View is showing me ‡ instead of ‡ (using html entities for clarity here, don't know what the underlying markup looks like. 2009-03-26T16:28:42 public opac seems fine. 2009-03-26T16:29:46 *** natschil has quit IRC 2009-03-26T16:30:08 jeff: i fixed that instance 2009-03-26T16:30:39 but haven't squashed the variation in the merge records interface, in which you get %u2021 instead :) 2009-03-26T16:30:47 dbs: ah, so this has been fixed? good deal. thanks! 2009-03-26T16:30:51 heh 2009-03-26T16:39:51 *** eby- is now known as eby 2009-03-26T16:42:49 dbs++ # charset/encoding fixin' 2009-03-26T16:43:38 *** phase_bb has quit IRC 2009-03-26T16:43:41 <_bott_> miker_: dbs: my ingest problem was local. Didn't like a custom entry in config.metabib_field and was dying quietly mid-ingest 2009-03-26T16:43:54 *** phase_bb has joined #openils-evergreen 2009-03-26T16:44:02 _bott_: interesting - something you had tailored? 2009-03-26T16:45:07 <_bott_> an entry to include a 9xx into metabib data, which I used for some item importing long ago 2009-03-26T16:45:18 ahh. interesting 2009-03-26T16:51:27 *** Meliss has quit IRC 2009-03-26T16:51:31 so hopefully we'll get 1.4.0.4 wrapped up with a bow and we can get some people banging on it and I'll just be a freak out in the wilderness :) 2009-03-26T16:55:20 <_bott_> I don't know you well enough to make the comments I've considered 2009-03-26T17:10:00 *** jamesrf has joined #openils-evergreen 2009-03-26T17:20:00 _bott_: I am relatively thick-skinned, and any insult can be salved with the application of beer 2009-03-26T17:23:33 plus, freaks in the wilderness generally don't care 2009-03-26T17:29:02 <_bott_> yeah, but when you find that freak that does care, you end up with... the Unibomber! 2009-03-26T17:35:55 berick: you've got mail 2009-03-26T17:36:13 if it smells like gunpowder, never you mind 2009-03-26T17:38:42 dbs: at my esi acct? 2009-03-26T17:40:04 dbs: just a bad joke about a letter bomb 2009-03-26T17:40:16 although I did send some scanned materials via email 2009-03-26T17:40:18 *** GlenM has quit IRC 2009-03-26T18:00:54 *** dbs has quit IRC 2009-03-26T18:45:09 *** eby has quit IRC 2009-03-26T19:26:18 *** greg-g has quit IRC 2009-03-26T19:31:23 *** jamesrf has quit IRC 2009-03-26T19:37:46 <_bott_> ugh, IE doesn't like my 1.4 skin, and reports a syntax error on an non-existent line ...when will the world be rid of this trash? 2009-03-26T20:02:31 *** djfiander has joined #OpenILS-Evergreen 2009-03-26T20:11:01 *** jamesrf has joined #openils-evergreen 2009-03-26T20:26:42 *** jamesrf has quit IRC 2009-03-26T20:27:16 *** phase_bb has quit IRC 2009-03-26T20:46:10 using yaz-marcdump how can I add a subfield to the output (916a). 2009-03-26T20:46:32 yaz-marcdump -f marc8 -t utf8 -i marc -o marcxml 2009-03-26T20:48:17 you don't. yaz-marcdump doesn't alter the contents of the marc records. it just changes the format. 2009-03-26T20:49:13 the best way to do something like that is probably using the perl MARC module. 2009-03-26T20:49:31 - ok -- need to check my records ( i thought I had a 916a field in there.. 2009-03-26T21:07:57 *** djfiander has quit IRC 2009-03-26T21:19:16 *** djfiander has joined #OpenILS-Evergreen 2009-03-26T21:26:35 *** dbs has joined #openils-evergreen 2009-03-26T21:27:34 miker_: just realized upgraders won't get the are (tag, subfield) index; probably not critical 2009-03-26T21:29:49 dbs: I can add that 2009-03-26T21:29:56 still adjusting the tag for release 2009-03-26T21:30:00 version numbers and such 2009-03-26T21:30:18 miker_++ 2009-03-26T21:31:24 ooh, good catch. 2009-03-26T21:31:47 *** phase_bb has joined #openils-evergreen 2009-03-26T21:31:49 ... of my own fumble 2009-03-26T21:35:00 details, details. 2009-03-26T21:35:52 *** greg-g has joined #openils-evergreen 2009-03-26T21:38:30 dbs: just the tag,subfield index, yes? that's all you added? 2009-03-26T21:38:54 yessir 2009-03-26T21:48:56 * miker_ taps fingers waiting on `for i in cs-CZ en-CA fr-CA hy-AM; do make LOCALE=$i install; done` 2009-03-26T21:53:27 mighty big of you to do it so that the poor EG users don't have to 2009-03-26T21:53:41 ha 2009-03-26T21:56:03 oh, I know - I could add that bit to the global make install, and watch phasefx explode as he just tries to build the client 2009-03-26T22:44:58 *booom* 2009-03-26T22:50:59 poor phasefx 2009-03-26T22:57:28 download page and wiki updated 2009-03-26T22:57:35 it's Official(tm) 2009-03-26T23:04:01 *** djfiander has quit IRC 2009-03-26T23:09:30 miker_++