IRC log for #koha, 2014-07-28

All times shown according to UTC.

Time S Nick Message
00:45 Mahdi joined #koha
00:46 Mahdi Hi there
00:46 Mahdi I am new to Koha, just installed it and got it up and running to test things out. However I'm having issues cataloging new books
00:46 Mahdi anybody here who can assist me
00:49 eythian wahanui: ask
00:49 wahanui Don't ask to ask, just ask.
00:52 Mahdi lol k
00:53 Mahdi basically I tried using z39.50 search to import a book
00:53 Mahdi and then i filled in the mandatory fields and saved it
00:53 Mahdi and later hit add item
00:53 Mahdi thinking it works
00:53 Mahdi but then it doesn't show up in the catalogue search
00:54 eythian I have to go for lunch, but describe your koha and linux installation/versions and the actual process you're using and the results you get, and someone may be able to make suggestions.
00:59 Mahdi running Koha version Using an AWS EC2 64 bit debian squeeze server.
00:59 Mahdi I just set installed things using linux commands I found online
01:00 Mahdi and now with an empty instance I wanted to start adding books
01:00 rangi hmm where did you get 3.15  from?
01:01 Mahdi i added the URL to my sources.list
01:02 Mahdi and then using some command instructions i found online i installed
01:02 rangi did you follow[…]on_Debian_Squeeze ?
01:02 Mahdi apt-get update
01:02 Mahdi yes essentially
01:05 rangi in that case, either zebra isnt running, or youve done something that stops the cronjobs running, or they are running with the wrong permissions, as otherwise it woudl show in your search results after 5-10 mins
01:06 Mahdi i just realize that the about koha page says zebra server seems to not be available
01:06 wizzyrea ah koha-start-zebra might fix you up then
01:06 eythian there is no in the packages
01:07 Mahdi ok i gotta go will try that command later
01:07 Mahdi appreciate the help
01:07 Mahdi thanks
01:08 rangi eythian: yeah i was puzzled by that too
01:08 wizzyrea seems odd.
01:09 eythian unless 3.16 in squeeze-dev identifies itself like that, but that'd be unexpected
01:09 rangi nope master is 3.17.000
01:10 eythian master is ahead
01:10 eythian as it doesn't build, last I looked.
01:12 rangi if you made one from master before 3.16 was released
01:12 rangi and havent since
01:12 rangi then yes it will be 3.15
01:13 rangi master always is on odd numbers
01:13 eythian I think it was an RC
01:14 eythian koha-common 3.16~git+20140521105956.3f9a8ccc
01:14 rangi yeah pre release, so yep that will be 3.15.something
01:18 * jcamins finds himself wondering how on earth someone ends up with squeeze-dev by accident.
01:20 eythian our $VERSION = ''; ... yep, that's it
01:20 eythian I'm curious how my package is labeled 3.16 then, it usually comes directly from that. Maybe I did it by hand or something, but it doesn't look like it.
01:20 eythian (on the other hand, it's the only explanation
01:20 eythian )
01:21 wizzyrea sunspots.
01:21 wahanui sunspots are what you get from staring at the sun for too long
01:49 Mahdi joined #koha
01:51 Mahdi_ joined #koha
02:01 tcohen joined #koha
03:09 eythian 3.14.09 uploading
03:10 eythian uploaded
03:10 eythian that was fast
04:49 Mahdi joined #koha
05:19 cait joined #koha
05:19 Mahdi joined #koha
05:53 cait good morning #koha
05:53 cait @later tell tcohen - can i pay in cookies? better not get the RM drunk
05:53 huginn cait: The operation succeeded.
06:33 cait @later tell oleonard could you take a look at bug 9043 maybe?
06:33 huginn cait: The operation succeeded.
06:40 cait left #koha
06:44 jseplae joined #koha
06:52 sophie_m joined #koha
06:57 alex_a joined #koha
06:58 alex_a bonjour
06:59 ashimema joined #koha
07:03 yohann joined #koha
07:10 cait joined #koha
07:10 cait good morning #koha
07:11 sophie_m hello #koha and cait
07:15 cait hi sophie_m
07:28 ashimema morning cait
07:28 ashimema morning all
07:32 hahaha joined #koha
07:32 hahaha hello
07:33 hahaha vietnam
07:34 gaetan_B joined #koha
07:34 gaetan_B hello
08:00 paul_p joined #koha
08:55 atheia joined #koha
09:41 willem joined #koha
09:41 willem hi there, I'm working on a php>mySQL connection to update koha automatically from another database. Someone warned me not to fiddle with koha tables as this might interfere with backups, or shadow files (something about Zebra). Is that indeed the case, or can I freely change info e.g. borrowers details or add borrowers? Thanks, Willem
09:51 cait borrowers is pretty safe
09:51 cait it depends on the tables
09:51 cait you shoudln't mess with bibliographic data
09:52 cait but not using an existing API has always the danger of things changing on updates
09:52 magnuse joined #koha
09:53 * magnuse waves
09:54 cait hi magnuse
09:54 willem OK thanks, @cait
09:55 magnuse kia ora cait
09:56 cait hi magnuse :) getting bored already? ;)
09:56 atheia Hi magnuse
09:57 cait hi atheia :)
10:00 magnuse cait: um, nah ;-) just dropping by while mini me is sleeping
10:00 magnuse kia ora atheia
10:00 cait heh
10:22 ashimema hi magnuse.
