IRC log for #koha, 2015-11-11

All times shown according to UTC.

Time S Nick Message
00:03 pj left #koha
00:05 brick yes =)
00:06 dcook I'd probably spin up a new instance in unimarc and try that instead
00:06 dcook Converting between unimarc and marc21 isn't a pleasant, or even perfectable, automated task
00:09 brick Yes, perhaps unwrap it was... ohhh демоны кто это поднимал
00:39 dbs joined #koha
01:22 wahanui joined #koha
01:30 dcook One of the downsides of inheritance... trying to find the damned module where a function is initially defined!
01:30 dcook The error message is for IO::Socket::IP... which is a parent of Net::HTTP... which is a parent of LWP::Protocol::http:::Socket...
01:31 dcook But IO::Socket::IP is a child of Socket...
01:31 dcook Which is referencing some other module in some interesting way
01:33 dcook The problem is in Socket::getnameinfo which is just an alias for Socket::fake_getnameinfo...
01:33 dcook And I'm sure the problem is Socket::unpack_sockaddr_in()
01:34 dcook But cannot see where unpack_sock_addr_in() is defined...
01:34 dcook I wonder if that means it's a C function...
01:36 pastebot joined #koha
01:39 rocio1 joined #koha
01:39 JoshB joined #koha
01:40 JoshB_ joined #koha
01:47 Francesca joined #koha
01:53 eythian it sounds like one
01:54 Francesca @wunder wlg
01:54 huginn Francesca: The current temperature in Wellington, New Zealand is 13.0°C (2:30 PM NZDT on November 11, 2015). Conditions: Light Rain Showers. Humidity: 88%. Dew Point: 11.0°C. Pressure: 29.92 in 1013 hPa (Rising).
02:12 tcohen joined #koha
02:12 tcohen yawn
02:12 tcohen rangi, I'll take a look thanks
02:12 tcohen @wunder cordoba, argentina
02:12 huginn tcohen: The current temperature in Bo Altos de San Martin - NW, Cordoba city, Cordoba City, Argentina is 18.9°C (11:10 PM ART on November 10, 2015). Conditions: Scattered Clouds. Humidity: 80%. Dew Point: 15.0°C. Pressure: 30.18 in 1022 hPa (Rising).
02:16 dcook But usually with C functions there should be some Dyna Loader or whatever yeah?
02:16 * dcook waves to tcohen
02:17 eythian
02:17 tcohen hi dcook
02:17 * Francesca waves at dcook and tcohen
02:17 dcook eythian: Yeah, that's what I'm looking at now
02:18 tcohen hi Francesca
02:18 dcook I don't see the function definition in the source though
02:18 dcook[…]t-2.020/
02:18 dcook ahh
02:18 dcook EXLoader
02:18 dcook XSLoader*
02:18 dcook There we go
02:19 dcook[…]ket-1.5/Socket.xs
02:20 dcook Hmm might not have mattered anyway..
02:21 tcohen hey, 56.3% code coverage on the tests, great
02:22 tcohen dcook: what are you using such a low-level aPI for?
02:22 dcook EBSCO EDS is getting an error and I can't find where it's from
02:23 dcook EDS plugin*
02:23 dcook Getting a 500 error locally... and getting a HTTP::Response object back when trying to contact the API with the following error:
02:23 dcook addr is not a string at /usr/lib/perl5/5.20.1/IO/Socket/
02:23 dcook So I think it'll be... gethostbyname?
02:24 dcook The error message looks familiar
02:24 dcook or gethostbyaddr maybe
02:24 dcook If it's talking about addr
02:25 tcohen it reminds me name resolution issues
02:25 dcook Pretty sure it's a name resolution issue
02:25 dcook Just not sure how exactly
02:26 tcohen did u already test fixing an address in /etc/hosts? only to make sure?
02:26 tcohen nevermind, i'm too sleepy
02:27 dcook Looks like I"m looking at too old a version of Socket.xs
02:27 dcook See the error message now..
02:27 dcook xs_getnameinfo in Socket.xs...
02:29 dcook Always wondered why CPAN doesn't seem to show XS files in the interface..
02:29 dcook Guess you can download the tar.gz...
02:32 JoshB joined #koha
02:32 dcook newXS("Socket::getnameinfo", xs_getnameinfo, __FILE__);
02:32 dcook Booyah
02:33 dcook Anyway, gotta run..
02:34 sophie_m joined #koha
02:35 dcook Wow... I'd never heard of SvPOKp() before...
02:35 dcook
02:38 dcook So I'm guessing LWP::Protocol::http::_get_sock_info() might be the culprit...
02:43 dcook I'm thinking the server is having trouble looking up information about itself?
02:47 dcook Hmm might be any number of things..
02:50 tcohen dcook: format c:
03:01 tcohen joined #koha
03:05 dac joined #koha
03:25 dac tcohen: :p
03:28 eythian tcohen: can you add this patch?
03:28 eythian (I'm just feeling too lazy to make a bug for it.)
03:32 tcohen eythian: np, but maybe sent it as attachment so i do it once i wake up?
03:32 eythian kk
03:35 * tcohen loves the existence of HTML::TableExtract
03:35 tcohen @later tell rangi I will just get rid of using a regexp for extracting translations info
03:35 huginn tcohen: The operation succeeded.
