← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
| Time | Nick | Message |
|---|---|---|
| 00:03 | NateC joined #koha | |
| 00:09 | smeagol joined #koha | |
| 00:17 | smeagol | feesh..nice fish.. |
| 00:24 | devap joined #koha | |
| 00:25 | devap left #koha | |
| 00:47 | Dyrcona joined #koha | |
| 01:41 | pastebot | "wajasu" at 127.0.0.1 pasted "zebra facet yaz-client result" (58 lines) at http://paste.koha-community.org/174 |
| 01:43 | wajasu | looking into facets. |
| 01:44 | dcook | Did you set up zebra::facet::title:w in your zebra config? |
| 01:45 | You could try zebra::facet::title:0 since you're using ICU | |
| 01:48 | wajasu | i've been putting <retrieval syntax="xml" name="zebra::facet::title:w"/> into ... zebradb/retrieval-info-bib-dom.xml |
| 01:48 | dcook | Also, wajasu, I don't think we actually have a "title" index set |
| 01:48 | You might try "Title" | |
| 01:48 | Caps might matter | |
| 01:48 | * dcook | can't remember the actual index he tried out in October |
| 01:48 | wajasu | i tried, but will try again. |
| 01:49 | got the same thing with Title. let me try 0 | |
| 01:51 | 0 instead of w get | |
| 01:52 | got error | |
| 01:52 | dcook | That's a bit bizarre |
| 01:53 | Let's see...I could probably take a few minutes to try this again | |
| 01:53 | I really need to remember to set up a dev machine one of these days.. | |
| 01:55 | wajasu | i even tried the retreival lines from comment 6 from bug 11232 |
| 01:55 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11232 new feature, P5 - low, ---, gmcharlt, NEW , Retrieve facets from Zebra |
| 01:56 | dcook | Hold on. I'm just getting it set up now :p |
| 02:01 | Hmm | |
| 02:03 | wajasu | maybe i forgat to put something in the yazgfs section |
| 02:03 | dcook | Mmm, I doubt it |
| 02:03 | Maybe I forgot to put something in instructions | |
| 02:03 | wajasu: Are you using a package install? | |
| 02:04 | wajasu | yes |
| 02:04 | dcook | Have you tried koha-zebra-restart? |
| 02:04 | wajasu | well i lied. |
| 02:04 | dcook | lol |
| 02:05 | wajasu | i created a VM and did koha packages to get dependencies, but then to a kohaclone dev install |
| 02:05 | the VM in about 5 week old. | |
| 02:05 | in/is | |
| 02:05 | dcook | In any case, if you can restart zebra itself, try that |
| 02:07 | wajasu | i have clean db with 3 records i pulled and caltaloged that have "heart" in the title |
| 02:07 | i've rebuild_zebra and restart zebrasrv | |
| 02:07 | dcook | That's not what I said to do :p |
| 02:07 | Ah | |
| 02:08 | I'm thinking that there is something missing in my directions.. | |
| 02:09 | wajasu | my base biblios find @attr 1=4 heart gets 4 hits. |
| 02:09 | its the elements line | |
| 02:10 | dcook | Yes, that would be it |
| 02:11 | wajasu | element zebra::facet::any:w,title:w,title:0 ends up giving me: |
| 02:11 | [25] Specified element set name not valid for specified database -- v2 addinfo 'zebra::facet::any:w,title:w,title:0' | |
| 02:12 | dcook | Yeah, same. |
| 02:12 | wajasu | maybe my retrieval should be combined |
| 02:12 | dcook | Same for any combo for me at the moment |
| 02:12 | O_o | |
| 02:13 | wajasu | i hope its not because its squeeze and an older zebrasrv yaz-client |
| 02:13 | when you were hacking could it have been a newer compiled version? | |
| 02:13 | dcook | Well, I'll try it on my wheezy vm in a minute |
| 02:13 | Well could've been | |
| 02:13 | Facetting is only supported from Zebra 2.... | |
| 02:13 | 2.1 | |
| 02:14 | Wait.. | |
| 02:14 | wajasu | zversion says 3 |
| 02:14 | dcook | Maybe 2.0.20 |
| 02:14 | wajasu | but that may be the default |
| 02:20 | got yaz-client version 4.0.11 | |
| 02:21 | option gives: | |
| 02:21 | search present delSet triggerResourceCtrl scan sort extendedServices namedResultSets | |
| 02:25 | I just recently tried: | |
| 02:25 | base biblios | |
| 02:26 | format xml | |
| 02:26 | wahanui | format xml is a client side transformation of whatever comes back from the server. |
| 02:26 | wajasu | elements zebra::facet::itype:w |
| 02:26 | f @attr 1=8031 BK | |
| 02:27 | and got 7 hits. then did a show and got a facet xml response, but text was G/A | |
| 02:28 | just grabbed 8031 from the ccl.properties mapping | |
| 02:29 | dcook | G/A ? |
| 02:30 | wajasu | is see this in the log |
| 02:30 | 21:16:20-19/03 zebrasrv(1) [log] dict_lookup_grep: (\x01G\x01\x02)/A\x01\x01\x01\x06\x01\x01\x01\x06 | |
| 02:30 | there is a G and /A in that string | |
| 02:31 | pastebot | "wajasu" at 127.0.0.1 pasted "zebra log" (9 lines) at http://paste.koha-community.org/175 |
| 02:32 | wajasu | facet first phase real=0.00 cat=index might be a lead |
| 02:35 | pastebot | "wajasu" at 127.0.0.1 pasted "zebra log -v all" (188 lines) at http://paste.koha-community.org/176 |
| 02:39 | * mtompset | is glad dcook and wajasu are talking zebra and facets. :) |
| 02:54 | dcook | Aha! |
| 02:54 | wajasu | Wha? |
| 02:55 | dcook | I was wondering why it wasn't working on my Debian VM, but it's because I was pointing to the wrong file... |
| 02:56 | Alas, that's for package installs | |
| 02:57 | So not really relevant to the task at hand | |
| 02:57 | wajasu | ok. be right back |
| 02:58 | wajasu joined #koha | |
| 03:00 | dcook | wajasu: I'm quite confident that the problem is with restarting Zebra |
| 03:00 | Double-check your koha-conf.xml to make sure that retrieval-info-bib-dom.xml is the correct file | |
| 03:01 | wajasu | ok |
| 03:02 | kathryn joined #koha | |
| 03:05 | dcook | So yeah...I think the problem is making sure you're changing the correct file, and then making sure that Zebra is restarting correctly |
| 03:15 | mtompset | dcook: less kohaclone/debian/scripts/koha-restart-zebra |
| 03:15 | look for the lone daemon line. ;) | |
| 03:15 | ^lone^long^ | |
| 03:15 | dcook | Yeah, I already looked there |
| 03:16 | I don't have root access for my dev install in any case, so it looks like I'll have to ask someone else to do it | |
| 03:16 | mtompset | You don't have root access for your dev install?! |
| 03:17 | dcook | It's not my server :p |
| 03:17 | Someone else runs it | |
| 03:17 | I just provide the code :p | |
| 03:17 | mtompset | So then... why don't you have a VM for your own development? |
| 03:18 | dcook | I don't really have my own development |
| 03:18 | But I do have a VM and I use packages and gitify :p | |
| 03:18 | And we wind back up at the beginning where I have to ask someone else to restart zebra in this case :p | |
| 03:18 | mtompset | I have 13 VM's to play with. :) |
| 03:18 | * mtompset | shouts, "MORE VMS!" |
| 03:19 | grins. | |
| 03:59 | dcook | Ahhh! |
| 03:59 | I'm awesome! | |
| 03:59 | * dcook | does a little dance |
| 03:59 | dcook | It was so obvious in the end |
| 03:59 | Gahhh | |
| 04:00 | Too much excitement. Not enough outlet... | |
| 04:00 | wizzyrea | hehe |
