2013-03-25T01:40:35 *** jcamins_ has joined #evergreen 2013-03-25T01:40:47 *** Rish has joined #evergreen 2013-03-25T01:41:58 *** misilot1 has joined #evergreen 2013-03-25T01:42:41 *** phasefx__ has joined #evergreen 2013-03-25T01:46:20 *** jcamins has quit IRC 2013-03-25T01:46:21 *** misilot has quit IRC 2013-03-25T01:46:21 *** phasefx_ has quit IRC 2013-03-25T02:17:25 *** Mark__T has joined #evergreen 2013-03-25T02:19:12 *** Rish has quit IRC 2013-03-25T02:19:39 *** Rish has joined #evergreen 2013-03-25T03:03:59 *** Rish has quit IRC 2013-03-25T04:06:55 *** Rish has joined #evergreen 2013-03-25T04:26:12 *** Rish has quit IRC 2013-03-25T04:40:34 *** rbecker has quit IRC 2013-03-25T04:41:00 *** Mark__T has quit IRC 2013-03-25T04:41:01 *** Mark__T has joined #evergreen 2013-03-25T04:41:03 *** phasefx__ has quit IRC 2013-03-25T04:43:16 *** rbecker has joined #evergreen 2013-03-25T04:44:37 *** phasefx_ has joined #evergreen 2013-03-25T04:49:04 *** rbecker has quit IRC 2013-03-25T04:51:08 *** rbecker has joined #evergreen 2013-03-25T05:17:38 *** Mark__T has quit IRC 2013-03-25T05:44:34 *** Mark__T has joined #evergreen 2013-03-25T06:43:25 *** Rish has joined #evergreen 2013-03-25T06:51:27 *** Rish has joined #evergreen 2013-03-25T07:00:20 *** timf has joined #evergreen 2013-03-25T07:42:41 *** misilot1 is now known as misilot 2013-03-25T07:56:57 is there a perm we need to grant to allow patrons to add items to lists in tpac? 2013-03-25T08:00:36 csharp: CREATE_MY_CONTAINER 2013-03-25T08:02:35 ok - thanks 2013-03-25T08:14:03 *** akilsdonk_ has joined #evergreen 2013-03-25T08:18:03 *** misilot has quit IRC 2013-03-25T08:19:59 *** collum has joined #evergreen 2013-03-25T08:26:58 *** DPearl has joined #evergreen 2013-03-25T08:32:05 *** kmlussier has joined #evergreen 2013-03-25T08:44:48 *** Dyrcona has joined #evergreen 2013-03-25T08:44:48 *** gdunbar has joined #evergreen 2013-03-25T08:49:58 afternoon 2013-03-25T08:51:25 *** robertl has joined #evergreen 2013-03-25T08:53:11 *** Callender_ is now known as Callender 2013-03-25T08:57:35 *** Meliss has joined #evergreen 2013-03-25T09:04:27 *** AaronZ-PLS has joined #evergreen 2013-03-25T09:09:05 It's there a library setting or global flag that allows an added field in a merged record to show in its own separate line, and not like this: 035 ‡a(OCoLC)54321 ‡a(DCL)12345 2013-03-25T09:14:10 *** akilsdonk_ has quit IRC 2013-03-25T09:21:27 *** yboston has joined #evergreen 2013-03-25T09:24:16 *** Mark__T has left #evergreen 2013-03-25T09:26:00 *** akilsdonk_ has joined #evergreen 2013-03-25T09:29:33 *** goooood is now known as eeevil 2013-03-25T09:42:25 *** krvmga has joined #evergreen 2013-03-25T09:42:55 *** Mark__T has joined #evergreen 2013-03-25T09:48:39 i want to make sure i understand something correctly. 2013-03-25T09:48:54 in the OPAC, i search for mark twain and get results 2013-03-25T09:49:56 then under Personal Author, i click "Paine, Albert Bigelow" and this gets me all the books about Mark Twain written by Paine 2013-03-25T09:50:06 it does not get me all books written by Paine 2013-03-25T09:50:28 is this the correct result? that the facets give me sub-searches of my original search? 2013-03-25T09:50:31 that's my understanding of it too, yes 2013-03-25T09:50:38 krvmga: That's right. It's a facet for refining the search that you already conducted. 2013-03-25T09:50:55 most excellent. just wanted to be sure. thx! 2013-03-25T09:50:58 :) 2013-03-25T09:55:16 *** hopkinsju has quit IRC 2013-03-25T09:58:08 then, before or after selecting the facet, you could select the author name from one of the individual search results, and that would give you a new author search based on the author name you selected. 2013-03-25T09:58:14 *** jeff_ is now known as jeff 2013-03-25T09:58:47 *** shadowspar has quit IRC 2013-03-25T10:04:26 *** mllewellyn has joined #evergreen 2013-03-25T10:15:50 *** ellie has joined #evergreen 2013-03-25T10:22:08 *** shadowspar has joined #evergreen 2013-03-25T10:22:36 In the OPAC, a search timeout looks identical to a "no results" search. I'm trying to find a way to detect the difference so I can inform the user, but I have not found it yet. Any suggestions? 