10:41 magnuse hiya ashimema
10:41 * magnuse runs off
10:47 bgkriegel joined #koha
10:50 bgkriegel hi
10:53 cait hi bgkriegel :)
10:53 cait and bgkriegel++
10:53 bgkriegel hi cait :)
10:54 cait thx for your cgi patches and all the sign offs and ... :)
10:56 sophie_m joined #koha
10:56 bgkriegel np :), and thanks to you
11:11 ashimema tcohen++ for taking up the slack in writing some ldap tests
11:21 marcelr joined #koha
11:21 marcelr hi #koha
11:27 collum joined #koha
11:27 clrh hello all
11:27 clrh khall: I am just asking why bug 11988 is "blocked" ? what does it mean ?
11:27 huginn Bug[…]_bug.cgi?id=11988 enhancement, P5 - low, ---, jonathan.druart, BLOCKED , Display basket group close date on late orders
11:28 clrh because "depends on" ?
11:32 paul_p joined #koha
11:49 cait clrh: taking a look
11:49 cait hi marcelr :)
11:50 marcelr hi cait
11:50 cait clrh: i thik because of the needs sign off on the depending bug
11:50 cait clrh: because it can't be pushed before the other bug starts moving
11:50 marcelr cait, clrh: not sure if keeping things in blocked is a great solution for that
11:51 cait marcelr: the other way is that it is in the pqa - but can't be pushed
11:51 cait i think it was moved to stop people getting irritated by it
11:51 clrh yep ok cait I understand, thanks ;)
11:52 cait marcelr: not ideal yes, but not sure we have another way to indicate - not yet
11:52 cait marcelr: i hope the conflict on your patches was not too big
11:52 marcelr not really
11:52 cait we have quite a few bigger things waiting
11:53 clrh if someone look at "blocked" when something is pushed, it is ok
11:53 marcelr yeah well hopefully the author still remembers :)
11:54 cait true
11:55 cait marcelr: Joubu had asked me to take a look at 2 patches for marc modfications, do you think you might have time to look at those?
11:55 marcelr depends
11:55 marcelr for most on the amount of time needed
11:55 cait i will show you the bug numbers one sec
11:56 cait bug 11319
11:56 cait bug 11413
11:56 huginn Bug[…]_bug.cgi?id=11319 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Marc modification templates improvements
11:56 huginn Bug[…]_bug.cgi?id=11413 normal, P5 - low, ---, jonathan.druart, Signed Off , Marc modification template has not the expected behavior if the condition and source fields are on the same field
11:56 cait i am not sure if it touches your area, probably not, but I haven't checked
11:57 cait i am aiming to start work on bug 11126 this week again
11:57 huginn Bug[…]_bug.cgi?id=11126 enhancement, P5 - low, ---, kyle, Signed Off , Make the holds system optionally give precedence to local holds
11:57 marcelr ok i will try to look at those this week somewhere
11:58 cait that would be cool
11:58 cait i have family visiting this week, so not a lot of time overall, so not sure how far i will get
11:58 marcelr no guarantees :)
11:59 cait none expected:)
11:59 cait just let me know if you can't make it at all, and i can check with you when done with the other...
11:59 cait i am a bit scared of everything touching holds
12:00 meliss joined #koha
12:01 marcelr holds is fun :)
12:01 ashimema any apt pinning pros here today?
12:02 marcelr two turns on the left please :)
12:02 cait apt ?
12:04 jcamins ashimema: the trick is to use large pins.
12:05 ashimema ;)
12:05 ashimema thanks jcammins..
12:05 jcamins What are you trying to pin?
12:05 ashimema trying to pin it such that our koha-common only pulls from our local koha repo.. but still allows the dependances from the community repo.
12:06 jcamins Hm.
12:06 ashimema of course the better approach would probably be to actually build the dependancies into our local repo too.. but i've not got to that yet.
12:06 jcamins I did that before.
12:06 ashimema :)
12:06 ashimema that's good news
12:06 jcamins I don't see why you need to build the dependencies.
12:06 jcamins Not that I can remember where I did it.
12:09 ashimema at the moment to run your own repo, you have to include both your repo and the community repo to pull koha-common dependencies from.
12:09 ashimema but yeah.. by haivng the community repo in my sources too.. by default when they update their pacakges, then they get  higher version number than mine so get installed nuking our customisations..
12:10 ashimema all good fun playing with pacakges.
12:10 jcamins Yeah, but if you have pinning configured it's not a problem, and who wants to repackage things unnecessarily?
12:10 ashimema indeed..
12:11 carmenh joined #koha
12:11 ashimema just means I've got to get my head around pinning again..  one of those things I touch about once every couple of years at the moment.
12:11 cait hm wasn't there a kopa-deps package or similar once?
12:12 alex_a joined #koha
12:14 pastebot "jcamins" at pasted "ashimena: I think?" (3 lines) at
12:25 francharb joined #koha
12:30 ashimema thanks jcammins.. i'm getting closer with that as a guidline..
12:30 ashimema but I don't think 've got here yet.
12:31 ashimema do you happen to know of any documentation that says what the different things you can put after 'release' are?
12:31 ashimema I've seen o=, n= and a= so far..
12:31 ashimema not sure what any of them mean.
12:31 jcamins I do not.