| 04:00 | dcook | We're doing our Zebra config all wrong for facets |
| 04:00 | wizzyrea | that is unsurprising to me. |
| 04:01 | dcook | Initial testing suggests that it'll be so easy to access Zebra's special element set |
| 04:01 | http://www.indexdata.com/yaz/d[…]ls.retrieval.html | |
| 04:01 | It specifies that "name" is optional... | |
| 04:01 | Wait... | |
| 04:01 | I need to double-check things | |
| 04:01 | * dcook | halts the celebration |
| 04:01 | dcook | I might be an idiot still |
| 04:03 | Nope! I was right! | |
| 04:03 | * dcook | continues dancing |
| 04:04 | dcook | We just need "<retrieval syntax="xml"/>" at the bottom within <retrievalinfo> |
| 04:04 | Then we regain access to the Zebra special retrieval elements | |
| 04:04 | That we lost by using <retrievalinfo> in the first place | |
| 04:05 | hdl posed the question in June 2009 and now there is an answer :) | |
| 04:30 | @later tell wajasu take a look at bug 11232 again. I've updated the instructions. | |
| 04:30 | huginn | dcook: The operation succeeded. |
| 04:32 | wajasu | trying out |
| 04:33 | dcook | It only fixes one of the problems though it appears.. |
| 04:33 | pastebot | "dcook" at 127.0.0.1 pasted "Zebra output on non-package install" (25 lines) at http://paste.koha-community.org/177 |
| 04:35 | "dcook" at 127.0.0.1 pasted "Zebra facet output on package install" (51 lines) at http://paste.koha-community.org/178 | |
| 04:40 | dcook | Hmm, I'm using DOM for both of them.. |
| 04:40 | Not sure about ICU | |
| 04:41 | Interesting... | |
| 04:41 | wahanui | hmmm... interesting is sometimes good and sometimes bad |
| 04:41 | wajasu | my koha-conf has a bunch of retrieval entries as well as the included one. |
| 04:42 | dcook | Yeah, those are fallbacks |
| 04:42 | It looks like my package install is non-ICU, while my tarball install is ICU | |
| 04:45 | wajasu | i haven't had any success yet. |
| 04:45 | dcook | I doubt you will while using ICU |
| 04:45 | wajasu | remind me how to turn off icu |
| 04:45 | dcook | Although if you replaced all the specific zebra::facet::*:* type entries with just a <retrieval syntax="xml"/> you should be able to do any facet you want in yaz |
| 04:46 | You'll need to change default.idx | |
| 04:46 | Possibly somewhere else. I'm not sure. | |
| 04:46 | I do have an idea though.. | |
| 04:47 | wajasu | now things make sense. those index w and index p definitions in there link stuff up |
| 04:48 | dcook | Link stuff up? |
| 04:49 | wajasu | the index w, and index p, must have to do with zebra::facet::Title:p and such |
| 04:49 | dpk joined #koha | |
| 04:50 | wajasu | can u paste your default.idx |
| 04:50 | dcook | Yeah, "zebra::facet::Title:p" would be a facet for the title phrase index |
| 04:53 | wajasu | i suppose i can do another make and choose chr |
| 05:03 | dcook | Ooo |
| 05:03 | wajasu: With your ICU install, try a facet of "itype:w" | |
| 05:03 | That's working for me | |
| 05:04 | As does pubdate | |
| 05:05 | wth... | |
| 05:05 | Suddenly title is working now too | |
| 05:06 | wajasu: Try something like "elements zebra::facet::Title:w,itype:w,pubdate:w,rtype:w" with ICU | |
| 05:06 | For some reason, it's working great | |
| 05:06 | Oh I wonder.. | |
| 05:07 | Oh wait...it's because I'm an idiot | |
| 05:07 | Probably at least | |
| 05:07 | wajasu | i haven't had any luck yet |
| 05:08 | dcook | I switched back to non-ICU and that's why it was working..derp derp |
| 05:09 | wajasu | ok i'll try switching back to non-ICU |
| 05:09 | dcook | Use "charmap word-phrase-utf.chr" instead of "icuchain words-icu.xml" in default.idx for whatever index type you want to try |
| 05:09 | Still...it's not a great solution if it doesn't work for ICU.. | |
| 05:09 | wajasu | i'll copy my default.idx fromt he package install into my kohadev |
| 05:09 | dcook | That's certainly one way to do it, if your package install doesn't use ICU |
| 05:10 | Doesn't hurt to understand what's going on in the config files though :p | |
| 05:11 | Of course, the reason it might not seem to work with ICU might be because the terminal isn't expecting utf8.. | |
| 05:14 | wajasu | i am thinking its a character conversion thing as well |
| 05:15 | dcook | Hmm, I just told yaz-client to use UTF-8 and that's not working.. |
| 05:15 | wajasu | maybe inputcharset outputcharset stuff |
| 05:15 | dcook | Yeah, I changed the output |
| 05:17 | Hmm, none of those options are doing much | |
| 05:23 | wajasu | http://lists.indexdata.dk/pipe[…]nuary/001859.html |
| 05:25 | http://lists.indexdata.dk/pipe[…]April/002190.html | |
| 05:25 | dcook | Interesting |
| 05:25 | wahanui | i think Interesting is sometimes good and sometimes bad |
| 05:25 | dcook | Gotta love how their bugzilla links are broken.. |
| 05:26 | wajasu | :) |
| 05:28 | dcook | Would you post those links on Bugzilla? |
| 05:28 | Probably makes more sense to focus on ElasticSearch and/or Solr anyway :p | |
| 05:30 | the :0 thing might fix this but I'm not sure.. | |
| 05:31 | wajasu | i did get facet counts working, where i retrieve all the results records (marcxml) and inspect hidden etc. but what if the system is big like 1million records. |
| 05:32 | dcook | How do you mean that you got facet counts working? |
| 05:32 | As in, at the Perl code level? | |
| 05:33 | Indexdata claims that Zebra special retrieval elements should work faster than using marcxml | |
| 05:33 | wajasu | yes. the other day. i have a bug 11909 that i was working with mtompset |
| 05:33 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11909 major, P5 - low, ---, matted-34813, ASSIGNED , Fix hidelostitems, OpacHiddenItems total count, prog them facet display |
| 05:33 | mtompset | -- which by the way is on my "get back to quickly" list. |
| 05:34 | I should go. I need sleep. | |
| 05:34 | Have a great day, #koha wajasu dcook. :) | |
| 05:34 | dcook | Then I'm not sure of the relevancy |
| 05:34 | ta mtompset | |
| 05:34 | I suspect that querying Zebra for facets from a million results would be a lot faster than iterating through a million results to generate a facet list | |
| 05:34 | Not that that was what you were precisely saying of course :p | |
| 05:35 | wajasu | i agree. that why i was investigating the facets. |
| 05:35 | now that i inderstand search so well. | |
| 05:37 | what i wish i could do is query zebra and just get back the biblionumber (Local-number) for a search. so i can grab items from the DB and count facets myself. | |
| 05:38 | but we get back a marcxml chunk and must decode with MARC::Record. | |
| 05:38 | irma joined #koha | |
| 05:39 | wajasu | what i noticed with my patch is as i went through more than N records that a syspref restricts, the resulting 5 facets for author changed. |
| 05:41 | so then you start realizing how much is left out. when i increased the number from 5 to 100, the catalog got more interesting because i saw more writers who wrote on a topic, or such. | |