2013-03-25T10:24:25 *** mrpeters has joined #evergreen 2013-03-25T10:24:58 morning guys....is there any harm in disabling open-ils.auth_proxy if we have no plans to use LDAP or any other authentication method 2013-03-25T10:25:56 by disable i mean just commenting it out in the opensrf.xml 2013-03-25T10:26:27 I think the OPAC might use it, not sure, though. 2013-03-25T10:26:40 orly...good to know 2013-03-25T10:27:43 seeing a bunch of this -- http://pastie.org/7111284 -- in logs 2013-03-25T10:38:04 mrpeters: You can always turn it off and see if patrons start complaining. :) 2013-03-25T10:38:33 heh yeah not in a gambling situation today 2013-03-25T10:41:54 *** ktomita has joined #evergreen 2013-03-25T10:45:28 mrpeters: I think you're safe to turn off open-ils.auth.proxy if you're not using ldap. 2013-03-25T10:45:41 ok, thanks 2013-03-25T10:45:47 http://coffeecode.net/egdocs/2.2/_authentication_proxy.html made it seem that way 2013-03-25T10:46:02 I just looked at OpenILS::WWW:Proxy::Authen that is used by a number of modules and it uses open-ils.auth 2013-03-25T10:48:08 Gave me another reason to rebase PhoneList.pm anyway. :) 2013-03-25T10:49:55 *** zerick has joined #evergreen 2013-03-25T10:50:33 *** gmcharlt_ is now known as gmcharlt 2013-03-25T10:51:01 *** gmcharlt has joined #evergreen 2013-03-25T10:51:27 thanks Dyrcona 2013-03-25T10:52:27 *** mmorgan has joined #evergreen 2013-03-25T10:55:47 *** Mark__T has quit IRC 2013-03-25T10:57:04 mrpeters: unless you turned it on, auth_proxy should already be off. The log messages you are seeing is the login code probing to see whether it is enabled, and currently those are going to happen no matter what. 2013-03-25T10:57:19 ok, so its noise 2013-03-25T10:57:30 yes 2013-03-25T10:57:56 hello. Does anyone happen to know where I can find the number of libraries that use Evergreen? 2013-03-25T10:58:21 I have seen the map Bob created earlier, but it does not seem to provide the total number and also looks outdated 2013-03-25T10:58:23 log_noise-- 2013-03-25T10:58:38 There's a lot of noise in the logs. 2013-03-25T10:58:54 ellie: probably not the best place to ask. you might try the general list serve. 2013-03-25T11:00:28 *** afterl has joined #evergreen 2013-03-25T11:01:27 mrpeters: It would actually be less noisy if the service were running with the 'enabled' flag set to false, but with something as important as login, I think not running at all is a good conservative default. 2013-03-25T11:01:39 fair enough. thanks. 2013-03-25T11:03:03 enabled is set to false in our case 2013-03-25T11:03:57 Apache2::RequestIO::print: (103) Software caused connection abort at /usr/lib/perl5/Template.pm line 176, referer: http://www.gapines.org/eg/opac/home 2013-03-25T11:04:09 anyone seen that^^? 2013-03-25T11:06:54 mrpeters: It does concern me a bit that all that log data happens twice. Is that a problem more generally with your log, or is it specific to this event? If the second, I may need to look into that a little more. 2013-03-25T11:09:01 should we have a listener/drone for open-ils.auth_proxy if its disabled? 2013-03-25T11:10:45 mrpeters: will do, thanks 2013-03-25T11:12:12 * Dyrcona regrets all the legacy junk brought over from Horizon. 2013-03-25T11:13:11 * dbs muses about cutting over to bcrypt with an upgrade script that just brute forces md5 hashes 2013-03-25T11:18:44 * dbs promotes http://evergreen-ils.org/dokuwiki/doku.php?id=dev:hackfest:eg2013 again - sign up for the hackfest! add interesting / fun hacks that you think people might want to work on! 2013-03-25T11:18:50 dbs++ 2013-03-25T11:18:54 * Dyrcona muses about shoving PKI down everyone's throat since any one-way hash can be brute forced given time and CPU cycles. 2013-03-25T11:19:04 berick++ # for actually having added stuff :) 2013-03-25T11:19:13 * Dyrcona takes a look. 2013-03-25T11:19:16 dbs: i'll be there! 2013-03-25T11:19:20 paxed++ # you too 2013-03-25T11:19:53 csharp: That software caused connection abort thing, we see that all the time. Basically it tends to boil down to someone clicking to another page or moving on without actually waiting for the system to give them the page they were looking for. 2013-03-25T11:20:29 * berick wonders if we could get anything useful done with auth/bcrypt at the hackfest 2013-03-25T11:21:12 csharp: Sort of like, they go to a TPAC page, and instead of waiting for results or for the whole page to be loaded, they click on another link, key in a new address, or otherwise leave. 