12:32 tcohen joined #koha
12:33 tcohen morning
12:34 tcohen hi cait
12:35 Dyrcona joined #koha
12:35 ashimema aha.. found it in the docs.. it's cryptic.. but its there
12:35 ashimema morning tcohen
12:35 tcohen hi ashimema
12:36 cait good morning tcohen
12:37 cait tcohen: what prompted the jag comment you left for me? :)
12:42 oleonard joined #koha
12:42 cait hello oleonard :)
12:42 tcohen hi oleonard
12:43 oleonard Hi
12:43 tcohen ldap?
12:43 wahanui it has been said that ldap is checked first, then it falls back to local Koha authentication
12:43 tcohen ldap?
12:43 wahanui ldap is checked first, then it falls back to local Koha authentication
12:44 tcohen ldap configuration?
12:45 tcohen cait, ashimema: link to LDAP configuration sample?
12:45 nengard joined #koha
12:45 ashimema as in a configuration you can test against.. or just an example of a configuration?
12:46 tcohen an example configuration
12:46 tcohen for writing unit tests
12:46 cait tcohen: i am not able to test aganst ldap
12:46 cait :(
12:46 ashimema There's a basci example of how to configure it in the perldoc for
12:46 cait ashimema might be more helpful
12:46 ashimema i believe
12:46 tcohen i'll mock the ldap conexion
12:46 cait i believe so too
12:46 cait and some in koha-conf i think?
12:47 ashimema awesome.. I took a quick look at the ldap::mock module thne got your email..
12:47 ashimema thanks for looking into it tcohen.
12:47 cait ashimema can then resuse your work for shibboleth *hides*
12:47 ashimema indeed..
12:47 ashimema I need to get my head around tests in general, then write them for shib too
12:53 druthb joined #koha
13:09 ssa joined #koha
13:10 francharb joined #koha
13:12 pastebot "ssa" at pasted "cataloque problem" (2 lines) at
13:17 kmlussier joined #koha
13:19 kmlussier joined #koha
13:25 cait ashimema++
13:25 ashimema what I do?
13:25 cait i won't tell :P
13:29 druthb cait++
13:33 cait he did do something!
13:34 druthb Wait, I need a *reason* to ++ you, cait?  Who knew?
13:52 ashimema hehe
13:56 kmlussier joined #koha
14:04 NateC joined #koha
14:06 gerundio joined #koha
14:10 fALSO joined #koha
14:10 fALSO Hi there!
14:10 fALSO Im having a little problem: on OPAC, when acessing a Subscription record, and clicking in "Serial colletion"
14:10 fALSO Template process failed: undef error - Invalid local time for date in time zone: Europe/Lisbon
14:10 fALSO Has anyone had this problem ?
14:11 cait fALSO: sorry, haven't seen that so far
14:11 cait which version are you using?
14:12 fALSO
14:12 cait hmm
14:13 cait fALSO: does it happen for every subscription or only a special one?
14:13 fALSO It was already reported on the ML but no fix- from what it seems
14:13 cait i haven't seen it on the list, hard to tell
14:13 fALSO http://koha.1045719.n5.nabble.[…]on-td5774440.html
14:14 fALSO 2011-09-14T11:09:01  <agJohn> In case anyone's interested, the mess with the "Invalid local time for date in time zone" seems to be resolved by upgrading to the 0.70 version of
14:15 fALSO found this on a log
14:15 cait ah
14:15 cait i am not here all the time :)
14:15 cait even if some people might say something else
14:15 fALSO But I installed Koha via apt-get
14:15 fALSO I cant force a upgrade of a module...
14:16 khall patches to update to TT submitted! Bug 9000
14:16 huginn Bug[…]w_bug.cgi?id=9000 normal, P5 - low, ---, kyle, Needs Signoff , misc/cronjobs/rss/ uses HTP
14:20 cait khall++
14:20 cait fALSO: i think best would be to file a bug
14:20 cait on bugzilla
14:21 cait with your findings
14:28 cait khall: around?
14:28 khall yo
14:30 fALSO Well that "fix" doesnt work
14:30 fALSO package DateTime;
14:30 fALSO {
14:30 fALSO $DateTime::VERSION = '0.75';
14:30 fALSO }
14:30 fALSO I already have a "newer" version that 0.70
14:31 fALSO than that
14:32 fALSO Interested on the backtrace? - some it if
14:32 fALSO
14:58 tgoat joined #koha
15:12 nengard hey all do we have minimum required versions for Perl and MySQL for Koha?
15:12 gerundio joined #koha
15:13 cait nengard: i think we normally add to the release notes if a newer version is required
15:13 nengard okey dokey, will check there
15:14 tcohen joined #koha
15:14 jcamins Perl 5.10, and MySQL 5+. But really if you use MySQL 5.0 your life will be sad so you should use 5.5.
15:16 nengard thanks jcamins
15:16 jcamins And there was a report of MySQL 5.6 not working, but that was also a non-Debian-based distribution of Linux, so I don't know if that's a hard and fast rule.
15:16 jcamins And, of course, MariaDB works just fine too.
15:17 jcamins Hm. I'm using MariaDB 5.5 and not 10.0, so I have no idea if MariaDB 10.0 works.
15:17 jcamins Maybe tcohen has tested it.
15:18 * tcohen is setting a Debian 7 + MariaDB node for jenkins, but haven't tried it yet
15:19 tcohen There are known issues with MySQL 5.6, and it is discouraged
15:20 nengard I'll tell the library that's asking 5.5
15:26 tcohen nengard: 5.6 changed some policy that prevents some stuff we usually do get done
15:26 cait ah
15:26 cait that was the webinstaller problem?