| 05:41 | so i was learning about new authors, etc. | |
| 05:41 | so i think facets are worth it. | |
| 05:43 | dcook | I didn't really follow most of what you said there, but yeah...the current way of doing facets really sucks |
| 05:43 | There might be some hope though: http://lists.indexdata.dk/pipe[…]ruary/002563.html | |
| 05:43 | Probably unrelated though | |
| 05:44 | wajasu | i'm sure if we could read the source code we could find out, and see if there is any ICU support code. |
| 05:46 | dcook | Alas, I don't know, and I have other things to do :/ |
| 05:46 | Definitely learned something today though and we are farther than we were | |
| 05:47 | wajasu | yup |
| 05:47 | thx | |
| 05:49 | dcook | You're welcome. It was interesting! |
| 06:23 | cait joined #koha | |
| 06:40 | paxed | how does the cataloguer set the framework used for a record? |
| 06:43 | indradg joined #koha | |
| 06:49 | wajasu | when i search (z3950) and get the Add MaRC Record page, there is a default framework dropdown. if other framework mappings have been setup by the admiin, i would expect they would be avaiable. |
| 06:51 | indradg | good morning #koha |
| 06:55 | magnuse | happy equinox, #koha! |
| 07:04 | paxed | hm. any reason why 008 value builder doesn't know how show the type of material correctly? (eg. LDR/06=a and LDR/07=s, 008 builder shows BKS, when it should show CR) |
| 07:04 | marc21^ | |
| 07:08 | * cait | waves |
| 07:09 | paxed | ah, it's bug 9093 |
| 07:09 | huginn | 04Bug http://bugs.koha-community.org[…]w_bug.cgi?id=9093 normal, P5 - low, ---, gmcharlt, NEW , 008 forgetting what material type was chosen |
| 07:09 | paxed | lovely. |
| 07:11 | our cataloguer just had a fit over that. | |
| 07:29 | alex_a joined #koha | |
| 07:29 | alex_a | bonjour |
| 07:43 | paxed | so, is there any kind of ETA for Rancor? |
| 07:44 | magnuse | bug 11559 is currently "failed qa" |
| 07:44 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11559 enhancement, P5 - low, ---, jweaver, Failed QA , Professional cataloger's interface |
| 07:45 | paxed | because the current cataloguing system is bad. |
| 07:45 | magnuse | so pianohacker will have to fix the issues and then it will go through signoff, qa and getting pushed by the rm |
| 07:45 | there was quite some interest around it in marseille, so i think it will make it into 3.16 | |
| 07:45 | but you never know... | |
| 07:46 | sophie_m joined #koha | |
| 07:46 | Joubu | hello |
| 07:46 | wahanui | what's up, Joubu |
| 07:46 | paxed | i have no idea how the koha maintainers haven't been killed by cataloguers. |
| 07:48 | reiveune joined #koha | |
| 07:48 | reiveune | hellok |
| 07:53 | SIT joined #koha | |
| 07:54 | paxed | what is the reserveconstraints table for? |
| 07:55 | lds joined #koha | |
| 08:11 | cait joined #koha | |
| 08:11 | cait | good morning #koha |
| 08:16 | magnuse | kia ora cait |
| 08:23 | paul_p joined #koha | |
| 08:26 | cait | morning magnuse and paul_p |
| 08:28 | marcelr joined #koha | |
| 08:28 | marcelr | hi #koha |
| 08:43 | magnuse | hiya marcelr |
| 08:43 | marcelr | hi magnuse |
| 08:43 | wahanui | hi magnuse are you spending much time in oslo? |
| 08:45 | laurence joined #koha | |
| 08:48 | cait | wahanui forget hi magnuse |
| 08:48 | wahanui | cait: I forgot hi magnuse |
| 08:48 | cait | hi marcelr - thx for continuing :) |
| 08:49 | marcelr++ | |
| 08:49 | marcelr | hi cait |
| 08:49 | continue what btw | |
| 08:49 | cait | QA |
| 08:49 | marcelr | ok now i see :) |
| 09:02 | Hopla joined #koha | |
| 09:03 | gerundio joined #koha | |
| 09:05 | Hopla | Hey koha i just upgraded frop 3.8 to 3.12 i have some problems with my old css file so i want to make a new one only i find there are 2 now (i wanted to copy one to edit it) one in ccsr and one in prog ... in 3.8 there was only 1 so i should copy both or just 1 in a new local css file ? |
| 09:08 | laurence joined #koha | |
| 09:09 | cait | ccsr and prog are themes |
| 09:09 | each has it's own css file | |
| 09:09 | but don't change the default css | |
| 09:09 | Hopla | i wont |
| 09:09 | cait | ok :) |
| 09:09 | you can just add your things to opacusercss system preference | |
| 09:09 | then you don't need to care about the files and it will just overwrite the default | |
| 09:09 | if you want to use a file, there are other preferences | |
| 09:10 | the theme you use can be found out by looking at the opacthemes system preference | |
| 09:10 | Hopla | sec |
| 09:10 | ok im using prog | |
| 09:12 | cait | ccsr is going to go away with 3.16 probably, it's responsive, but there is a new theme bootstrap since 3.14 that's going to be the new and only default theme at some point |
| 09:12 | Hopla | so if i make a copy off the opac.css edited it and point to it in opaclayoutstylesheet it would be ok |
| 09:12 | cait | if you worked like that in the past, it should work again |
| 09:12 | Hopla | ow |
| 09:13 | thx fr the info :) | |
| 09:13 | cait | if you only have minor changes, i'd recomment just changing what you want to change and use opacusercss - it's bit easier to maintain |
| 09:13 | sure | |
| 09:13 | Hopla | aha |
| 09:13 | searching that one | |
| 09:14 | cait | yu can put in the css there directly and change anytime by logging in |
| 09:14 | also on updating, you won't miss the changes from the new css, so probably less work to make it fit again | |
| 09:15 | Hopla | ill try i only change logo + colours so |
| 09:20 | cait | ah yeah, probably easier then :) |
| 09:54 | vfernandes joined #koha | |
| 09:54 | vfernandes | hi :) |
| 09:54 | cait | hi vfernandes :) |
| 09:54 | vfernandes | where I can see the koha roadmap and new features/enhacements for 3.16? |
| 09:56 | cait | you could do a search in bugzilla |
| 09:56 | for pushed since last release | |
| 09:56 | you can create searches by status change - that should work | |
| 09:56 | and master as version | |
| 09:57 | vfernandes | i thought there was a simple way :D |
| 09:57 | cait | there is also a release notes script that can be fun - it's on the git repo |
| 09:58 | it's still a bit time to go, so hard to tell what will make it | |
| 09:58 | petter joined #koha | |
| 09:58 | cait | i woud look at what's already pushed |
| 09:58 | i think it was get-bugs.pö | |
| 09:59 | .pl - http://git.koha-community.org/[…]30ed94fabd6db4ea0 | |
| 09:59 | hi petter | |
| 09:59 | petter | hi cait! |
| 10:03 | magnuse | kia ora petter! |
| 10:03 | petter | hei magnus! |
| 10:03 | magnuse | cait: .pö - is that for german perl scripts? ;-) |
| 10:03 | vfernandes | but there is any new features that you have any information cait? |
| 10:06 | cait | i'd have to think - there are some :) the dashboard always shows the last five enh/features pushed |
| 10:07 | http://dashboard.koha-community.org/ | |