2013-03-25T11:21:22 dbs: have you had a look at the db-seed-i18n script improvement i did? 2013-03-25T11:21:25 berick: mebbe. try dropping it in at the server level and see how things blow up at the JavaScript level. Oh javascript, you so slowwwwww 2013-03-25T11:21:34 Generally just noise. Though it's kind of made me wonder what's so slow that they quit. 2013-03-25T11:21:42 paxed: no, saw the bug go by but that was about it 2013-03-25T11:21:52 time approaching 0 2013-03-25T11:22:19 bshum: thank you 2013-03-25T11:22:45 dbs: These are ideas for presentations or mini-sessions or just what you want to hack on? 2013-03-25T11:23:10 can't hurt to add it.. /me does so 2013-03-25T11:23:45 Dyrcona: a mix. the mini-session / presentation stuff is essentially learning from previous hackfests that lots of people show up with enthusiasm but need an intro to get productive 2013-03-25T11:24:04 ok, then. 2013-03-25T11:24:13 Will throw up a few things.... ;) 2013-03-25T11:24:27 Ideally, we would be able to use lightning talks to report on the [2,3,4] cool things we accomplished during actual hacking too 2013-03-25T11:24:37 Dyrcona++ 2013-03-25T11:25:28 Curse you page locking! ;) 2013-03-25T11:25:38 Dyrcona: try now 2013-03-25T11:25:43 Now I know how the database feels. ;) 2013-03-25T11:28:29 *** hopkinsju has joined #evergreen 2013-03-25T11:42:39 mrpeters: sorry to disappear for a bit. Double-checking in master, it looks like I was wrong about the level at which it is disabled by default. In a stock config, the service runs, but 'enabled' is false. If yours isn't running (as opposed to running but not 'enabled'), that would account for some of the noise, but it is certainly designed to fail gracefully on the client end either way. 2013-03-25T11:43:13 ok, so we could just comment it out of opensrf.xml to prevent it from s tarting and creating that noise? 2013-03-25T11:48:27 mrpeters: If things are working properly, the least noise will happen in the default config, where the service can successfully check for 'enabled'. If your service is actually running and you still get that 503, that concerns me. 2013-03-25T11:48:44 [Attached]root@brick03-head: ps aux | grep auth_proxy 2013-03-25T11:48:44 opensrf 17903 0.0 0.1 148712 41672 ? Ss 10:50 0:00 OpenSRF Listener [open-ils.auth_proxy] 2013-03-25T11:48:44 opensrf 17959 0.0 0.1 148716 40584 ? S 10:50 0:00 OpenSRF Drone [open-ils.auth_proxy] 2013-03-25T11:48:44 opensrf 18163 0.0 0.1 148712 40616 ? S 10:51 0:00 OpenSRF Drone [open-ils.auth_proxy] 2013-03-25T11:48:51 seems like its running 2013-03-25T11:50:06 mrpeters: I'm not well versed on multi-brick setups, but the errors are from brick06-head. Do we know it is running there as well? 2013-03-25T11:50:13 yep 2013-03-25T11:52:13 mrpeters: ok, I will assume then at this point there is some small bug in the way the 'enabled' check happens, I'll add it to my TODOs. Thanks for bringing it up. 2013-03-25T11:52:26 no problem. ping me if you need more inforation 2013-03-25T11:52:36 mrpeters: thanks, I will 2013-03-25T11:52:40 *** jdouma has joined #evergreen 2013-03-25T11:52:42 or email -- mpeters@emeralddata.net 2013-03-25T12:10:10 *** mcarlson has joined #evergreen 2013-03-25T12:15:09 *** ellie has quit IRC 2013-03-25T12:24:51 *** mcooper has joined #evergreen 2013-03-25T12:30:10 *** smyers_ has joined #evergreen 2013-03-25T12:30:13 *** misilot has joined #evergreen 2013-03-25T12:50:28 *** krvmga has quit IRC 2013-03-25T12:52:48 *** bradl_ is now known as bradl 2013-03-25T13:07:16 *** pmurray_away is now known as pmurray 2013-03-25T13:10:08 anyone have any documentation on recovering from an offline session that died in the middle of processing? 2013-03-25T13:10:54 * mrpeters thinks services got stopped in the middle of processing, leading to 500 internal server errors 2013-03-25T13:19:27 *** j_scott has joined #evergreen 2013-03-25T13:20:35 *** mcarlson__ has joined #evergreen 2013-03-25T13:21:05 *** mcarlson has quit IRC 2013-03-25T13:22:19 nevermind....found the log files. i will just create a new session after importing them. 2013-03-25T13:22:46 for the logs --- /openils/var/data/offline//pending/orgunitid/sessionkey/ 2013-03-25T13:25:28 *** kmlussier has quit IRC 2013-03-25T13:28:25 *** Rogan_ has joined #evergreen 2013-03-25T13:35:40 *** dkyle has joined #evergreen 2013-03-25T13:39:32 Does anyone else have 2.4 test box running? 