03:37 tcohen night
03:44 eythian new master packages uploading
03:47 dcook eythian: Is there a way to use perl -d to step until it can't step anymore?
03:48 dcook Or do you have to "continue until position"?
03:48 mario joined #koha
03:48 mario .
03:50 eythian dcook: define "can't step any more"?
03:51 dcook In this case, I think a fatal error
03:51 eythian that terminates the program. You may as well just do "run" then
03:51 dcook "run"?
03:52 eythian well, press "c"
03:52 dcook Basically, I just want to step until it dies so I can see which line did it
03:52 eythian ah
03:52 eythian it doesn't break there per se
03:52 eythian doesn't the error message tell you that?
03:54 dcook Well, the error message tells me what line number of the module, but it's a pretty generic function
03:54 dcook So I'm hoping to look at the function that called that function
03:54 dcook As that'll tell me all the things
03:54 eythian what you should be asking is "how do I get a backtrace on an exception"
03:54 dcook hehe
03:54 dcook Yes, yes I should
03:54 dcook That would be great
03:55 eythian I can't answer that, but I'm sure I've seen that it's possible
03:55 eythian I think the Carp module can do it
03:55 dcook Yeah, I think it can too, although only if used in the right place?
03:55 eythian i.e. you load it in a special way with -M and it'll something something stacktrace
03:55 dcook Yeah?
03:55 dcook That'd be awesome
03:55 eythian[…]ing-a-Stack-Trace
03:56 eythian oh
03:56 eythian that's not what you need
03:56 eythian that's only good if it's already using that
03:56 dcook Nopes :/
03:56 eythian there must be some way, I'm sure
03:56 dcook I'd think so, eh?
03:56 * eythian hates modules that don't use carp to give good messages.
03:57 dcook ^
03:57 dcook Oh ho... did I find something..
03:57 dcook use Carp;
03:57 dcook $SIG{ __DIE__ } = sub { Carp::confess( @_ ) };
03:57 dcook Yes
03:58 dcook Beautiful
03:58 dcook eythian: That one is a keeper
03:58 eythian heh
03:58 dcook Fortunately, that confirms what I already suspected... really good having that confirmation though
03:58 dcook I wonder if I can get any more information..
04:00 dcook I was actually working at catching signals for INT and... TERM the other day
04:01 dcook Working on a listener for a unix socket but I couldn't trap the INT while listening on the socket
04:01 dcook Could catch the INT before or after I think... but while listening it was too far into the C I think to listen to Perl
04:02 dcook Of course, the documentation makes it sound a lot more technical than that :p
04:02 eythian heh
04:02 eythian it's not something I've really played with
04:03 dcook Hopefully you won't need to!
04:03 dcook I was thinking about that debugger/logging talk we saw though where he was talking about using IPC to communicate with daemons
04:04 dcook So I'm all like "oh yeah, I can use a unix socket for that, sure"
04:04 dcook Which actually works pretty well, although I need to do some more research into buffering from unix sockets...
04:05 dcook I know with a TCP socket that you need to read in the buffer as you can get multiple packets for a single message, but unsure about unix sockets...
04:05 dcook I think it's just a stream so you shouldn't have to buffer... you just read until you get a pre-defined terminator
04:06 dcook But yeah... I think signals still count as IPC and it's common to use SIG{TERM} to force a config reload or even a daemon restart... but I can't catch that SIG{TERM} while waiting for a unix socket connection
04:06 dcook Anyway... that's totally unrelated to today's issues ^_^
04:07 dcook Except catching the SIG{DIE} and using Carp on it... I wish that I would've thought to do that years ago!
04:07 eythian My understanding is you should treat TCP and UNIX sockets pretty much the same, but I'm no expert at this
04:07 dcook So far, that's my understanding as well, but yeah... more research needed
04:08 dcook I think the maximum size of the messages varies but I think there IS still a maximum and that a person needs to be wary of that...
04:08 dcook So you don't wind up truncating the message by accident..
04:08 dcook Even if the protocol I use will be super short and straight forward... I rather do something the right way
04:13 dcook Back to my actual problem...
04:15 dcook So the error message comes from the generic function IO::Socket::IP::_get_host_service... I knew that
04:15 dcook But the specific function which calls that generic function is IO::Socket::IP::peerhost
04:15 eythian check the network
04:15 dcook So there's an issue with the peer host address
04:15 dcook How do you mean?
04:15 eythian well, it's supposed to be doing something.
04:15 dcook Ah, that could still be our DNS, couldn't it..
04:16 eythian do that with unix tools and see what happens
04:16 eythian yeah
04:16 eythian or firewalling or something
04:17 dcook Hmm couldn't do a traceroute on it
04:17 dcook But can do other things like google
04:17 dcook Cannot handle "host" cmdline arg
04:17 dcook Name or service not known
04:17 eythian that seems suspicious
04:18 dcook When I googled the original message, people said vaguely that it had been down to DNS issues..
04:18 dcook If I go a wget, it works
04:19 dcook Well, actually, it doesn't..
04:19 dcook It's a 400
04:19 dcook But it's still doing the name lookup correctly
04:19 eythian from the server that's running it?
04:19 dcook Yeah, same server for both the wget and traceroute
04:19 eythian and that is running the code?