| 10:07 | wahanui | http://dashboard.koha-community.org/ are the better stats to optimise |
| 10:14 | magnuse | enhancements and new features pushed to master since 2013-11-22: http://bugs.koha-community.org[…]%20apply&bug_stat |
| 10:14 | us=Pushed%20to%20Master&bug_status=Pushed%20to%20Stable&bug_status=RESOLVED&bug_status=VERIFIED&bug_status=CLOSED&bug_status=BLOCKED&chfield=bug_status&chfieldfrom=2013-11-22&chfieldto=Now&chfieldvalue=Pushed%20to%20master&list_id=93066&query_format=advanced | |
| 10:14 | whoa, that's a long url | |
| 10:15 | indradg joined #koha | |
| 10:15 | magnuse | short version: http://tinyurl.com/o2l47s6 |
| 10:16 | hm, it just shows one new feature - everything might not be correctly tagged... | |
| 10:17 | cait | i think we are quite strict with 'new feature' but not so many more modules we can write... so maybe we should be a bit less strict :) |
| 10:18 | magnuse | :-) |
| 10:18 | well, there are some features waiting to get in, i guess | |
| 10:18 | like rancor | |
| 10:22 | cait | hm i forgot to make tea today. |
| 10:22 | * cait | wanders off |
| 10:23 | vfernandes | thanks magnuse and cait |
| 10:51 | magnuse | woohoo - bug 10003 |
| 10:51 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=10003 enhancement, P5 - low, ---, tomascohen, Passed QA , koha-* scripts (packages) should provide tab-completion in bash |
| 10:54 | Dyrcona joined #koha | |
| 11:20 | BTT joined #koha | |
| 11:20 | * BTT | slaps francharb_afk around a bit with a large fishbot |
| 11:21 | BTT left #koha | |
| 11:22 | francharb | good morning |
| 11:22 | BT joined #koha | |
| 11:25 | p_vdk joined #koha | |
| 11:38 | * magnuse | will have french cheese for lunch, but only one kind... |
| 11:58 | meliss joined #koha | |
| 12:05 | drojf joined #koha | |
| 12:05 | drojf | hi #koha |
| 12:13 | petter joined #koha | |
| 12:13 | oleonard joined #koha | |
| 12:16 | oleonard | Hi #koha |
| 12:17 | cait | hi oleonard :) |
| 12:20 | nlegrand | hey #koha ! |
| 12:42 | nengard joined #koha | |
| 12:55 | oleonard | Seems like people are entering search terms into Koha list emails like it's Google. |
| 12:57 | drojf | it isn't? ;) |
| 13:07 | tcohen joined #koha | |
| 13:09 | Dyrcona joined #koha | |
| 13:09 | rhcl joined #koha | |
| 13:14 | rhcl left #koha | |
| 13:34 | kivilahtio | m de rooy |
| 13:34 | does anyone now his nick? | |
| 13:34 | oleonard | irc regulars? |
| 13:34 | wahanui | irc regulars is at http://wiki.koha-community.org/wiki/IRC_Regulars |
| 13:35 | kivilahtio | oleonard: thanks :D |
| 13:35 | marcelr: about bug 11974 | |
| 13:35 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11974 enhancement, P5 - low, ---, gmcharlt, NEW , Enable unix socket connections for database connections. |
| 13:35 | marcelr | hi kivilahtio |
| 13:35 | kivilahtio | hi there :D |
| 13:36 | good suggestion btw | |
| 13:36 | I am not very good with Makefile.pl | |
| 13:36 | marcelr | :) |
| 13:36 | kivilahtio | or make whatsovere |
| 13:36 | it is bad for my health | |
| 13:36 | marcelr | you can do it ! |
| 13:36 | kivilahtio | I was hoping you could do it |
| 13:36 | marcelr | ;) |
| 13:36 | kivilahtio | that would force you to sign off my patch |
| 13:36 | and I could sig of your pathc | |
| 13:37 | this way we both get a sign off and commit | |
| 13:37 | win win | |
| 13:37 | you scratch my back i scratch yours? | |
| 13:37 | marcelr | better offer? |
| 13:37 | kivilahtio | :) |
| 13:37 | I'll take a look ral quick | |
| 13:37 | but if ANY issues arise... | |
| 13:37 | marcelr | so good |
| 13:37 | markvandenborre | I am looking into the best way to simplify entry of sheet music books into koha |
| 13:38 | we only keep a relatively minor subset of metadata about them | |
| 13:38 | any pointers towards relevant bits of documentation? | |
| 13:38 | cait | markvandenborre: sounds like you might want to use a custom framework |
| 13:38 | markvandenborre | cait: as in not koha? |
| 13:39 | there seem to be provisions in koha for sheet music... | |
| 13:39 | even ISMN (sheet music) numbering lookup | |
| 13:39 | cait | hm? |
| 13:40 | a custom framework would be just a catalouging form defined witht he fields you want to use | |
| 13:40 | to make data entry easier | |
| 13:40 | a bibliographic framework | |
| 13:40 | are you using marc21 or unimarc? | |
| 13:40 | markvandenborre | right, I was already guessing that this was some bilbiographic or koha specific terminology, not a custom web framewrork :-) |
| 13:41 | tgoatley joined #koha | |
| 13:41 | markvandenborre | unimarc iirc, I chose the EU specific version |
| 13:41 | cait | unimarc is only used in a few countries |
| 13:41 | france, italy... | |
| 13:41 | markvandenborre | Belgium? |
| 13:41 | cait | depending on where you are located marc21 might make sense |
| 13:41 | markvandenborre | what am I going to get the best support on? |
| 13:41 | cait | i can't really tell you, maybe look what your national library uses, or what z39.50 servers in your regions use |
| 13:41 | markvandenborre | ok |
| 13:42 | cait | germany is all marc21 now |
| 13:42 | but that's something that is hard to change later | |
| 13:42 | because the formats are different | |
| 13:42 | so worth investigating a bit | |
| 13:42 | markvandenborre | it's not incredibly important |
| 13:42 | very small library | |
| 13:42 | 2702 books | |
| 13:42 | cait | also documentation about the format |
| 13:42 | markvandenborre | some audio cd's and only a tiny number of dvd's |
| 13:43 | cait | both formats are similar and then very different |
| 13:43 | markvandenborre | with no big increase on either a close or faraway horizon |
| 13:43 | cait | in which fields are used |
| 13:43 | markvandenborre | can you point me at more specific docs about this "custom framework" reference you just made |
| 13:43 | "custom framework" is horrible as a web search term | |
| 13:43 | cait | look in the manual for bibliographic frameworks |
| 13:43 | petter | olli: about your socket patch - did you measure the speedup yourself? |
| 13:43 | 30-40% | |
| 13:44 | is it for real? | |
| 13:46 | NateC joined #koha | |
| 13:46 | markvandenborre | heh, the links to the manual dropped off my screen |
| 13:47 | when I clicked the documentation page | |
| 13:47 | so I just now looked at it more closely | |
| 13:47 | they were entirely invisible when arriving to http://koha-community.org/documentation/ | |
| 13:48 | don't know if there's anyone with the ambition to make the docs more prominent, but it would have helped stupid me | |
| 13:49 | the manual seems to be quite comprehensive | |
| 13:55 | cait: so the idea is I remove all the tags I don't need from a copy of the default framework, correct? | |
| 13:55 | marcelr | cait: bug 9032 |
| 13:55 | huginn | 04Bug http://bugs.koha-community.org[…]w_bug.cgi?id=9032 enhancement, P5 - low, ---, m.de.rooy, Signed Off , Share a list |