2013-03-25T13:39:56 Rogan_: I have master as of 20130319, that's pretty close to 2.4. 2013-03-25T13:40:07 I'm seeing video formats in the advanced search duplicate and curious if it's a "my system" or actual bug issue. 2013-03-25T13:40:43 Rogan_: Do you mean in the drop down? 2013-03-25T13:40:57 Exactly. 2013-03-25T13:41:26 Rogan_: I'll have a look as soon as my current rebuild finishes. 2013-03-25T13:41:40 Appreciate it. 2013-03-25T13:41:56 I removed one of three testing branches that I also added. 2013-03-25T13:42:20 *** jdouma_ has joined #evergreen 2013-03-25T13:44:05 Takes a while to build all of the partial updates. :( 2013-03-25T13:45:28 *** jdouma has quit IRC 2013-03-25T13:45:32 Bah, that's what multitasking is for. 2013-03-25T13:45:46 Me, I'm learning more about the TPAC than I ever wanted to know right now. 2013-03-25T13:46:51 are you learning that it is awesome :) 2013-03-25T13:47:25 It's very nice I admit, I just didn't want to get into the customization this much. 2013-03-25T13:47:53 Rogan_: Well, I'll clear out my /openils/var/updates directory later. That'll save some time on future builds. :) 2013-03-25T13:48:13 :) 2013-03-25T13:48:13 Rogan_: feedback on what you think the defaults should be would be awesome, to spare others the same pain 2013-03-25T13:48:31 * mrpeters doesnt seem to have duplicated rows in Video Format on my master 2013-03-25T13:48:55 but i dont have a dropdown...i have a selection box 2013-03-25T13:49:08 Rogan_: I don't see anything doubled on advanced search. 2013-03-25T13:49:34 Rogan_: we have doubled entries in advanced search. I suspect it could be related to upgrading from previous versions 2013-03-25T13:49:40 Interesting. I'll put it on my "look into later" rather than "worry about documenting bug" list then. 2013-03-25T13:49:52 http://laurentian.concat.ca/eg/opac/advanced?loc=105 2013-03-25T13:50:17 * dbs will now poke, as it might help two sites :) 2013-03-25T13:50:29 Ah, so it isn't just me. 2013-03-25T13:50:40 dbs: You don't have blu-ray doubled. I'll bet it does have something to do with upgrades. 2013-03-25T13:51:01 yep, double entry in config.videorecording_format_map 2013-03-25T13:51:31 The same things are doubled on mine and same things that aren't, aren't. 2013-03-25T13:53:49 Same. One is opac_visible TRUE, the other line is FALSE 2013-03-25T13:54:10 So it keeps both options in the config.coded_value_map 2013-03-25T13:54:22 * dbs doesn't have opac_visible in his production tables yet. hmm. 2013-03-25T13:54:35 Maybe it's just a master thing 2013-03-25T13:54:51 *** Rogan_ has quit IRC 2013-03-25T13:54:54 But that's the only difference between two lines in my table where the value was like "D-2" 2013-03-25T13:55:19 bshum: But is master ignoring the OPAC Visible flag? Do both show up in your advanced search? 2013-03-25T13:55:35 * Dyrcona decides to check his database rather than speculate. 2013-03-25T13:55:37 Dyrcona: I have to enable the field to find out, we don't use that option in our advanced search. 2013-03-25T13:55:51 I was just following that yes, there do appear to be dupe entries 2013-03-25T13:57:01 In config.coded_value_map there's an opac_visible column, and th vr_format entries are all duped for us with opac_visible = TRUE 2013-03-25T13:57:35 *** Rogan_ has joined #evergreen 2013-03-25T13:57:40 Make it easy to dedupe though, as theer is a synthetic key on the table 2013-03-25T13:58:50 I have no dupes in config.coded_value_map and opac_visible is apparently ignored by advanced search, at least. 2013-03-25T13:59:07 And no, I'm not looking at this in the staff client. I'm actually using the OPAC. 2013-03-25T13:59:41 Betacam and Betacam SP have opac visible false, but they show up in my opac, just for the record. 2013-03-25T13:59:56 *** misilot1 has joined #evergreen 2013-03-25T14:01:24 Rogan_: A related question, how is search performance on your 2.4 test system? 