15:27 tcohen yes
15:36 tcohen jcamins++
15:46 cait left #koha
15:54 tcohen do we have test::mockobject¿
15:58 tcohen ashimema: could you please add me to the QA Trello board?
15:58 tcohen (so I can put my comments there?
16:00 ashimema done ;)
16:05 ashimema hmm.. no cait.
16:05 wahanui well, no cait is here
16:15 * oleonard is reminded of the Hitchhiker's Guide to the Galaxy Infocom game in which the player (as Arthur Dent) has in his inventory "No tea"
16:33 gaetan_B bye!
16:34 gerundio hey everyone
16:34 wahanui everyone is, like, doin' the hot new dance the Cracked Out Kitty Tail Shiver
16:34 Mahdi joined #koha
16:34 gerundio tricky problem about Koha underlying date time parse
16:34 gerundio[…]/lisbon?year=1979
16:34 gerundio according to that source after 23:59:59 on the 1979-04-01 in Portugal came 01:00:00
16:34 gerundio which means that 00:00:00 didn't exist
16:35 gerundio perl's DateTime::Format::DateParse::parse_datetime throws an error when we insert 1979-04-01 without any timestamp to it
16:35 gerundio which means that it adds 00:00:00 to verify the date
16:35 gerundio this is the expected result, but how should Koha respond to this?
16:35 gerundio I have a subscription starting at 1979-04-01 which results in: Software error: Template process failed: undef error - Invalid local time for date in time zone: Europe/Lisbon
16:48 gerundio the hour 1979-04-01 00:00:00 doesn't exist but Koha wanted to know if the day 1979-04-01 is valid... and it is
16:56 nengard hi all, is something up with the wiki or is it just me? I don't see the menu on the left or the login link on the top anymore
16:57 oleonard Looks okay to me nengard
16:57 nengard hmmm
16:57 nengard my computer pissing me off lately
16:57 nengard be right back
16:59 nengard joined #koha
16:59 nengard okay all back to normal
17:10 atheia left #koha
17:47 gerundio joined #koha
17:49 pianohacker joined #koha
18:18 Mahdi joined #koha
18:51 kmlussier joined #koha
18:59 cait joined #koha
19:18 francharb joined #koha
19:21 nengard oleonard any chance you have a good tutorial on how to make a 2 column responsive table that I could put in the opacmainuserblock
19:25 oleonard Responsive in what way nengard?
19:25 nengard so that the 2 columsn become 1 if necessary
19:25 nengard columns
19:25 cait one moving below the other?
19:26 nengard yes
19:26 nengard sorry for being unclear
19:29 oleonard nengard: You probably don't want a table, you probably want to use one of the built-in Bootstrap layouts
19:33 * oleonard is trying to find a good example
19:35 bgkriegel joined #koha
19:37 pastebot "oleonard" at pasted "2 equal-width columns which stack if necessary" (5 lines) at
19:37 nengard I used the word 'table' loosely :)
19:38 nengard so bootstrap has that built in?? That span6 says to make it 2 equal columns?
19:38 nengard and if so is there documentation on that? So like if I wanted 3 columns or 4 or whatever
19:38 oleonard nengard: Yes. 6 of 12 total columns.
19:38 nengard omg!!!
19:38 oleonard nengard: You can use any number of spanX divs adding up to 12
19:41 nengard that is SOOO Awesome :)
19:41 nengard I think that might be a call for a tutorial article
19:41 nengard :)
19:41 cait oleonard: do you need css to tell it when to break or is that already done?
19:42 oleonard The default Bootstrap CSS has a built in breakpoint for it but I'm not sure what the exact number is off the top of my head
19:43 oleonard Bootstrap 3 offers better control for more than one breakpoint
19:43 cait hm i haven't really worked with responsive stuff yet
19:43 cait but i will have to learn :)
19:43 nengard oleonard++
19:43 cait so would i want to accopany that html with some css to make it do waht i want?
19:43 tcohen @later tell ashimema i've pushed my work on ldap tests to my github repo, named qa_8148_ldap
19:43 huginn tcohen: The operation succeeded.
19:44 tcohen @later tell ashimema it is WIP, but I'm getting closer
19:44 huginn tcohen: The operation succeeded.
19:44 cait or should it behave reasonably fine in that use case without doing more?
19:45 oleonard You just have to test it to see if the built-in breakpoint works well for you.
19:45 tcohen cait: please behave
19:45 * pianohacker just came up with an evil master plan
19:45 cait oleonard: ok :)
19:45 cait and i am looking for bootstrap and breakpoints if i want to change it
19:45 cait tcohen: what did i do?
19:45 cait anyway... brb :)
19:45 tcohen "16:44 <cait> or should it behave reasonably fine..."
19:46 pianohacker I can get the web-based syntax highlighting editor library I'm using for Rancor integrated by adding it to the system preferences editor! muahaha
19:46 tcohen bye #koha
19:46 pianohacker this calls for a celebratory espresso, brb
19:47 tcohen pianohacker++
19:47 oleonard What are you using pianohacker?
19:47 ashimema cheers tcohen..
19:47 pianohacker oleonard: codemirror ( It's pretty magical
19:47 oleonard Oops. I have to leave!
19:48 ashimema Interesting enough pianohacker..
19:48 ashimema i've got a branch doing exactly that somewhere..