04:20 dcook yep
04:21 dcook Can't ping it..
04:21 eythian sounds like there's a firewall blocking ICMP.
04:21 eythian mtr will tell you where
04:23 dcook I'll look at mtr then :)
04:26 dcook mtr shows this:
04:26 dcook Failed to resolve host: Name or service not known
04:26 eythian right
04:26 eythian that's suspicious
04:28 dcook The hell..
04:28 wahanui hmmm... .. is that cause of the custom css, or is it that way for everyone
04:29 dcook I just keep trying and now mtr works
04:29 dcook Well... for a shortened version
04:29 eythian check your DNS servers and make sure they all work
04:29 dcook mtr
04:29 dcook that works
04:29 dcook mtr
04:29 dcook that doesn't
04:29 eythian it won't
04:29 eythian mtr is a prettier traceroute
04:29 eythian it doesn't take paths
04:30 dcook Ah, that would be logical
04:30 dcook silly dcook
04:30 dcook in that case, traceroute works too :p
04:30 eythian but failing DNS is a bad sign
04:31 dcook Well, that's what I'm confused about now
04:31 dcook When I pull my head out and don't use paths for mtr and traceroute... it works
04:31 eythian oh right
04:31 dcook But I'm still getting that IO::Socket::IP error
04:31 dcook Which... could mean that it's getting something bad in there
04:31 dcook I suppose?
04:31 eythian what's the error?
04:31 dcook addr is not a string at /usr/lib/perl5/5.20.1/IO/Socket/
04:31 wahanui i already had it that way, dcook.
04:32 dcook I'd like to know what addr is exactly
04:32 eythian hmm
04:32 dcook Oh right... it's this
04:32 dcook Well...I don't have the data
04:32 dcook $self->peername
04:36 dcook sub peerhost { my $self = shift; ( $self->_get_host_service( $self->peername, NI_NUMERICHOST, NIx_NOSERV ) )[0] }
04:36 dcook That's where my stacktrace points me
04:37 dcook I figure it must be using getpeername ([…]tpeername.2.html)
04:37 eythian probably
04:37 eythian though, addr shouldn't be a string there too.
04:37 eythian who knows
04:39 dcook Maybe I'm wrong about getpeername...
04:39 dcook _get_host_service... passes $self->peername to..
04:39 dcook getnameinfo() which is defined in Socket.xs
04:39 dcook Midway through it croaks as addr doesn't appear to be a string according to SvPOKp()
04:40 dcook
04:41 dcook Similarly, generic C code must call the SvGETMAGIC() macro to invoke any 'get' magic if they use an SV obtained from external sources in functions that don't handle magic.
04:41 dcook O_O
04:46 dcook I'm guessing $sock->peername is an object?
04:47 dcook And SvGETMAGIC(addr); is supposed to turn it into a string...
04:47 dcook I don't know. The code looks fine?
04:48 dcook Ahhh the peername method is from IO::Socket
04:48 dcook Good ol' inheritance...
04:49 dcook Which is interesting..
04:49 dcook ${*$sock}{'io_socket_peername'} ||= getpeername($sock); ([…]lib/IO/
04:58 rocio joined #koha
05:21 Francesca joined #koha
05:29 mario joined #koha
05:48 saiful joined #koha
05:50 cdickinson joined #koha
06:08 rocio left #koha
06:19 dcook eythian: I think I know why I was getting that error...
06:19 dcook Well, vague idea
06:19 dcook I think there was a problem with the initial connection
06:19 dcook So $sock isn't what getpeername expects it to be
06:20 dcook We changed the URL to a few different things and it works
06:20 dcook So I think there's something up with EBSCO's API
06:20 * dcook is surprised he hasn't heard anyone else complaining about EDS in that case though...
06:22 dcook I really should've realized sooner actually
06:22 dcook Because there is an earlier request to a different URL with the same hostname and getpeername is working for that
06:23 dcook It really looks like something isn't working right for[…]est/createsession
06:26 dcook Actually, I think the connection is successful, but then the server interrupts it or drops it or something...
06:26 dcook Other endpoints work
06:26 dcook And... it appears that EBSCO has indeed changed the API and there is a new plugin version out as of this afternoon
06:38 dcook Looks like a bad authtoken...
06:39 magnuse joined #koha
06:53 cait joined #koha
07:05 * magnuse waves
07:08 * cait waves back
07:09 magnuse cait: i'll have a look at bug 13642 as soon as i can find the time for it
07:09 huginn Bug[…]_bug.cgi?id=13642 enhancement, P5 - low, ---, hector.hecaxmmx, Needs Signoff , Adding new features for Dublin Core metadata
07:21 Francesca joined #koha
07:30 cait magnuse: someone asking where to find your work :)
07:30 cait left #koha
07:41 ashimema joined #koha
07:47 Traumatan joined #koha
08:03 Joubu hi
08:09 Viktor joined #koha
08:14 mveron Good morning everybody :-)
08:15 mveron Joubu: Thanks for taking care of Bug 14946, I signed off.
08:15 huginn Bug[…]_bug.cgi?id=14946 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Remove C4::Dates from files acqui/*.pl
08:16 Joubu thanks mveron
08:16 mveron Joubu++
08:23 magnuse mveron++ Joubu++
08:37 Sirenia joined #koha
08:58 cait joined #koha
09:01 saiful joined #koha
09:16 cdickinson joined #koha
09:26 cait left #koha
09:26 cait joined #koha
09:31 bumby Joubu: Hi, martin here, working on bug 15116
09:31 huginn Bug[…]_bug.cgi?id=15116 normal, P5 - low, ---, koha-bugs, Failed QA , show circ buttons at the top of batch circ page
09:31 Joubu hi bumby
09:31 bumby trying to figure out a good way to handle this "print slip" shortcut-link
09:31 bumby I can see your point in it beeing a bit redundant, and perhaps even confusing :P
09:32 bumby the idea was to make it so that you wouldn't have to navigate by mouse (which seems be the only reasonable way to reach the print-slip submenuitem)
09:33 bumby do you thank it might be a better solution to automaticly open the "Print"-menu and put focus on the "print slip" item?
09:34 Joubu no :)
09:34 bumby :D
09:34 Joubu It's certainly not a good idea to open something by default
09:34 Joubu Did you try to define shortcut for this link?
09:36 bumby didn't know there was a shortcut system in koha, but if so, that certianly seem like a good idea
09:36 cait hm maybe a cookie + checkbox? to open by default?
09:37 cait so first ime you click a link, later on it just opens it
09:37 bumby The single item checkout page will print a slip if you hit enter in the barcode textbox if it's empty
09:37 cait bumby: that's driven by a system preference
09:37 cait CircAutoPrintQuickSlip
09:40 bumby hm, perhaps there could be a CircAutoPrintQuickSlip that will automaticly print a slip after a checkout, without any user interaction needed. I'm a bit unsure of the usability of that though, don't know too well how the system is actually used in reality :P
09:45 cait iam not sure you can print in the background in a browser currently
09:46 cait so there will probably always be a pop-up
09:52 bumby true, I can see how that would be very annoying
09:53 bumby I'll go with Frédérics suggestion to limit my patch to what was actually requested, and perhaps open a new one for this feature.
09:58 cait sounds like a good idea
10:15 bdonnahue joined #koha
10:43 mveron Joubu, I signed off blocking bugs for Bug 14969
10:43 huginn Bug[…]_bug.cgi?id=14969 enhancement, P5 - low, ---, jonathan.druart, ASSIGNED , Remove C4::Dates from serials/*.pl files
10:47 drojf joined #koha
10:47 drojf hi #koha
10:55 khall joined #koha
11:16 mveron Lunchtime...
11:16 wahanui lunchtime is a terrible time for a meeting
11:23 Francesca joined #koha
11:33 andreashm joined #koha
11:35 * andreashm waves
11:38 drojf hi andreashm
11:46 andreashm hi drojf!
12:01 andreashm anyone working with a union catalogue here?
12:06 drojf andreashm: what do you mean by "working with"?
12:09 andreashm well, that have libraries that are connected to a union catalogue.
12:09 andreashm preferably one that uses bibliographic and holdings records.
12:09 andreashm I know BSZ does, but cait isn't around...
12:10 cait am back now :)
12:10 andreashm sweet!
12:13 andreashm potentially, if you where to use OAI-PMH to harvest records from the union catalogue into your local Koha instance (if you even have that workflow), and for server reasons there where separate flow for bib and holdings set up: would you prefer to first download and import bibs and then connect relevant holdings records or the other way around (holdings first, and the connect and download/import the bibs)?
12:13 andreashm heh, I don't know if I've ever written such an hypothetical question
12:14 cait um
12:14 cait i think first the bibliographic
12:14 drojf can you have holdings without bib records in koha?
12:14 cait in case it's a new record
12:14 cait but i might be missing something :)
12:14 andreashm The Swedish way of doing it is to look at the holdings first, and then connect the bibs... just because you do most changes in the holdings records. I'm wondering if we are alone in this workflow.
12:15 cait hm in our workflow
12:15 andreashm drojf: I don't think so - these will have to be merged. Which is a little difficult since Koha doesn't have holdings records.
12:15 cait we merge information from the holdings into the bilbiographic record
12:15 cait so actually we merge and import once
12:15 cait because the holdings include information about local URLs, classification, subjects...
12:16 cait on my slides this is what the "converter" does that we use withour nightly imports
12:16 andreashm yes, the import should preferably only be once.
12:16 cait maybe another difference is that we get always matching records
12:17 cait so we get a file set
12:17 drojf cait: is the workflow in the bsz wiki? that will probably be interesting for the seminar
12:17 andreashm even if going for holdings first, these would have to be staged somehow I guess to be marged when the bibs are downloaded... otherwise Koha would have difficulties. =)
12:17 cait with the bibliographic and holdings and authorities each night - that is specific to the library
12:17 cait drojf: it's on some of my slides i think... the wiki is probably not qute up to date
12:18 andreashm cait: we have a similar functionality... batch exports. but we would like it to be much, much faster... thus OAI-PMH.
12:18 cait andreashm: we have solved that outside of Koha basically
12:18 cait we create a marc file for koha .... and then use the staged tools from command line
12:18 cait for oai... we'd probably still have a conversion step that happens outside of koha
12:18 andreashm cait: that makes sense
12:21 cait not ideal... but hm
12:21 cait not sue how to solve it differently as long as the holdings are separate
12:21 andreashm cait: on your ealier question: new records should be included even if looking at holdings first, as long as the bib actually has any holdings? (and for us, if a bib doesn't have holdings we don't want it in Koha)
12:21 cait if they were included in the bilbiographic file... maybe you could have an xslt conversion step or something like that in koha deal with it
12:22 cait andreashm: sorry, i don't quite understand
12:22 andreashm yeah, it's a bit difficult.
12:23 andreashm you said that the reason for looking at bibs first is in case there is a new record. but a new record should have a holding as well?
12:24 * andreashm wishes Koha had holdings records, if only for this reason
12:25 magnuse there is no way to have itemtype icons show up in opac result lists, right? even if records have 942$c set?
12:26 magnuse ah, item-level_itypes. but that changes several things at once...
12:28 cait magnuse: i thought there is bug for that somewhere
12:30 cait hmmm
12:30 magnuse yeah, i think so too
12:30 cait andreashm: i am not sure... holdings are quite complicated
12:30 cait you could probably teach koha about them, but it would be complicated
12:31 magnuse bug 7492
12:31 huginn Bug[…]w_bug.cgi?id=7492 enhancement, P5 - low, ---, paul.poulain, NEW , Split item-level_itypes in two
12:31 cait or could turn out to be... the 852 does not exactly match our 952
12:53 tcohen joined #koha
12:53 magnuse kia ora tcohen!
12:54 tcohen hi!
12:54 cait hi tcohen
12:55 tcohen hi!
12:55 tcohen it is beta time!
12:55 magnuse w00t!
12:55 cait :)
12:57 nengard joined #koha
13:06 andreashm .
13:13 cabillman joined #koha
13:19 cabillman Hi everyone. I'm starting to prep for a 3.22 upgrade. I would like to split our opac + staff access onto two seperate servers. Are there any guides for that kind of setup?
13:20 Joubu tcohen: I have just attached a quick followup on 14544, please push :-/
13:20 Joubu (debug mesg in opac-search)
13:20 cabillman sepecifically zebra - i'm not sure if i need to run one instance and have it listen on the network. Or if I can run two instances one on each server
13:20 tcohen Joubu: was about to, as soon as i got notified by chrome .-D
13:27 andreashm cait: regarding automatic renewals - having that on, means that a user cannot manually renew an item? is that correct?
13:47 saiful joined #koha
13:50 tiuna joined #koha
13:52 tiuna Hello everyone!, I've a small question, I need to setup some rule for calculate a suspension based on the double o overdues days, someone of you knows?
13:58 mario joined #koha
13:59 tcohen tiuna: again?
14:01 tcohen you usually set the frecquency of the calculation to x and the suspension days to y, so for each x overdue days, it will add y suspension days.
14:01 tcohen For example: x=1 and y=2 will add 2 suspension days for each overdue day
14:01 tcohen so the double
14:02 tiuna where I setup X & Y?
14:04 magnuse joined #koha
14:19 huginn New commit(s) kohagit: Bug 15102: Fix capitalization for some strings on XSLT detail pages <[…]230039d839404e5c9> / Bug 14553: Silencing warn triggered when clearing item rating on OPAC <[…]607349adbf16344be> / Bug 14998: Restore previous behavior <
14:23 cait cabillman: hm actally i am not sure if waht you want to do is possible - separating opac and staff. separating mysql, zebra and koha I should be possible
14:24 cabillman cait: we already have separate mysql server
14:24 cabillman cait: we had a situation where a bunch of classes were teaching opac use and it was causing slow downs for the librarians running checkouts
14:25 cabillman my thought was to seperate them to minimize the impact clients can have on the staff functions
14:32 talljoy joined #koha
14:36 cait I haven't heard about separating staff and opac... I think in theory maybe you could have 2 kohas pointing to the same database... but that's more something you shoudl discuss with a system administrator (not me :) )
14:36 cabillman cait: thanks :)
14:36 cabillman cait: would i be better off posting this to the mailing list?
14:37 drojf cabillman: don't use the a production server to teach maybe?
14:37 drojf anyway i am interested in the question in general :)
14:38 cabillman drojf: the situation was k-6 students being tought how to use the opac to find books
14:38 cabillman i'm sure our librarians want them using the real opac...
14:38 cait hm also showing htem the user accounts i guess - it's a normal thing
14:39 cait maybe the queston is why it slowed down your system so much
14:39 drojf what cait says. i mean, what number of people using the opac at once do we talk about?
14:39 drojf i would not expect that to happen
14:40 drojf at least not on a server where you consider splitting staff and opac instead of upgrading your hardware ;)
14:40 cabillman to give a little bit of context... our librarians have like 30 students line up with two books each and try to check them out as fast as possible
14:40 cabillman we have 10 buildings
14:40 cabillman since they have kids lining up they are SUPER sensitive to slow downs
14:41 cabillman like a 3 second pause causes my phone to ring...
14:41 cabillman something must have gone wrong on one of the student machines because apache had like 200 requests to the opac in the span of 10 seconds
14:44 cabillman we have it running in our vmware cluster. I've already given it over 16 gigs of ram just for web/zebra box. mysql is on a different vm
14:47 cait maybe split of zebra indexes?
14:48 cabillman does anything get written to zebra? or is it just for seraching
15:01 cait it#s used for searching - the indexes are updated when a record or item record is changed
15:21 talljoy good morning #koha
15:23 * druthb waves to talljoy.
15:23 Traumatan joined #koha
15:41 talljoy hiya druthb
15:57 JesseM_ joined #koha
16:12 francharb joined #koha
16:18 JoshB joined #koha
16:26 wnickc joined #koha
16:35 mveron Jobu++ for Bug 14969
16:35 huginn Bug[…]_bug.cgi?id=14969 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Remove C4::Dates from serials/*.pl files
16:37 * mveron has to prepare a meeting
16:40 mveron joined #koha
16:50 bag heya
17:01 nengard_ joined #koha
17:17 Joubu tcohen: erk, back to the dark side, master is unstable
17:19 cait :(
17:19 cait cya all later
17:20 bag bye bye
17:21 Joubu tcohen: ha no, it's U14
17:22 Joubu Hi and bye bag :)
17:22 Joubu See you tomorrow #koha
17:22 cait left #koha
17:22 bag hi Joubu
17:22 bag bye bye Joubu
17:49 nengard__ joined #koha
17:49 JesseM2 joined #koha
17:52 nengard__ left #koha
17:52 nengard joined #koha
17:58 tcohen khall: are you trying to run kohadevbox inside a VM?
17:58 khall not sure what you mean. I'm trying to set it up on my macbook. I was able to set it up on my mac pro just fine last week!
17:58 khall any ideas?
17:58 wahanui any ideas are welcome :)
17:58 tcohen khall: yes, you are missing ansible!
17:59 khall thanks! that's not in the installation instructions afaict
17:59 khall I wonder how I managed to have it installed on my desktop ; )
18:00 bag heh hi there tcohen
18:00 khall for some reason I was thinking that was something vagrant would bootstrap install on the vm. thanks tcohen!
18:00 tcohen khall: that would actually make sense
18:00 tcohen for instance, mtompset didn't find a way to run it on a windows host
18:00 tcohen and installing ansible inside the VM would make things easier
18:02 wnickc joined #koha
18:07 tcohen khall: thanks for pointing the missing ansible info
18:07 tcohen i wasn't aware of the opmission
18:24 cait joined #koha
18:54 huginn New commit(s) kohagit: Bug 13799: (PM followup) Dependency updates <[…]1150c2a44acefd67b>
18:57 JesseM2 joined #koha
19:01 cdickinson joined #koha
19:03 pianohacker khall: rt 30345
19:08 * cait waves
19:09 JoshB joined #koha
19:13 * pianohacker waves at wrong channel
19:24 tcohen later #koha
19:25 bag khall[…]_bug.cgi?id=15176
19:25 huginn Bug 15176: enhancement, P5 - low, ---, koha-bugs, NEW , Add a sanity check when starting SIP2 (permissions)
19:26 cait pianohacker: we are the wrong channel???
19:26 pianohacker cait: to mention an internal ticket yes :)
19:27 bag nothing to see here (bag waves his hand like a ginny-jedi)
19:33 Viktor joined #koha
19:38 Traumatan joined #koha
20:09 JesseM2 joined #koha
20:27 * wizzyrea waves
20:27 wizzyrea @later tell wnickc thanks, I will give it a look :D
20:27 huginn wizzyrea: The operation succeeded.
20:33 Viktor joined #koha
20:37 * cait waves
20:43 bag hey wizzyrea
20:43 wahanui well, wizzyrea is having serious trouble with the last bits of words today.
20:46 wizzyrea forget wizzyrea
20:46 wahanui wizzyrea: I forgot wizzyrea
20:55 Francesca joined #koha
20:58 tcohen joined #koha
21:00 nengard joined #koha
21:07 mveron-away joined #koha
21:07 mveron-away Gooed evening / daytime #koha
21:07 mveron-away Good
21:07 * mveron-away should put his glasses...
21:12 mveron tcohen: Failing test on bug 14985 was due to missing dependency on bug 15166
21:12 huginn Bug[…]_bug.cgi?id=14985 enhancement, P5 - low, ---, veron, Failed QA , Remove C4::Dates from 6 files in folder C4/*.pm
21:12 huginn Bug[…]_bug.cgi?id=15166 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Make output_pref accept a string as parameter
21:12 mveron ...and hi everybody :-)
21:14 mveron @wunder Allschwil
21:14 huginn mveron: The current temperature in Basel, Switzerland is 3.0°C (10:00 PM CET on November 11, 2015). Conditions: Clear. Humidity: 93%. Dew Point: 2.0°C. Pressure: 30.33 in 1027 hPa (Steady).
21:15 cait hi mveron :)
21:15 mveron Hi cait :-)
21:15 bag @weather austin, tx
21:15 huginn bag: The current temperature in West End, Austin, Texas is 29.1°C (3:15 PM CST on November 11, 2015). Conditions: Mostly Cloudy. Humidity: 58%. Dew Point: 20.0°C. Pressure: 29.80 in 1009 hPa (Steady).
21:17 mveron Oh, that is a little bit warmer...
21:19 wnickc joined #koha
21:19 bag heh it’s too hot here mveron
21:19 bag and humid
21:20 bag this is where I came from yesterday
21:20 bag @weather 97215
21:20 huginn bag: The current temperature in Kittyland Love Center, E Tabor, Portland, Oregon is 9.1°C (1:19 PM PST on November 11, 2015). Conditions: Mostly Cloudy. Humidity: 94%. Dew Point: 8.0°C. Windchill: 9.0°C. Pressure: 30.35 in 1028 hPa (Falling).
21:20 bag HAHA even more humidity :P
21:20 wizzyrea kittyland love center?
21:20 wizzyrea what happens there?
21:20 bag funny right
21:20 mveron The best temperature is around 24°C... And dry...
21:21 bag yeah :)
21:30 tcohen mveron-away: thanks for pointing that out
21:30 tcohen any QA team member willing to test 15166?
21:32 tcohen jenkins_koha: health Koha_Master_D7
21:32 jenkins_koha Koha_Master_D7: Health [Clover Coverage: Conditionals 38% (6970/18426)(54%), Test Result: 0 tests failing out of a total of 17,216 tests.(100%), Build stability: No recent builds failed.(100%): http://jenkins.koha-community.[…]ha_Master_D7/584/
21:32 tcohen jenkins_koha: botsnack
21:32 jenkins_koha tcohen: you're so kind to me!
21:32 bag I can take a look tcohen
21:33 cait bug 15166
21:33 huginn Bug[…]_bug.cgi?id=15166 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Make output_pref accept a string as parameter
21:33 cait i am kind of on my way to that... but wouldn't mind anyone beating me to it :)
21:33 wizzyrea for your testing amusement,
21:34 tcohen wizzyrea: wildcard?
21:34 cait hm?
21:34 wizzyrea not sure, gmcharlt did the procurement of the certs
21:34 wizzyrea if it works, we're winning :)
21:34 wizzyrea I think so
21:35 * wizzyrea ought to go back and reread all the things he's said to me again, I've slept since I last did
21:35 tcohen gmcharlt++
21:36 tcohen wizzyrea: it shows green
21:36 tcohen so it works heh
21:36 wizzyrea fantastic :)
21:36 tcohen i would like to mention that here we are dropping nginx for reverse proxying
21:37 tcohen (we use a wildcard cert on the proxy pointing to the hosts that serve the services, for easier deployment of the certs)
21:37 tcohen but we found that at some high amount of requests the default nginx configurations didn't help
21:37 tcohen and they required re-compilation
21:37 tcohen to change them
21:38 tcohen unless you pay nginx pro or smth like that
21:39 tcohen works too
21:39 wizzyrea \o/ should do, that's what I told it to do. :D
21:40 cait wizzyrea: looks good - but firefox tells me it has blocked some unsafe things
21:40 wizzyrea oh right yeah I need to clear the cache
21:40 wizzyrea I've fixed that
21:40 cait images?
21:40 wahanui images are showing a bit ... hope that doesn't break things
21:40 wizzyrea there will be a bit of that
21:40 wizzyrea yeah
21:40 cait it's usually what happens to me
21:40 cait :)
21:41 wizzyrea that's what happens when you use a full path instead of a relative one
21:41 wizzyrea you have to fix all that kind of stuff.
21:41 wizzyrea I figured I'd get the homepage all golden, then make my way through
21:44 cait wizzyrea++
21:44 wizzyrea theoretically you shouldn't see any blocked things on the homepage now
21:45 wizzyrea (I cleared the cache)
21:45 wizzyrea (you might have to clear your cache too)
21:45 wizzyrea (I don't, if that's encouraging)
21:49 wizzyrea RewriteCond %{HTTP_REFERER} !^http://(.*)?koha-community\.org [NC] < would this mean that the rule only applies to http://
21:49 wizzyrea I am so dumb at rewrite rules.
21:50 cait i have never written one, you are way ahead :)
21:50 cait wizzyrea: hm still the little sign for me
21:50 cait but can't see what it exactly complains about
21:50 wizzyrea if you click the sign
21:50 wizzyrea and click "more information
21:51 wizzyrea then click "media"
21:51 wizzyrea it shows all good for me
21:51 wizzyrea but interested to know what yours says
21:51 cait yeah for me too
21:51 wizzyrea plus I'm never sure if the simple cache dump ever works
21:51 wizzyrea so might have to do a full one :)
21:51 cait thy are all from https://....
21:51 cait not obvious what it#s annoyed about
21:52 cait oh i have an idea
21:52 cait hn nope that didn't work
21:55 wizzyrea what page specifically are you looking at?
21:55 wizzyrea just the home page?
21:55 wizzyrea and with what browser
21:56 cait firefox
21:56 cait the main page
21:56 wizzyrea do you mean the half grey one or the triangle exclamation point one?
21:56 cait triangle
21:56 wizzyrea because mine says "content is blocked" but shows me the lock.
21:56 cait it does show the lock, but with a triangle
21:57 cait green lock, grey triangle
21:57 eythian hi
21:58 eythian dcook: so I heard there's an issue with the EDS API
21:58 cait hm a sign off ofr bug 15030 would be nice
21:58 huginn Bug[…]_bug.cgi?id=15030 blocker, P5 - low, ---, philippe.blouin, Needs Signoff , Certain values in serials' items are lost on next edit
21:58 cait item loss bug
22:00 wizzyrea eythian: did alvet stop by yesterday? he might today.
22:01 wizzyrea cait: I think I see something that might be doing it
22:01 wizzyrea 2 shakes
22:02 eythian wizzyrea: he didn't
22:09 bag why are my tests failing
22:09 bag silly tests
22:10 wizzyrea is it creators.t?
22:10 wizzyrea on a kohadevbox?
22:10 wizzyrea are you using koha-shell, and not setting the perl5lib?
22:10 wizzyrea to your git checkout
22:10 bag yes creators.t
22:11 JoshB joined #koha
22:11 wizzyrea don't worry about it
22:11 tcohen bag: you should run tests using
22:11 bag but I want to :P
22:11 wizzyrea mason did a patch for it
22:11 tcohen sudo koha-shell kohadev
22:11 tcohen cd kohaclone
22:11 tcohen prove t/Creators.t
22:12 wizzyrea do you have your koha-shell set to point to your git checkout tcohen?
22:12 tcohen nope
22:12 wizzyrea I almost always have this problem on a gitified install.
22:12 tcohen i just cd kohaclone once in
22:12 tcohen wizzyrea: we fix it in kohadevbox:ansible
22:12 tcohen by dropping a .bashrc file inside kohadev user
22:13 tcohen home
22:13 wizzyrea[…]_bug.cgi?id=14602
22:13 huginn Bug 14602: minor, P5 - low, ---, mtj, Pushed to Master , Fix failing t/Creators.t test, when using koha-shell
22:13 wizzyrea this must be a difference between ubuntu and debian, since it's uneven in appearance
22:14 tcohen wizzyrea: i do it in jessie
22:14 wizzyrea oh yay that one's been pushed.
22:14 wizzyrea that explains why I don't see it on master anymore ^.^
22:14 bag yeah I need to upgrade to jessie soon
22:14 tcohen
22:15 wizzyrea dunno, I get that one all the time
22:15 tcohen wizzyrea: i switch between wheezy, trusty and jessie each time i start my day
22:15 wizzyrea anyway, it happens sometimes, and I don't much worry about it.
22:15 wizzyrea because I know why it' shappening and others don't seem to have the problem.
22:18 tcohen wizzyrea: this is what we put in the instance's home directory[…]nstance_bashrc.j2
22:18 wizzyrea sweet
22:18 tcohen that way all tests (including those that go through apache) can be run inside the koha-shell
22:19 tcohen next step will be that once bag pushes Joubu's selenium scripts, we add the dependencies to kohadevbox too
22:19 bag :)
22:28 cait Bug 15030 - Certain values in serials' items are lost on next edit - anyone? :)
22:28 huginn Bug[…]_bug.cgi?id=15030 blocker, P5 - low, ---, philippe.blouin, Needs Signoff , Certain values in serials' items are lost on next edit
22:30 wnickc cait: I couldn't recreate, I thought was just me, but maybe more details on process?
22:31 cait hm let me see
22:32 cait where did you edit?
22:32 cait in cataloguing or from the serial collection page?
22:33 cait i think that it happens when you reedit a serial issue item from within the serials module
22:34 wnickc I don't remember right now, but I can try again when back in office next week
22:35 wnickc I thought I tried both
22:36 wnickc and hi cait ;-)
22:39 Francesca joined #koha
22:40 cait hi wnickc :)
22:40 cait wnickc++
22:40 bag man I pay attention to kyle and pianohacker talking and cait steals my qa bug from under me :P
22:40 bag cait++
22:40 cait hehe
22:40 bag heh
22:40 cait there are enough left!
22:40 bag yessem
22:40 cait anthing dates would be good and all bugs
22:40 bag we are showing wnickc and barton git aliases and qa-tools now
22:41 bag will be back in a bit
22:41 bag :)
22:41 cait i am on bug 14969 now
22:41 huginn Bug[…]_bug.cgi?id=14969 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Remove C4::Dates from serials/*.pl files
22:42 cait if you have soome good git aliases... put them on the wiki :)
22:43 wnickc wiki++
22:43 * wnickc is always surprised at how much is there
23:07 cait wnickc, wizzyrea:
23:07 cait the text in the images tab - does it work?
23:07 wizzyrea seems ok to me
23:07 cait it was before:
23:07 cait please <a>upload</a> one.
23:08 wizzyrea yes, much better then
23:08 cait oki
23:08 cait better for translations too :)
23:08 wizzyrea You could even change the button to say "Upload file" and remove the please select... text
23:09 wizzyrea which I think is even better for translation?
23:09 wizzyrea fewer words anyway.
23:09 wizzyrea and not, I think, less usable.
23:11 wizzyrea we have a lot of places where we have too many words.
23:14 bdonnahue joined #koha
23:50 cait wizzyrea: argh - missed your comments
23:50 cait i passed qa on it after you saidit was ok :)
23:51 cait would you mind adding your idea to the bug? it's bug 15154
23:51 huginn Bug[…]_bug.cgi?id=15154 minor, P5 - low, ---, koha-bugs, Passed QA , Allow correct translation for upload local cover image
23:52 cait @later tell drojf can you take a look at bug 15058? updates to the german web installer
23:52 huginn cait: The operation succeeded.
23:58 cait night
23:58 cait left #koha
23:59 nengard left #koha

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