2013-03-25T14:01:54 looks like dbwells fixed this in https://bugs.launchpad.net/evergreen/+bug/1091831 no? 2013-03-25T14:01:54 Launchpad bug 1091831 in Evergreen "Inconsistent config.update_coded_value_map() in codebase" (affected: 1, heat: 6) [Medium,Fix committed] 2013-03-25T14:01:56 *** misilot has quit IRC 2013-03-25T14:02:02 *** misilot1 is now known as misilot 2013-03-25T14:03:03 Search has been good. 2013-03-25T14:03:26 It's noticably faster to me but we're loaded on SSD drives too so I don't know how much is upgrade and how much is that. 2013-03-25T14:03:35 or maybe unfixed it... 2013-03-25T14:03:57 Rogan_: indeed, that will help. and newer PG 2013-03-25T14:04:10 Rogan_: Are you testing with production data or with Concerto? 2013-03-25T14:04:10 eeevil: yeah, I would say no, pending further investigation. and it might depend on what fixes were applied when, in what order, yada yada 2013-03-25T14:04:28 using production data, copy is only about two weeks old in fact 2013-03-25T14:05:00 Rogan_: Thanks for the input. I'm not happy with search performance on my development machine at the moment. It seems to be worse to me. 2013-03-25T14:05:06 So, it might be fixed might not but we upgraded from 2.3 so it might inherited so to speak from a past upgrade. 2013-03-25T14:05:51 DELETE FROM config.coded_value_map ccvm WHERE exists (SELECT 1 FROM config.coded_value_map c1 WHERE c1.id < ccvm.id AND c1.ctype = ccvm.ctype AND c1.value = ccvm.value AND c1.opac_visible = ccvm.opac_visible); -- might work out 2013-03-25T14:06:21 Dyrcona: I've found that our production database with new QP isn't doing so hot with broad searches. Like looking up "Connecticut" in our test systems. 2013-03-25T14:06:44 Or like "flowers" 2013-03-25T14:06:47 Or "white" 2013-03-25T14:07:00 Basically crummy unfocused searches for keyword, etc. 2013-03-25T14:07:04 (and with that, now our advanced search has no dupes in video format or language any more. yay) 2013-03-25T14:07:09 Usually just spin and spin forever till timeout 1 minute later. 2013-03-25T14:07:14 "It" 2013-03-25T14:07:14 I'm doing author searches mainly. 2013-03-25T14:07:37 dbs: Yeah "it" was definitely broken now. 2013-03-25T14:07:54 * Dyrcona tries it as title search. 2013-03-25T14:08:28 We've been holding back from new QP in production while we kicked the tires in our test systems. 2013-03-25T14:08:32 bshum: that was my fear re moving away from the LIMIT-restricted (if imperfect) core search query with secondary visibility checks to tsbere's reimplementation of an all-SQL search 2013-03-25T14:08:49 bshum: that's a good datapoint to have, thank you for testing and reporting 2013-03-25T14:09:49 anyone going to attend https://www.pgcon.org/2013/schedule/events/611.en.html ? 2013-03-25T14:09:53 (for history purposes, it as all-SQL before staged_fts (which was then followed by query_parser_fts), and we moved to the restricted, 2-stage search because of performance) 2013-03-25T14:11:04 does TPAC support lib_ips.txt? has the formatting changed? 2013-03-25T14:11:05 dbs: I wonder if I could talk my boss into sending either tsbere or myself. 2013-03-25T14:11:12 dbs: I haven't checked out the whole talks schedule yet, but I've been thinking about that one. 2013-03-25T14:11:35 * dbs wishes the description would get completed some day soon. 2013-03-25T14:12:51 Wow! I should go for BSDCan and stay for PGcon. If I take a week or two of vacation in between, I could just travel from Vancouver to Ottawa. 2013-03-25T14:13:20 Ok, a month of vacation.... ;) 2013-03-25T14:13:27 heh 2013-03-25T14:15:07 Wow. (author: stephen king && title: it) or whatever that comes out to from the advanced search is not working either. 2013-03-25T14:15:11 "Database Development with Git, Sqitchin, and pgTAP" would be an awesome tutorial, don't think I'm going to make it in time for that though 2013-03-25T14:15:12 *** Rogan_ has quit IRC 2013-03-25T14:15:41 Oh. it just finished. on the second try, btw. 2013-03-25T14:15:49 I think I signed up for that one. And some others. 2013-03-25T14:16:07 cached results from the first time actually completing in the background, I assume 2013-03-25T14:16:14 bshum++ 2013-03-25T14:16:16 dbs: yes. 2013-03-25T14:16:25 mrpeters: it does.. see http://ln-s.net/+tQO -- PerlSetVar OILSRedirectTpac "true" 2013-03-25T14:19:26 huh. "Squitchin" seems to only appear in that talk description. assuming it's either a typo or something unreleased related to http://sqitch.org/ (which I think dbs has mentioned before) 2013-03-25T14:19:54 yep, was called sqitch last year. 2013-03-25T14:24:03 Dyrcona / bshum / others (Rogan_ left...): I'm considering the removal of the all-SQL change to the QP changes, bringing back the search.query_parser_fts() stored proc ... I'd hate to have massive speed regressions in search for 2.4. it was claimed that testing shoed the all-sql version was "as fast or faster" but that seems to be untrue for your datasets ... please keep the datapoints coming. in particular, if 2.3 is showing a big difference for 2013-03-25T14:24:03 you on the same major version of PG, that is very important. as you have time, of course. 