19:48 ashimema it's at least a couple of years old mind ;)
19:48 ashimema made the html, css and jscript block much nicer to edit..
19:49 ashimema though I never thought it would get in without a more reaosnable need for the library (like your rancor project) so iwas doing the oposite.. waiting for rancor to get in ;)
19:50 cait heh
19:53 ashimema I tell a lie pianohacker.. it was ace code editor I played with.. not codemirror..
19:53 ashimema[…]s/ace_integration
19:53 ashimema I look forward to seeing it done 'properly' though ;)
19:54 ashimema codemirror is an awesome project.. I'd love to see it in koha.
20:01 Mahdi joined #koha
20:04 pianohacker hi ashimema :)
20:04 pianohacker how's the shib stuff going?
20:04 ashimema it's all good fun mate..
20:05 ashimema I've been back on it again relatively recently..
20:05 ashimema the base patch has morphed a bit with bug fixes here and there as I've come across them..
20:05 ashimema We're using it in production in a few places now.. even have it working on a multi-tenancy setup via pacakges!
20:06 ashimema I need to write some unit tests for it before it'll ever pass QA though..
20:06 cait and on my test server :)
20:06 ashimema and I'de like to put some of the base work for allowing you to select matching attribute through the config as early as possible.. currently that's very hackish..
20:06 Mahdi joined #koha
20:06 ashimema how about you pianohackers.. hows your world.
20:07 ashimema s/pianohackers/pianohacker/
20:08 ashimema I'll need to rebase your patches on it at some point too..
20:08 Mahdi joined #koha
20:08 ashimema I opened up a couple of bugs a while back with your code in.. for auto-provisioning and staff client..
20:09 ashimema the staff client one I think will need some hacking.. but the auto-provisioning I think should re-apply reasonably easily
20:10 pianohacker oh, excellent.
20:12 cait pianohacker: how is rancor?
20:12 ashimema Hope you don't mind.. I grabbed your commits from your repo and pushed them up to bugzilla for those patches.. so I could continue to work on them.. but give you initial attribution
20:13 pianohacker cait: good, currently redoing the macro language due to some intellectual property concerns
20:14 cait ah yeah
20:14 cait that would be better
20:14 pianohacker ashimema: don't mind at all :)
20:14 cait jus tdon't want too long with submiting
20:14 cait i really want to see it happen for 3.18
20:14 cait i tihnk macro could go separate maybe?
20:14 pianohacker same
20:15 pianohacker the macro stuff is leading some refactoring that really needed to happen, but we'll see. I'm trying to get another patchset done before school starts
20:31 ashimema intellectual property aye..
20:31 ashimema they didn't like you ripping off their macro language then..
20:31 ashimema I always find that a bit of an odd complain.. hey your making your app compatible with ours.. don't do that.
20:32 ashimema I suppose in this case your taking away their custom.
20:34 pianohacker well, it wasn't a sure thing, but I was talking to the funding company onsite and they mentioned that the license had something against that
20:34 pianohacker I think the new one is much better anyway, but I'm biased
20:35 ashimema nice.
20:35 ashimema well documented too I presume ;P
20:36 ashimema I'm really looking forward to having a really standardised macro language for our marc manipulation.
20:36 pianohacker it will be :)
20:36 pianohacker here, let me throw the example I made into a pastebin
20:36 ashimema would love to see some of the bulk modification stuff start to use it so we have one way to do it throughout.
20:37 ashimema awesome..
20:37 wahanui That'll be $1 for the awesome jar, ashimema
20:37 * ashimema is getting really distracted tonight.. these upgrades are going to take forever.
20:38 pastebot "pianohacker" at pasted "ashimema: new rancor macro language (wip)" (45 lines) at
20:47 ashimema looks interesting pianohacker..
20:47 ashimema I can't remember the original syntax to be fair so have nothing to compare to ;)
20:49 pastebot "pianohacker" at pasted "ashimema: original" (74 lines) at
20:49 pianohacker and OCLC's is, as far as I can tell, built on Visual Basic :P
21:17 ashimema I see..
21:17 ashimema Your syntax is certainly less verbose.. but I'm feeling it's going to be more complex to learn.
21:19 pianohacker yeah, that's the tradeoff
21:20 ashimema all good fun..
21:20 ashimema I'd have probably stuck with the keywords from crud.. create, read, update and delete.. but then I've been working with restful api's too much of late ;)
21:21 ashimema how do you navigate in your syntax.. and can you still copy/paste?
21:21 ashimema ah.. yeah.. first line for copy/paste..
21:21 * ashimema feels silly for missing that
21:21 ashimema actually..the copy paste syntax is much much nicer.. I like.
21:22 pianohacker the main idea was to allow actual text manipulation
21:22 pianohacker 260c = {260c} [electronic editiony
21:22 pianohacker typo aside there's the idea
21:23 ashimema how would you go about changing just the seocnd occurence of a tag?
21:24 ashimema is there a 'next' keyword for instance?
21:25 mtompset joined #koha
21:25 pianohacker currently none. there's some ideas on referring to specific subfields but nothing for specific tags yet. still fleshing out that part
21:25 ashimema coolios..
21:25 mtompset Greetings, #koha.
21:25 ashimema well it all looks very interesting..
21:25 mtompset Greetings, ashimema.
21:25 ashimema will follow it with baited breath.
21:25 ashimema hi mtompset
21:26 mtompset Has anyone ever made a change to a template page, only to have the unchanged source still show up on the page when you reload it in the browser?