| 13:56 | marcelr | i would have preferred to not include the notices for all the other languages at all :) |
| 13:56 | but that project got killed rather quickly | |
| 13:57 | markvandenborre | just answered my own question here after reading the f*cking manual |
| 13:58 | marcelr | ga zo door markvandenborre |
| 13:59 | markvandenborre | marcelr: I was hoping to find an easy way to delete tens of tags we don't need at once |
| 13:59 | am I overlooking something | |
| 13:59 | or is that just not possible? | |
| 13:59 | marcelr | command line? |
| 14:00 | markvandenborre | marcelr: where is this even stored, I presume in the sql db |
| 14:00 | marcelr | you mean user tags or marc tags? btw.. |
| 14:01 | markvandenborre | I have made a copy of the default framework, which I suppose contains only marc tags |
| 14:01 | and I'm deleting all kinds of information from that | |
| 14:02 | marcelr | ok you mean tags in framework |
| 14:02 | markvandenborre | with the goal of creating a very _simple_ |
| 14:02 | marcelr | you can hide them |
| 14:02 | markvandenborre | ah? please tell me more about that? |
| 14:02 | marcelr | look for tables marc_tag.. marc_subfield.. etc (please see the docs :) |
| 14:02 | maximep joined #koha | |
| 14:03 | cait | markvandenborre: better not delete but hide |
| 14:03 | in my experience | |
| 14:03 | and you can use export/import and edit as spreadsheet | |
| 14:04 | markvandenborre | ok... so now the question becomes how to "hide" |
| 14:04 | cait | export from koha |
| 14:04 | * markvandenborre | starts looking through the manual |
| 14:04 | cait | i think -5 shoudl do it |
| 14:04 | in the hideen | |
| 14:04 | markvandenborre | this feels like very much a welcoming irc channel, more so than many others |
| 14:05 | marcelr | this feels like a quote :) |
| 14:05 | markvandenborre | thank you |
| 14:05 | marcelr | you can add quotes via a bot |
| 14:06 | markvandenborre | thank you for the info, but I'm more interested in getting this up and running |
| 14:07 | cait: -5? hideen? | |
| 14:07 | I have this export open in libreoffice now | |
| 14:07 | cait | first |
| 14:07 | ah | |
| 14:07 | there is a hidden column | |
| 14:07 | in the spreadsheet | |
| 14:08 | markvandenborre | there is no column with the header "hidden" as far as I can see |
| 14:08 | but maybe you mean something more magic? | |
| 14:10 | I see column headers tagfield, liblibrarian, libopac, repeatable, mandatory, authorised_value, frameworkcode | |
| 14:10 | ah | |
| 14:10 | there is more info lower in the csv | |
| 14:15 | so for a lot of info, this is clear enough | |
| 14:17 | is there some overview page where I can find the meaning of all these unimarc fields? | |
| 14:17 | some are quite cryptic | |
| 14:17 | jcamins | unimarc? |
| 14:17 | wahanui | i guess unimarc is http://www.ifla.org/en/publica[…]ted-documentation |
| 14:18 | pastebot | "petter" at 127.0.0.1 pasted "bug 11974 before / after benchmarking" (34 lines) at http://paste.koha-community.org/179 |
| 14:18 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11974 enhancement, P5 - low, ---, gmcharlt, NEW , Enable unix socket connections for database connections. |
| 14:30 | petter | kivilahtio: I get a minor speeup from using domain sockets, from few percent to at best 19% faster |
| 14:30 | kivilahtio | petter: how are you testing? |
| 14:30 | petter | I'd be happy to sign of |
| 14:30 | by the benchmarking scripts in misc/load_testing | |
| 14:30 | check the linked paste | |
| 14:30 | http://paste.koha-community.org/179 | |
| 14:31 | rocio joined #koha | |
| 14:31 | nengard | oleonard around? |
| 14:31 | oleonard | Yes |
| 14:31 | kivilahtio | petter: a sec I almost have the Makefile.PL modified |
| 14:32 | nengard | I have a library that wants to hide the guarantor info on the child record add/edit form |
| 14:32 | the borrowerunwanted field isn't working | |
| 14:32 | and the section has no id | |
| 14:32 | tips on hiding it with jquery? | |
| 14:32 | marcelr | kivilahtio++ |
| 14:32 | kivilahtio | it is a pain :) |
| 14:32 | and I can feel the maintenance issues | |
| 14:33 | petter | hm, now that i look at it actually some benchmarks are slower |
| 14:33 | needs more testings | |
| 14:33 | more benchmarks i mean | |
| 14:34 | oleonard | nengard: I shouldn't even ask, but... Why do they want to hide it? |
| 14:34 | nengard | i don't know the answer to that ... I just do what i'm asked to do :) |
| 14:34 | Probably they don't want to link the child to an adult but do want other child functionality like the j to a cron | |
| 14:35 | oleonard | Do they think you can't create a child record without going through the "add child" link? |
| 14:35 | What does it have to do with the J to A cron? | |
| 14:36 | nengard | wait |
| 14:36 | i had a typo let me see if it works now | |
| 14:37 | Yes you can ... | |
| 14:37 | create a child without going through that link | |
| 14:37 | kivilahtio | petter: marcelr: go for it! bug 11974 |
| 14:37 | nengard | you just go to 'New Patron' |
| 14:37 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11974 enhancement, P5 - low, ---, gmcharlt, NEW , Enable unix socket connections for database connections. |
| 14:37 | nengard | and choose the patron type you want :) |
| 14:37 | and the J to A Cron turns a child to an adult when they reach the right age per the patron category | |
| 14:38 | marcelr | kivilahtio: i will be looking at it (in some stage of the process) |
| 14:38 | but not now | |
| 14:38 | gaetan_B joined #koha | |
| 14:39 | oleonard | nengard: I'll give up trying to understand, but I think their request might indicate a misunderstanding of how Koha works on their part that you might want to investigate. |
| 14:39 | nengard: $("legend:contains('Guarantor information')").parent().hide(); | |
| 14:41 | petter | olli - how to test the Makefile stuff. I always installed using packages then gitify.. |
| 14:41 | ? | |
| 14:45 | kivilahtio | petter: erm |
| 14:45 | petter: go to the folder where your INSTALL.ubuntu is | |
| 14:46 | petter: maybe the easiest way is to use those fancy virtual sandboxes you have | |
| 14:47 | petter: but from the INSTALL.ubuntu folder, run perl Makefile.PL | |
| 14:47 | answer questions | |
| 14:47 | make | |
| 14:47 | make test | |
| 14:47 | nano blib/KOHA_CONF_DIR/koha-conf.xml | |
| 14:47 | petter | So does the packages use the Makefile in anyway? |
| 14:48 | kivilahtio | see if mysql_socket is there with hte value you gave |
| 14:48 | petter: no idea | |
| 14:48 | petter | If packages are the recommended & preferred way, who uses the perl Makefile way? |
| 14:48 | kivilahtio | but if you have the source from git, you should have Makefile.PL |
| 14:48 | petter | off course |
| 14:48 | I have it | |
| 14:48 | kivilahtio | petter: don't ask me |
| 14:48 | petter: if you do crazy stuff like me you will enjoy the Makefile.PL | |
| 14:48 | petter | haha |
| 14:48 | never | |
| 14:49 | Anyway, I did some more benchmarkings | |
| 14:49 | kivilahtio | I dont think you need to do make install (to overwrite your configs) |
| 14:49 | petter | It seem indeed to be about 10% speed up on average |
| 14:49 | ok | |
| 14:49 | kivilahtio | just verify that the file there is modified |
| 14:49 | petter | sure, got it |
| 14:50 | Did you run the benchmarking scripts? | |
| 14:50 | kivilahtio | i tried, i failed miserably |
| 14:50 | I will write my own | |
| 14:50 | petter | but its easy |
| 14:50 | kivilahtio | I got like 100% error |
| 14:50 | dunno | |
| 14:50 | need to take a look there | |
| 14:50 | but now i need to configure zebra + mariadb + koha with unix sockets | |
| 14:50 | petter | cpan install HTTPD::Bench::ApacheBench |
| 14:50 | then run the scripts | |
| 14:50 | kivilahtio | yeah |
| 14:51 | like i said i got some errors | |
| 14:51 | petter | you should measure! |
| 14:51 | kivilahtio | jaja |
| 14:51 | petter | if not you dont know if it works |
| 14:51 | kivilahtio | could be |
| 14:55 | nengard left #koha | |
| 15:11 | nengard joined #koha | |
| 15:20 | cait left #koha | |
| 15:55 | nengard | anyone around to help me find a bug report? |
| 15:55 | I'm looking for a bug report I think we have that talks about the date shown for serials in the opac | |
| 15:55 | how it's not the date every library wants to see | |
| 16:02 | reiveune | bye |
| 16:02 | reiveune left #koha | |
| 16:15 | wajasu | nengard: maybe bug 8296 |
| 16:15 | huginn | 04Bug http://bugs.koha-community.org[…]w_bug.cgi?id=8296 enhancement, P5 - low, ---, julian.maurice, Needs Signoff , Add descriptive (text) published date field for serials |
| 16:26 | tgoatley joined #koha | |
| 16:31 | nengard left #koha | |
| 16:51 | nengard joined #koha | |
| 16:56 | bag | magnuse: HELLO there |
| 17:08 | drojf joined #koha | |
| 17:08 | pianohacker joined #koha | |
| 17:09 | pianohacker | hello |
| 17:15 | rhcl joined #koha | |
| 17:17 | Dyrcona joined #koha | |
| 17:17 | fredericd joined #koha | |
| 17:24 | laurence left #koha | |
| 17:26 | mtompset_away joined #koha | |
| 17:27 | mtompset | Greetings, #koha/. |
| 17:30 | rhcl | hihi |
| 17:31 | indradg joined #koha | |
| 17:34 | magnuse | bag: HI |
| 17:47 | alphaman joined #koha | |
| 17:52 | Brooke joined #koha | |
| 17:52 | Brooke | o/ |
| 17:57 | francharb | \o |
| 17:57 | Brooke | we gots us any big city fancy pants label makin' folkses? |
| 18:04 | gadzukes there are actual pictures in the label creator :) | |
| 18:09 | Dyrcona joined #koha | |
| 18:11 | wajasu | should i start testing signoffs with wheezy for master from now on? |
| 18:17 | cait joined #koha | |
| 18:18 | mtompset | wajasu: What do you mean? |
| 18:22 | wajasu | i wasn't sure if squeeze is the target debian, or wheezy. i wondered if i needed to build my test VM with wheezy |
| 18:30 | mtompset | I would guess that squeeze is okay until some time in 2015. :) |
| 18:30 | https://wiki.debian.org/DebianReleases | |
| 18:31 | rhcl joined #koha | |
| 18:32 | alphaman | Anybody have any experience using John Wohler's sip2.class.php? I'm having trouble getting a stable connection to 3.14 with it. |
| 18:35 | pianohac1er joined #koha | |
| 18:51 | mtompset | Did you just ask about PHP? |
| 18:51 | @karma php | |
| 18:51 | huginn | mtompset: php has neutral karma. |
| 18:52 | mtompset | @karma perl |
| 18:52 | huginn | mtompset: perl has neutral karma. |
| 18:52 | alphaman | yup. sorry... |
| 18:52 | I was worried I'd get neutralized... | |
| 18:52 | mtompset | No need to apologize. Everyone does it. They are just ashamed to admit it. ;) |
| 18:53 | alphaman | In all fairness, I've been doing Perl far longer than PHP… |
| 18:53 | sometimes you just gotta dig down in the muck to find a solution | |
| 18:55 | yay, tho I walk through the valley of PHP, I shall fear no evil, for Perl is my shepherd, I shall not stack dump… | |
| 18:58 | rhcl | alphaman: just out of curiosity are you trying to connect some specific app to Koha? |
| 19:03 | alphaman | homegrown. trying to auth patron's cards for our Drupal server so they can access our online resources |
| 19:17 | Dyrcona joined #koha | |
| 19:21 | paul_p joined #koha | |
| 19:25 | jenkins_koha | Project Koha_Docs_3.12.x build #52: FAILURE in 47 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/52/ |
| 19:26 | Project Koha_Docs build #458: FAILURE in 18 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/458/ | |
| 19:30 | Project Koha_Docs_3.14.x build #32: FAILURE in 14 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/32/ | |
| 19:39 | Project Koha_Docs_3.12.x build #53: STILL FAILING in 4.5 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/53/ | |
| 19:44 | Project Koha_Docs_3.14.x build #33: STILL FAILING in 4 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/33/ | |
| 19:46 | drojf joined #koha | |
| 19:52 | meliss joined #koha | |
| 19:54 | jenkins_koha | Project Koha_Docs_3.12.x build #54: STILL FAILING in 5 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/54/ |
| 19:56 | Project Koha_Docs build #459: STILL FAILING in 4.8 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/459/ | |
| 19:59 | Project Koha_Docs_3.14.x build #34: STILL FAILING in 3.9 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/34/ | |
| 20:09 | Project Koha_Docs_3.12.x build #55: STILL FAILING in 4 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/55/ | |
| 20:13 | Dyrcona joined #koha | |
| 20:14 | jenkins_koha | Project Koha_Docs_3.14.x build #35: STILL FAILING in 3.6 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/35/ |
| 20:24 | nengard | hmm jenkins isn't helping me - he's not telling me why the manual is failing :( -- "No problems were identified. If you know why this problem occurred, please add a suitable Cause for it" |
| 20:24 | jenkins_koha | Project Koha_Docs_3.12.x build #56: STILL FAILING in 4.8 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/56/ |
| 20:25 | nengard | I can't make it stop failing if you don't tell me what I did wrong jenkins!! |
| 20:25 | :( | |
| 20:26 | jenkins_koha | Project Koha_Docs build #460: STILL FAILING in 5.3 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/460/ |
| 20:27 | bag | nengard: can you get to where the docs are? |
| 20:27 | mtompset | @later tell papa So tired trying to figure out bug 11213 test plan. I think I finally got it. |
| 20:27 | huginn | mtompset: The operation succeeded. |
| 20:27 | nengard | bag you mean git? yes |
| 20:27 | usually jenkins emails me though with the exact error | |
| 20:27 | he's being a .... not nice bot .... today | |
| 20:27 | magnuse | it fails awfully fast... |
| 20:28 | mtompset | "Cannot allocate memory" That's the problem, I figure. |