2013-03-25T14:24:28 s/shoed/showed/ 2013-03-25T14:24:35 thanks berick 2013-03-25T14:24:50 eeevil: You got cut off at "in particular, if 2.3 is showing a big difference for..." 2013-03-25T14:25:10 bshum: thanks: " in particular, if 2.3 is showing a big difference for you on the same major version of PG, that is very important. as you have time, of course." 2013-03-25T14:25:24 *** bmills has joined #evergreen 2013-03-25T14:26:08 fwiw, we've been using PG 9.1 for both pre- and post- new QP. That's been our basic point in master testing so far. 2013-03-25T14:26:24 eeevil: I'm considering backing it out on a dev branch and reloading my database from production to see the difference on my dev machine. 2013-03-25T14:26:37 berick: our lib_ips.txt file is in place and that line is uncommented, but IPs aren't redirecting... how would one troubleshoot that? 2013-03-25T14:26:42 The difference between my dev machine and production is night and day. 2013-03-25T14:27:04 bshum: I'm using Pg 9.2 in development. 2013-03-25T14:27:18 "libarian by day, party animal by night" 2013-03-25T14:27:46 Dyrcona: I'll see if I can whip up a branch that positively reverts the changes (bring back the code instead of rewriting history) if that would be helpful. it's how I'd want to do it for master 2013-03-25T14:28:08 Dyrcona: and, just the all-SQL to 2-stage bits, not the nesting and whatnot 2013-03-25T14:28:38 berick: do you know if there are relevant log entires to show whether or not lib_ips.txt is being looked at? 2013-03-25T14:28:46 *** blongwel has joined #evergreen 2013-03-25T14:29:07 csharp: are you using anything in front of apache, such as a pound proxy? 2013-03-25T14:29:22 just ldirector 2013-03-25T14:30:06 eeevil: Ok. I'll be happy to load that. I was just going to make a new branch based on master and revert the relevant commits for testing. 2013-03-25T14:30:07 csharp: nevermind then. a proxy will complicate client-ip based things, but that doesn't apply if you're not using a proxy. :-) 2013-03-25T14:30:40 Dyrcona: hopefully just a logical inversion of 6d8872cf120caf67ad6f65995b2c5155fa5ab652 and ab9fb958e387a20cfe9fafb6035fb72bc5f1fb3f 2013-03-25T14:30:40 [evergreen|Thomas Berezansky] Remove dependence on search.query_parser_fts proc - 2013-03-25T14:30:41 [evergreen|Thomas Berezansky] Remove search.query_parser_fts from schema - 2013-03-25T14:31:12 mrpeters: see WWW/Redirect.pm ... there are several log lines in there 2013-03-25T14:31:30 eeevil: I'll try reverting those two and see what happens. 2013-03-25T14:31:34 "csharp" at 204.193.129.146 pasted "our settings for redirection" (14 lines) at http://paste.evergreen-ils.org/18 2013-03-25T14:32:01 *** asimon has joined #evergreen 2013-03-25T14:35:44 I'm testing the TPAC under 2.3.4. Search results screen do not show all of the items. For example, searching a title with 92 hits shows only items 1 and 2 on the first screen (although it says 1-10 of 92), items 11, 12, and 13 on the second screen, and so on. Is anyone aware of this TPAC problem? 2013-03-25T14:36:14 asimon: that symptom is consistent with records lacking a 901$c value -- is this your first foray into TPAC? 2013-03-25T14:36:32 this should be a FAQ! 2013-03-25T14:36:55 * dbs goes 2013-03-25T14:37:21 jeff: Yes. Let me check the records for a 901|c and I'll report back. 2013-03-25T14:37:59 dbs++ 2013-03-25T14:38:05 dbs++ 2013-03-25T14:38:51 ok, so it seems "Default and force patrons to TPac" is what breaks lib_ips.txt redirection 2013-03-25T14:40:06 *** rbecker has quit IRC 2013-03-25T14:41:24 Dyrcona: there are a goodly number of conflicts on the first commit there... I'm working through them, then I'll push a (untested) branch for you to compare with yours 2013-03-25T14:42:03 eeevil: Ok. I noticed the conflicts. I was just looking through log to see if another commit needs reversion first. 2013-03-25T14:42:09 *** rbecker has joined #evergreen 2013-03-25T14:45:25 jeff: OK, so I did a search for "Hunt for Red October" in the JSPAC, which resulted in the correct display of 18 hits. The same search in TPAC resulted in 18 hits, but display of only two (one on each page). All the records have a 901|c. 2013-03-25T14:51:03 Hi, I was curious where I would go to edit the size of the User Params lists in the reports module. For instance, some of our library names and shelving locations get cut off in the scrollable list of choices and we'd like to widen them so that the names could fit or be horizontally scrollable if possible? 2013-03-25T14:52:16 *** bkuhn has joined #evergreen 2013-03-25T14:53:26 jeff: Once I select a record to view, I can use the Previous and Next links to move back and forth among the 18 records. 2013-03-25T14:53:27 Dyrcona: perl syntax passes, at least, in working/collab/miker/revive-qp-fts 2013-03-25T14:54:16 eeevil: Ok. I'll give that a look. 2013-03-25T14:54:29 Dyrcona: the conflict resolution went different directions for each, because of post-commit drift 2013-03-25T14:55:13 eeevil: All right, then. I'll use your branch instead of my own. I just dropped the HEAD code without knowing. 2013-03-25T14:55:23 Dyrcona: I /believe/ you could 1) put the stored proc back in your db and 2) install the perl parts of that branch and avoid reloading all the data 2013-03-25T14:55:36 fyi: e0623504fb0466e line +RedirectMatch 301 ^/$ /eg/opac/home is what breaks lib_ips.txt 2013-03-25T14:55:36 [evergreen|Thomas Berezansky] Default and force patrons to TPac - 2013-03-25T14:55:48 or, you'd know quickly (by failed queries) if that won't be enough, and it should be quicker 2013-03-25T14:56:22 mrpeters: That makes sense now that I think about it. Couldn't have both ^/$ pointing at different things. 2013-03-25T14:56:26 simply commenting it out seemed to allow lib_ips.txt to do its redirection again. i wonder if that line is even needed. base URL still redirects to TPAC 2013-03-25T14:56:53 Dyrcona: perl, and the bits in opensrf.xml that handle core query limit and superpages 2013-03-25T14:57:09 eeevil: I'll give that a whirl. 2013-03-25T14:57:14 thanks 2013-03-25T14:57:57 bmills: I think that'd be some sort of development change (there isn't an option to change the size of those views on the report interface page). 2013-03-25T14:58:09 Probably have to look for the page which generates the page. 2013-03-25T14:59:06 mrpeters: it might be useful to add to the comment for the redirector that you should comment out the ^/$ redirect above it if you're using it. 2013-03-25T14:59:20 +1 2013-03-25T14:59:28 not a bad idea jeff 2013-03-25T14:59:49 mrpeters: i don't think it matters if the non-OpenILS::WWW::Redirect redirect for ^/$ is pointing at tpac or jspac. 2013-03-25T15:00:13 yeah, i think thats what we found 2013-03-25T15:01:06 should i bug this, or just push a branch to working with that comment 2013-03-25T15:01:15 bshum: thanks 2013-03-25T15:04:03 *** ericar has joined #evergreen 2013-03-25T15:08:18 * Anyone have any ideas (other than a missing 901|c) why records are not displaying in TPAC search results but do display when moving between record screens? 2013-03-25T15:10:03 asimon, we got the same problem before. And after fixing missing 901 records, the issue was solved. 2013-03-25T15:10:30 Maybe it's not just that they're not missing but they're mismatched somehow? Do the bib record IDs for the 901c match the actual bib in question? 2013-03-25T15:11:09 (is just spit-balling, has no clue why else this would break having never seen them broken personally) 2013-03-25T15:12:14 mrpeters: I think either way would work. It's a small enough change that I'd be okay signing off and pushing the line through with a new comment. 2013-03-25T15:12:30 That said, feel free to open a bug ticket for it and include your code if you're not ready to do it right away. 2013-03-25T15:12:36 Then we might not forget it later. 2013-03-25T15:12:43 bshum: i just made a bug, ill let someone who can more eloquently explain the problem add the comment 2013-03-25T15:13:12 * mrpeters doesnt have a vm handy to push code anyway at the moment 2013-03-25T15:15:27 bshum: Yes, the 901|c fields match the 001 fields in each record. 2013-03-25T15:23:19 asimon: it should match the id field on biblio.record_entry 2013-03-25T15:24:37 *** timf has quit IRC 2013-03-25T15:24:50 *** rbecker is now known as RBecker 2013-03-25T15:30:21 *** bkuhn is now known as bkuhnIdle 2013-03-25T15:31:48 *** ericar has quit IRC 2013-03-25T15:32:08 *** asimon has quit IRC 2013-03-25T15:32:33 *** robertl has quit IRC 2013-03-25T15:40:26 eeevil: I am going to have to try your changes on a fresh database. 2013-03-25T15:40:59 Dyrcona: boo ... what errors are you seeing? it may be my conflict resolution.. 2013-03-25T15:41:53 eeevil: I don't see any errors, but I'm not sure there was any speed up. That could just be my dev server being slow. 2013-03-25T15:42:55 Dyrcona: ah, cool. so, you go results, at least ... I'm curious if you captured any of the search queries, mainly to confirm that you're getting what we'd expect at the db level 2013-03-25T15:43:39 eeevil: No, I didn't but the list of the results are so far that same as on our production server. 2013-03-25T15:44:41 eeevil: Would that be in the postgres logs or elsewhere? 2013-03-25T15:45:42 Dyrcona: if you're logging "slow" queries, then in the PG log, yep. otherwise, no. you can grep the PG logs for "bib search" to find the starting line number if they're there 2013-03-25T15:46:02 eeevil: Ok. 2013-03-25T15:46:04 we mark searches with an SQL comment for easier findification 2013-03-25T15:50:53 eeevil: Nope. We are apparently not logging slow queries in our development database. 2013-03-25T15:53:14 *** kmlussier has joined #evergreen 2013-03-25T15:57:33 *** Meliss has quit IRC 2013-03-25T16:13:54 I will try the code on clean data tomorrow. 2013-03-25T16:16:00 *** pdurbin has joined #evergreen 2013-03-25T16:16:33 good job with the faceted search 2013-03-25T16:18:26 *** smyers__ has joined #evergreen 2013-03-25T16:18:28 *** bkuhnIdle is now known as bkuhn 2013-03-25T16:20:58 *** smyers_ has quit IRC 2013-03-25T16:22:17 *** akilsdonk_ has quit IRC 2013-03-25T16:26:38 hmm, from a quick look at your code it seems like you've implemented faceted search within postgres as opposed to, say, solr or elastic search 2013-03-25T16:35:34 *** ktomita_ has joined #evergreen 2013-03-25T16:37:24 *** ktomita__ has joined #evergreen 2013-03-25T16:37:54 *** ktomita has quit IRC 2013-03-25T16:40:26 *** ktomita_ has quit IRC 2013-03-25T16:44:43 *** afterl has left #evergreen 2013-03-25T16:52:14 *** jdouma has joined #evergreen 2013-03-25T16:53:41 *** collum has quit IRC 2013-03-25T16:55:08 *** jdouma_ has quit IRC 2013-03-25T16:56:16 *** bmills has quit IRC 2013-03-25T16:58:24 *** ktomita has joined #evergreen 2013-03-25T17:00:53 *** ktomita__ has quit IRC 2013-03-25T17:02:48 *** artunit has quit IRC 2013-03-25T17:02:59 *** artunit has joined #evergreen 2013-03-25T17:07:10 *** ktomita_ has joined #evergreen 2013-03-25T17:09:11 *** kmlussier has quit IRC 2013-03-25T17:10:16 *** alexlazar has joined #evergreen 2013-03-25T17:10:16 *** ktomita has quit IRC 2013-03-25T17:10:51 *** mrpeters has left #evergreen 2013-03-25T17:17:31 *** mmorgan has left #evergreen 2013-03-25T17:18:19 *** dbwells has quit IRC 2013-03-25T17:18:35 *** dbwells has joined #evergreen 2013-03-25T17:24:59 *** Dyrcona has quit IRC 2013-03-25T17:48:14 *** mllewellyn has left #evergreen 2013-03-25T17:52:44 *** zerick has quit IRC 2013-03-25T18:01:03 *** stevenyvr2 has joined #evergreen 2013-03-25T18:13:10 *** hopkinsju has quit IRC 2013-03-25T19:06:52 *** smyers__ has quit IRC 2013-03-25T19:21:44 *** bkuhn has quit IRC 2013-03-25T19:27:48 *** yboston has quit IRC 2013-03-25T19:53:49 *** jdouma has quit IRC 2013-03-25T20:14:08 *** stevenyvr2 has left #evergreen 2013-03-25T20:24:58 *** mcooper has quit IRC 2013-03-25T20:38:36 *** blongwel has quit IRC 2013-03-25T21:39:58 *** pmurray is now known as pmurray_away 2013-03-25T21:42:44 *** mcarlson__ has quit IRC 2013-03-25T21:57:40 *** pmurray_away is now known as pmurray 2013-03-25T22:03:04 *** pmurray is now known as pmurray_away 2013-03-25T22:26:27 *** RBecker is now known as rbecker 2013-03-25T22:35:42 *** zerick has joined #evergreen 2013-03-25T22:37:41 *** zerick has quit IRC 2013-03-25T23:45:09 *** mcarlson has joined #evergreen 2013-03-25T23:48:49 pdurbin: that's right... we're one of the few projects that uses PostgreSQL full-text search, as well as home-rolled faceting (mostly rolled by that eeevil character)