21:26 mtompset This is rather frustrating, and I'm positive I forced the reload.
21:27 cait ?
21:27 cait are you looking at the right opac template?
21:28 cait the right file?
21:28 cait which one is it?
21:28 jeff occasionally. good idea in that case to 1) ensure that your browser isn't being annoyingly cache-happy, 2) ensure that the file you're editing is the same across all servers (if dealing with a multi-server/multi-brick/whatnot setup) [er, possibly disregard #2 -- this is #koha and not #evergreen) and 3) ensure that you're editing the right file in the first place -- i've made that mistake more than twice.
21:29 jeff (surprisingly, quite similar troubleshooting steps despite my confusion over which channel i was reading)
21:30 mtompset Actually, I opened up IE and it displayed the correct change, so I suspect Firefox is caching on me.
21:30 mtompset Any idea on how to tell Firefox to stop it?
21:34 mtompset There... much better. :)
21:35 cait what did you do? :)
21:35 mtompset It was caching the page.
21:35 mtompset I forced firefox to clear its cache.
21:35 mtompset and reloaded the page.
21:40 mtompset And it is the offline circulation page... I was looking at bug 10466.
21:40 huginn Bug[…]_bug.cgi?id=10466 minor, P5 - low, ---, dcook, RESOLVED FIXED, Lists: Use "size" as names/hash keys leads to an unexpected results when using Template::Toolkit (name of a virtual method there)
21:40 mtompset I mean 10446.
21:40 mtompset I mean bug 10446.
21:40 huginn Bug[…]_bug.cgi?id=10446 trivial, P5 - low, ---, kyle.m.hall, Failed QA , System preference maxoutstanding is poorly named
21:57 cait hmmm
21:57 cait someone else seeing broken diacritics in marcxml exports from the detail page?
21:59 mtompset is this on a hidden record? :)
21:59 cait hm a normal one in opac
21:59 mtompset no hidden items involved?
22:00 cait hm i don't think so
22:00 mtompset let me see... broken diacritics...
22:00 mtompset detail page.
22:00 wahanui i think detail page is great
22:00 cait all other exports are ok
22:00 cait the only one giving me trouble is marcxml
22:01 cait <subfield code="a">The R&#xE4;uber</subfield>
22:02 cait that should have been: The Räuber
22:03 mtompset so export in MARCXML?
22:05 cait yep
22:05 mtompset But isn't &#xE4; ä?
22:07 mtompset I have the same ugliness... my diacritics view just fine looking at the marcxml file in IE, but I have the &#...; encoding.
22:08 cait hm
22:08 cait confusing
22:08 cait i am not sure if this is a bug or not, but i don#t rmember it doing this in 3.12
22:08 cait i use the marcxml export a lot to check files
22:09 cait hm i am wrong
22:09 cait it's like this there too
22:09 cait i think it#s time to sleep
22:10 mtompset Have a good rest, cait. :)
22:11 cait thx!
22:11 cait left #koha
22:19 eythian hi
22:21 ashimema I must be up late...
22:21 ashimema there's an eythian at the beginning of his day instead of the end.. I don't see that very often these days.
22:21 ashimema hi
22:22 eythian heh
22:22 eythian hiya
22:26 ashimema at some point, when I'm more awake, eythian I may want to pick your brains about packages and adding shibboleth support to them.
22:27 eythian shibboleth the authentication system?
22:27 ashimema the bit I've not worked out.. tbh not looked at yet.. is that I'de like to add commented vhost configs per instance to the virtualhosts during koha-create.
22:28 ashimema yes.. shibboleth as in authentication
22:28 eythian Oh OK.
22:28 ashimema :)
22:28 eythian that probably wouldn't be too hard.
22:28 ashimema my patches have been slowly working their way to being pretty stable..
22:28 eythian Hey, do you happen to know M5? I want to move the templates to that at some stage.
22:29 eythian Maybe this could be an excuse...
22:29 ashimema erm.. pass I'm afraid.
22:29 eythian ah well
22:29 ashimema got a link?
22:29 eythian it'd actually look almost the same as it would now, but will let us do a bit more
22:30 eythian oh, M4, not M5.
22:30 eythian
22:31 ashimema ahh.. I see
22:31 ashimema cool
22:31 pianohacker eythian: the only quirk with shibboleth as we've figured out how to do it is that it depends on an external daemon (shibd). Not a new thing but a quirk to deal with
22:31 ashimema another one to add to my list of things to learn.
22:31 pianohacker eythian: and is this the same terror involved in autoconf?
22:31 eythian ashimema: I used it for something else a few years about, it took about 10 minutes of skimming the docs to understand it.
22:32 eythian pianohacker: other daemons are fine, the init system will manage them.
22:32 ashimema indeed.. it does rely upon the standard packaged shibboleth service provider packages in debian.
22:32 eythian pianohacker: yeah, and sendmail configs.
22:32 ashimema it shouldn't matter in the packages though..
22:33 eythian No, I wouldn't think so.
22:33 pianohacker so... not necessarily bad in of itself but we have a couple slightly overcomplicated examples to avoid?
22:33 eythian yeah
22:33 pianohacker and are you talking about the .tt files or the config file templates?
22:34 ashimema we'd just need to document that to use it, you'de need to uncomment the vhost configuration bit.. add some config to koha-conf (like you already have to for the ldap auth's and similar) and install the shibboleth daemon packages..