| 20:28 | nengard: Can you reboot the machine it is building on? | |
| 20:28 | nengard | not sure what machine that is .... |
| 20:28 | magnuse | i can see "ERROR: Workspace has a .git repository, but it appears to be corrupt." here: http://jenkins.koha-community.[…]_Docs/460/console |
| 20:28 | nengard | I type my fancy git commands and it magically updates |
| 20:28 | bag is it on our server? | |
| 20:29 | mtompset | nengard: free -m |
| 20:29 | How much is left? | |
| 20:29 | When the script runs, it probably is a memory hog. | |
| 20:29 | bag | mtompset: are you talking about the git server? or the jenkins server? |
| 20:29 | jenkins_koha | Project Koha_Docs_3.14.x build #36: STILL FAILING in 4.4 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/36/ |
| 20:29 | magnuse | i also see "Building remotely on Galen in workspace /var/lib/jenkins/workspace/Koha_Docs" - maybe gmcharlt knows something about it? |
| 20:29 | mtompset | http://jenkins.koha-community.[…]_Docs/460/console |
| 20:30 | ERROR: Workspace has a .git repository, but it appears to be corrupt. | |
| 20:30 | ewww.... | |
| 20:31 | nengard | is it corrupt on my local machine? or is that something that I have no control over? |
| 20:31 | it all seems okay from my end | |
| 20:32 | magnuse | yeah, looks like a problem on the jenkins server |
| 20:33 | nengard | whew |
| 20:33 | so i'm all good :) | |
| 20:34 | magnuse | from that console output it looks like the git repo on the jenkins server is corrupt, maybe because of some lack of memory (is my guess, based on looking at the error messages) |
| 20:35 | bag | nengard what are you doing to gmcharlt's memory :P |
| 20:35 | nengard | breaking it |
| 20:39 | jenkins_koha | Project Koha_Docs_3.12.x build #57: STILL FAILING in 7.9 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/57/ |
| 20:39 | nengard | :P |
| 20:40 | nengard left #koha | |
| 20:41 | jenkins_koha | Project Koha_Docs build #461: STILL FAILING in 6 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/461/ |
| 20:44 | Project Koha_Docs_3.14.x build #37: STILL FAILING in 4.8 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/37/ | |
| 20:47 | Project Koha_Docs build #462: STILL FAILING in 4.8 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/462/ | |
| 20:53 | Project Koha_Docs build #463: STILL FAILING in 5.6 sec: http://jenkins.koha-community.[…]ob/Koha_Docs/463/ | |
| 20:55 | rangi | its not actually corrupt |
| 20:55 | its just the box is out of ram | |
| 20:55 | not on the jenkins server either, but on galens node | |
| 20:56 | mtompset | rangi++ # yay for a diagnosis. :) |
| 20:56 | Dyrcona | gmcharlt is giving a presentation. I'll let him know after he's finished. |
| 20:57 | * cait | waves |
| 20:58 | jenkins_koha | Project Koha_Docs_3.12.x build #58: STILL FAILING in 3.8 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/58/ |
| 20:59 | rangi | Dyrcona: thanks, ill disable the node for the time being |
| 21:00 | http://jenkins.koha-community.org/computer/Galen/ | |
| 21:09 | Irma joined #koha | |
| 21:09 | magnuse | so Viktor's public library has been testing koha under plack on a server i set up for them, and it has not blown up while they have been testing the functions they usually use |
| 21:10 | they are now begging me to let them test koha+plack for one day in real use :-) | |
| 21:10 | rangi | sweet, we run the opac under plack in production |
| 21:10 | however without fixing the zconn subroutine | |
| 21:10 | in C4::Context | |
| 21:11 | it isnt safe, it randomly gives no search results (connection has gone away) | |
| 21:11 | but if you fix that, its mostly safe enough, (there are more issues but we plan on fixing them soon) | |
| 21:15 | magnuse | rangi: cool. and "the zconn subroutine" would be fixed by your patch which is floating around somewhere? |
| 21:15 | if there are known issues it would be really good to have them documented somewhere | |
| 21:15 | rangi | yep thats the plan |
| 21:16 | but sysprefs is one | |
| 21:16 | fixing the memcached config is another | |
| 21:16 | we cant be having env variables in apache .. | |
| 21:18 | eythian | hi |
| 21:18 | magnuse | "sysprefs is one" - does that mean there is a problem with sysprefs in general, or that changing them can be troublesome? |
| 21:18 | hiya eythian | |
| 21:22 | * cait | waves |
| 21:23 | * magnuse | waves to cait |
| 21:23 | pianohacker | hi all |
| 21:23 | magnuse | hiya pianohacker |
| 21:24 | pianohacker | hey magnuse |
| 21:24 | how was your flight back? | |
| 21:24 | magnuse | looong :-) |
| 21:24 | well, not as long as yours, but... | |
| 21:24 | cait | hi pianohacker |
| 21:24 | :) | |
| 21:25 | magnuse | we left our apartment at 05:00 am and arrived at home at 20:00 pm |
| 21:25 | 3 flights and lots of waiting | |
| 21:25 | pianohacker: how was yours? | |
| 21:25 | pianohacker | magnuse: past a certain point, it's just long and exhausting |
| 21:25 | magnuse | yeah... |
| 21:26 | * cait | nods |
| 21:26 | magnuse | leonard is now ~22 months and has flown 22 times |
| 21:26 | pianohacker | magnuse: mine was decent. Got patted down twice, really should have shaved, and customs in the US was way understaffed, but everything happened in time :) |
| 21:27 | magnuse | that is something :-) |
| 21:27 | pianohacker | magnuse: wow. I hope you don't keep that up, that's a lot of time in the air! |
| 21:32 | magnuse | pianohacker: it is. huge carbon footprint for a small boy. it will be a lot less once he starts kindergarten and his mothers starts working again... |
| 21:33 | pianohacker | ahh, yeah... he's coming with you to oslo, I'm guessing? |
| 21:34 | tcohen joined #koha | |
| 21:37 | magnuse | not so far, but we have talked about spending a couple weeks there before summer |
| 21:37 | we'll see | |
| 21:37 | * magnuse | wanders off |
| 21:37 | pianohacker | bye |
| 21:41 | tcohen | rangi: has anyone profilled if caching syspref fetching is worth the trouble? |
| 21:41 | i guess mysql has to be good for caching such trivial queries | |
| 21:42 | jcamins | tcohen: it is *not*. |
| 21:42 | eythian | jcamins: which comment are you refuting? |
| 21:43 | jcamins | Oh. |
| 21:43 | Right. | |
| 21:43 | tcohen | the second? |
| 21:43 | wahanui | the second is, like, by design |
| 21:43 | jcamins | tcohen: yes, I've profiled, and it's not worth it. |
| 21:43 | Maybe with Plack it could be worthwhile, but that's kind of iffy, too. | |
| 21:43 | And the absolute worst thing you can do performance-wise is to preload all the sysprefs. | |
| 21:44 | bag | right |
| 21:46 | jcamins | Well, no, the worst thing you can do performance-wise is preload all the sysprefs and discard them all. |
| 21:46 | Or, I suppose, try to cache the entire database. | |
| 21:47 | eythian | yeah, it's worth noting that in many cases deserialisation can be pretty expensive. |
| 21:48 | jcamins | It's better if you're using JSON::XS, but it's still bad. |
| 21:48 | eythian | I found a case a little while ago where the memoisation of a function made it slower. |
| 21:49 | papa joined #koha | |
| 21:49 | eythian | (because it was doing its own internal caching anyway, which has its own problems. But someone had added the memoisation without benchmarking before/after.) |
| 21:49 | wajasu | could it be worth using the memcached option for zebra queries? and possibly namedResults for paging |
| 21:49 | eythian | zebra is not something that is slow. |
| 21:49 | you do a search and profile it, zebra doesn't really show up at all. | |
| 21:49 | tcohen | fetching the frameworks is |
| 21:49 | bag | it's just a pain :P |
| 21:49 | eythian | frameworks are slow. |
| 21:49 | wajasu | premature optimization then |
| 21:50 | eythian | they should be handled in a better way. |
| 21:50 | tcohen | question: would it be a bad thing to force a collation for SQL queries on the Perl side? |
| 21:51 | wajasu | what if the mapping framework could be generated as an XSLT transform that zebra uses and maybe we return MARC as JSON |
| 21:51 | just a wild idea | |
| 21:51 | eythian | nothing about that sounds like it's likely to be an improvement :) |
| 21:51 | tcohen | MiJ++ |
| 21:52 | jcamins | Well, MiJ is faster than Marcxml. |
| 21:52 | cait left #koha | |
| 21:52 | tcohen | i was looking for a way to have dbic tell me the attributes of a db connection (character set, collation) |
| 21:52 | eythian | http://debian.koha-community.o[…]h-cached/nytprof/ <-- wajasu, start here. |
| 21:53 | tcohen | and found that we could actually force those from the code |
| 21:53 | jcamins | (n.b. this was not a vote for making a change like that, just observing that MiJ is faster) |
| 21:53 | eythian | http://debian.koha-community.o[…]obin/opac-search/ <-- or here for a version with memcache off. |
| 21:54 | rangi | the thing for remembering with caching |
| 21:54 | its not for speed | |
| 21:54 | its for scalability | |
| 21:54 | memcached scales more easily than mysql | |
| 21:55 | wajasu | i wish i could get the biblionumbers for a resultset from zebra without having to process the marcxml chunk. then i could write batched queries against the db to determine if an item is hidden. |
| 21:56 | tcohen | you can do it |
| 21:56 | jcamins knows how :-P | |
| 21:57 | eythian: the cached version takes longer? | |
| 21:57 | eythian | yes |
| 21:58 | tcohen | found the bottelneck? |
| 21:58 | pianohacker | rangi: that reminds me, did you get a chance to look at the Koha::Database slowness? |
| 21:58 | rangi | nope, ive been buried in ncip |
| 21:59 | pianohacker | that's a heavy, smelly pile to be buried in good sir |
| 21:59 | eythian | tcohen: it's mentioned on bug 11051 |
| 21:59 | huginn | 04Bug http://bugs.koha-community.org[…]_bug.cgi?id=11051 enhancement, P5 - low, ---, robin, Signed Off , Performance of opac-search |
| 22:01 | wajasu | by the way, is it better to run C4::Context::preference once at the top of the script instead of in the loops latter when deciding to process? |
| 22:02 | eythian | you should avoid calling functions in loops anyway, just to be safe. |
| 22:02 | (in general, anyway.) | |
| 22:03 | (often it's not optional, but if you can readily store something in your own namespace, it's probably good to.) | |
| 22:03 | but again, premature optimisation etc etc. | |
| 22:03 | so do what looks best and then see. | |
| 22:03 | if your loop happens 5 times, no one cares. If it happens 5,000 times, then maybe worth refactoring. | |
| 22:03 | your profiler will let you know. | |
| 22:05 | wajasu | is opac-search really geared toward returning 1000 results. since the zebra sort seems like its configured for that many results. |
| 22:05 | eythian | I don't really know, I've not look at that sort of thing very deeply. |
| 22:07 | wajasu | i know we git the hits, then page around 20 per page, so we grab 20 marcxmls to present. |
| 22:07 | tcohen | actually 1000 might be a low value |
| 22:08 | wajasu | when the build facets runs it also has a max records to preocess syspref and that constrains the amount of marcxml and then the mysql db items retrived to calucalte facet counters. |
| 22:09 | tcohen | facets aren't even facets |
| 22:09 | * tcohen | hides |
| 22:09 | wajasu | true |
| 22:10 | but we must process all the items for the biblionumbers for a given result set. | |
| 22:11 | tcohen | context? |
| 22:11 | wahanui | i think context is everything? |
| 22:12 | wajasu | i'm wondering if i can just generate chunked queries supplying a subset of the biblionumbers, that use the where clause to match the OpacHideItem rules. |
| 22:12 | tcohen | ask zebra for several biblionumbers for using them? |
| 22:13 | bag | @quote random |
| 22:13 | huginn | bag: Quote #175: "*oleonard is waiting for a good scientist -> hulk patron category transition script" (added by wizzyrea at 03:44 PM, January 04, 2012) |
| 22:13 | wahanui | i already had it that way, huginn. |
| 22:14 | wajasu | got to run. |
| 22:47 | grex_ joined #koha | |
| 23:33 | pianohacker | amen |
| 23:47 | gmcharlt | rangi: jenkins slave process got bloated; I've killed it and put it back in the cluster |
| 23:51 | tcohen joined #koha | |
| 23:52 | jenkins_koha | Starting build #38 for job Koha_Docs_3.14.x (previous build: STILL FAILING -- last SUCCESS #31 1 day 10 hr ago) |
| 23:52 | Yippee, build fixed! | |
| 23:52 | wahanui | o/ '`'`'`'`'`'`'`'`'` |
| 23:52 | jenkins_koha | Project Koha_Docs_3.14.x build #38: FIXED in 5 min 30 sec: http://jenkins.koha-community.[…]a_Docs_3.14.x/38/ |
| 23:52 | Nicole C. Engard: update local cover images info | |
| 23:52 | tcohen | @wunder cordoba, argentina |
| 23:52 | huginn | tcohen: The current temperature in Bo Alto de San Martin, Cordoba City, Argentina is 20.6°C (8:50 PM ART on March 20, 2014). Conditions: Clear. Humidity: 81%. Dew Point: 17.0°C. Pressure: 29.71 in 1006 hPa (Steady). |
| 23:56 | dcook | tcohen: How long until you head to London? :) |
| 23:56 | @wunder london, england | |
| 23:56 | huginn | dcook: The current temperature in London, United Kingdom is 8.0°C (11:50 PM GMT on March 20, 2014). Conditions: Clear. Humidity: 93%. Dew Point: 7.0°C. Windchill: 6.0°C. Pressure: 29.68 in 1005 hPa (Steady). |
| 23:56 | pianohacker | bye |
| 23:56 | dcook | ta, pianohacker |
| 23:56 | pianohacker | cya dcook |
| 23:59 | jenkins_koha | Starting build #59 for job Koha_Docs_3.12.x (previous build: STILL FAILING -- last SUCCESS #51 1 day 10 hr ago) |
| 23:59 | Yippee, build fixed! | |
| 23:59 | wahanui | o/ '`'`'`'`'`'`'`'`'` |
| 23:59 | jenkins_koha | Project Koha_Docs_3.12.x build #59: FIXED in 1 min 55 sec: http://jenkins.koha-community.[…]a_Docs_3.12.x/59/ |
| 23:59 | Nicole C. Engard: update local cover images info |
← Previous day | Today | Next day → | Search | Index