22:34 NateC joined #koha
22:35 eythian so at the moment we're doing sed search-and-replace, instead of doing that we'd emit some defines, but I think this'll let us turn on/off blocks or comments in a nicer fashion.
22:35 eythian pianohacker: the config file templates
22:35 ashimema sounds useful
22:35 pianohacker ahhh, okay. Yeah, that's definitely better than our current hacked-together system
22:36 ashimema indeed...
22:36 eythian it won't allow updating config files, though I see magnus has done some stuff towards that.
22:36 eythian (for koha-conf specifically)
22:36 ashimema go magnus++
22:36 ashimema I wish I had more time to work on koha at the moment..
22:36 eythian ashimema: what vhost configuration does it require?
22:37 ashimema currently It all quick dives in and out for upgrades.
22:37 ashimema for single instance it's a simple
22:38 ashimema <Location />
22:38 ashimema AuthType shibboleth
22:38 ashimema require valid-user
22:38 ashimema </Location>
22:38 ashimema or some such..
22:38 ashimema but for multi-instance you'de need to set an applicationID for shibboleth to bind to.
22:38 eythian Oh, I see
22:39 eythian[…]-perl_1.03-1.html <-- pianohacker, that made it to the NEW queue once I fixed up some issues
22:39 ashimema so.. for example I have the following in one of our multi-tenant systems
22:39 ashimema # Optional Shibboleth Configuration
22:39 ashimema Alias /shibboleth "/home/koha/shibboleth/"
22:39 ashimema <Location />
22:39 ashimema ShibRequestSetting applicationId bmcc
22:39 ashimema AuthType shibboleth
22:39 ashimema ShibRequireSession Off
22:39 ashimema #Require valid-user
22:39 ashimema Require shibboleth
22:39 ashimema </Location>
22:40 pianohacker thanks eythian!
22:40 ashimema though I'de need to remind myself what all that does before advising to use it yet ;)
22:40 pianohacker I really don't think I can convince mod-gzip to do this, so that's really helpful
22:40 ashimema you using shib in multi-tenant anywhere piano
22:40 pianohacker no, not at the moment. Can't you tie the application ID to the URL in the shib config, though?
22:41 ashimema you can indeed.. but it makes the configuration at the shibboleth deamon end much much much more difficult and error prone.
22:42 ashimema I've found tiling to applicationID's a massive amount easier.
22:42 pianohacker haha I'll take your word for that. I walk a very narrow path through the shib docs and try to ignore the darkness on either side
22:43 ashimema yeah.. unfortunately I've had a rather large number of implementations for it recently.
22:43 ashimema have ended up delving far deeper than i'de have liked to ;)
22:43 eythian I didn't realise it was very common at all
22:43 ashimema pretty much all the uni's in the UK use it now..
22:43 ashimema and allot of the big firms use it.. we have a few law firms with it.
22:44 pianohacker yup, shibboleth is (unfathomably) the easiest way to tie into a lot of other auth systems
22:44 pianohacker using a second piece of middleware one of our clients is authenticating users in Koha against active directory (automatically!)
22:44 ashimema on the other hand.. I'm yet to hear of a single place in the UK using CAS... so it wouldn't surprise me if it's a rpetty british thing to use shibboleth
22:45 ashimema indeed.. we have shib's authenticating against active directory.. openid.. google account, facebook accounts.
22:45 ashimema it's a remarkably used system over here.
22:46 eythian interesting. Here we're usually doing LDAP/AD
22:46 pianohacker ashimema++
22:46 ashimema I've got a few test IdP's up these days.. means I can login to a fairly large number of our koha customer using my google login credentials ;)
22:46 pianohacker ^ needs to be said. I was trying to do SAML directly, and he pointed me at it in a chance conversation at KohaCon '13
22:47 eythian We've been talking about implementing SAML into Koha for years now.
22:47 ashimema We have a fair few ldap's too.. but peeps seems to be getting skittish about the security of their username/password going onto remote servers.
22:47 ashimema shibboleth = saml in my book..
22:47 pianohacker Don't try to do it using Net::SAML or Net::SAML2. The first is confusing and doesn't support SAML 2, the second is nice and easy but very limited
22:47 ashimema it's a common misconception in my opinion that they are different things..
22:48 pianohacker but they are, though
22:48 ashimema saml is the actual token that gets passed around.. shibboleth is how it gets passed..
22:48 pianohacker SAML's a protocol, shibboleth is a tool that implements it
22:48 eythian We're also looking at doing SSO within an AD environment some time soonish (SSO, not just authentication.) It actually looks fairly easy and we have one of the Samba devs on staff here.0
22:48 ashimema that's my understanding. ;)
22:48 pianohacker oh heh jinx :)
22:48 eythian I didn't know that
22:48 ashimema shibboleth/saml is sso ;)
22:49 pianohacker eythian: how are you looking at passing credentials from client -> server?
22:49 nengard left #koha
22:49 pianohacker our client used Okta's browser plugin -> Okta -(SAML)> Shibboleth -> Koha
22:49 eythian pianohacker: some magic I don't understand. Mostly it'll be done by apache.
22:49 ashimema for instance.. that law firm I was talking about.. they login to their windows computer at the beginning of the day and don't have to type a password for the rest of their day.
22:49 eythian I think it's an IE/NTLM thing
22:50 ashimema We didn't have to worry about that side.. but yeah.. they're also using IE/NTLM
22:51 ashimema crazy clever stuff..
22:52 ashimema would be awesome to understand how it works with samba though.. if you get around to it ;)
22:54 ashimema pianohacker.. are your shib customers all using cardnumber or userid to match upon?
22:54 eythian Apparently it's pretty easy, an apache module just passes the username to Koha if it's valid, like SSL client certs, or basic auth do.
22:54 ashimema we've got a few that are using email and now I'm seeing a need for matching to an extended patron attribute..
22:54 pianohacker ashimema: userid/email
22:55 eythian matching an extended attribute? That sounds messy.
22:55 ashimema take it your just using the basic patch then.. which currently checks whatever is passed to it against cardnumber.. then if no match userid (and you must have added email)
22:56 ashimema I'm wanting to add 'which should I match against' t the config but haven't gotten around to it yet.
22:57 ashimema shouldn't be that messy to include one of the extended attributes eythian.. afterall, you can set an attribute to be enforced unique.
22:57 eythian I did that for the SSL client cert stupport
22:57 ashimema we use attributes in a few cases as a match point for some institutions for our background user data updates.
22:57 eythian there's a syspref for it
22:57 ashimema is the the pki_auth stuff.
22:57 ashimema is the the pki_auth stuff?
22:59 eythian yeah
23:01 ashimema I shall take a look for inspiration ;)
23:03 ashimema hmm.. no auth_with_pki :(
23:03 ashimema I assume that means it's all in
23:04 eythian I think so. It's been some years since I wrote it
23:05 ashimema If I were to refactor one of the first things I'd do is to try and factor out a fair chunk of that code ;)
23:05 ashimema haha..
23:05 ashimema but that's just a dream ;)
23:06 ashimema I think any code that's around 2000 lines in one file is usually a tangled mess that can be split into more manageable chunks.
23:07 eythian yeah
23:07 eythian really, it needs plugins or something
23:07 ashimema indeed it does..
23:07 eythian so you can just switch on the auth methods you want
23:07 ashimema the Auth_with_x series isn't terrible.. but their not great either to be fair
23:12 tcohen joined #koha
23:13 Mahdi joined #koha
23:14 pianohacker It'd be kind of nice if Koha had a standard hook/lightweight AOP system
23:14 tcohen night
23:15 * tcohen thinks mvc could be a good start
23:15 pianohacker Koha::Hook->attach('before_checkpw', sub { ... }) # for standard registered hooks
23:15 pianohacker a good idea, by far, though it addresses a different piece imo
23:16 pianohacker this could be a nice way to make CanBookBeIssued, for instance, a lot less messy
23:16 eythian I'm somewhat sceptical about AOP, it seems like it'd make debugging harc.
23:16 eythian d
23:16 eythian It's like INTERCAL's COMEFROM statement. And no one wants to be like INTERCAL in real life.
23:18 pianohacker haha yeah. That's why I'd say a more formal hook system would be a bit better
23:18 * ashimema shows his lack of a computing background by not really knowing what aop is
23:18 pianohacker Being able to just throw extra code before/after/around subs would be kind of cool but terrible to debug, yes :)
23:18 pianohacker ashimema:[…]ction-to-aop.html something like this, but with less java
23:19 eythian <-- perhaps this for more context :)
23:19 papa joined #koha
23:20 pianohacker "Threaded Intercal?" dear lord
23:20 eythian hi there papa
23:20 * ashimema runs away scared
23:35 tcohen ashimema: did you get some time to look at my branch?
23:35 ashimema had a very quick peak..
23:36 tcohen[…]tree/qa_8148_ldap
23:36 ashimema I left a comment on github actually..
23:36 ashimema idn't have time to really 'think' about the test though.. only comment was to point at where I'de started as we seemed to be going in different directions to start with..
23:37 ashimema no actually sure my comment make things better of worse though
23:37 tcohen how do i see the comment?
23:37 ashimema it's a comment on the last commit
23:37 ashimema[…]31eaaabaa8a20b7dd
23:38 ashimema basically I didn't get very far at looking at using Test::Net::LDAP::Util to Mock an LDAP server..
23:38 ashimema looks like your building a mock ldap object instead if I'm reading it right.
23:40 tcohen i am
23:41 ashimema after reading your code I wasn't sure which approach made more sense to me.. ;)
23:41 ashimema sorry.. I've not been of much help really.
23:43 tcohen i like that lib
23:43 tcohen i'm not sure how would we control its behaviour, htou
23:43 tcohen it seemed to me that we used a small set of methods/output vals
23:43 tcohen and having them mocked with Test::MockObject gave us more control
23:43 tcohen anyway, I like the lib
23:44 tcohen if you want to work with it, reuse my config mock
23:44 ashimema Yeah.. MockObject is certainly more controllable from our end.. but with that it's also more complex to setup?
23:45 ashimema I've not delved deep enough into that lib yet to work out whether we'de actually come up against it's limits.
23:46 tcohen ashimema: just take a look at the "bind" method for Test::Net::LDAP::Mock, it only returns a successful bind
23:47 tcohen so, we cannot write a regression test for the current bug :-D
23:49 ashimema ah.. yeah.. I see now..
23:49 ashimema that's actually pretty usless..
23:50 ashimema I hadn't got that far down the page yet :(
23:50 ashimema that's a shame.. it would have been great to have such a mocking tool
23:50 ashimema prebuilt
23:56 tcohen ashimema: i agree
23:57 tcohen the lib would work for !auth_by_bind,

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary