Time  Nick         Message
23:58 rangi        awesome
23:57 jcamins      rangi: http://git.koha-community.org/gitweb/?p=contrib/global.git;a=summary
23:57 mtompset     Ah... RaiseError in the connect statement.
23:55 eythian      or DBI rather
23:55 eythian      the DBD docs tell you about that I think
23:54 mtompset     DBD::mysql::st execute failed: Table 'kohadata.TRYMEOUT2' doesn't exist at ./test2.pl line 21.
23:53 mtompset     I want to handle them myself.
23:52 mtompset     I know there will be messages like "db doesn't exist' or "db already exists".
23:52 mtompset     I don't want output.
23:52 mtompset     perhaps.
23:51 rangi        you arent confusing warns with errors again?
23:51 mtompset     I'm trying that, but it isn't working.
23:50 rangi        http://search.cpan.org/~doy/Try-Tiny-0.12/lib/Try/Tiny.pm
23:49 mtompset     Is there a nice way to silence errors and determine there was one?
23:29 pianohacker  bye
22:04 mtompset     Now I'm on my desktop computer... let's get coding. :)
22:04 mtompset     Greetings, #koha.
21:32 Brooke       I gots to go though
21:32 Brooke       hope it was fun
21:32 Brooke       oh wow
21:28 cait         for started on monday, came back tonight
21:28 cait         travelling for work
21:28 cait         lol no
21:26 Brooke       unless you're in the states, in which case WHERE!?
21:26 Brooke       wow, you were out late
21:20 cait         just got home, cookies welcome
21:20 cait         hi Brooke and thx
21:20 * Brooke     hands cait an Akoma Chocolate Chip Cookie
21:19 Brooke       its' Cait!
21:18 Brooke       how's it shakin on the left siiiiiide?
21:12 bag          heya Brooke
20:49 Brooke       o/
19:53 mtompset     Have a great day (24 hour period), everyone.
19:53 * druthb     blinks…  now I've heard everything.  Scott Joplin's "Pine Apple Rag"…on handbells.
19:47 mtompset     Because. :P
19:46 libsysguy    bah, traffic is already bad....wwwwwhhhyyyyyy
19:46 druthb       Right.
19:46 libsysguy    :p
19:46 libsysguy    oh thats right, just washington DC
19:45 * druthb     never said that!
19:45 rangi        the queryparser that jcamins did is a zillion times better
19:45 mtompset     oleonard: Good Lord! Say you didn't do that. Say you didn't enter the labyrinth.
19:43 libsysguy    what is it druthb says, Search.pm is so dense not even good search results can escape
19:43 oleonard     Quick trip into Search.pm right before quitting time? Never a good idea.
19:42 libsysguy    I was trying to get some elbow action in there with trex arms
19:42 libsysguy    oh yes, I did mean that
19:42 mtompset     libsysguy: What is o[? Don't you mean o<? :P
19:41 mtompset     oleonard: Yes, yes it is. God be with you. :)
19:39 libsysguy    Random: \o/   ^o^   o[    /o\
19:38 oleonard     I have 9 minutes to discover how hard that would be before leaving for a week of vacation.
19:37 oleonard     Hm, it would certainly simplify things if "SEARCH_RESULT.norequests" were true for available items if AllowOnShelfHolds is off...
19:34 oleonard     Uh... "if" no options are enabled...
19:34 oleonard     ...but I don't want to show an empty "actions" label no options are enabled
19:33 oleonard     Urgh, the logic with which to tell if an "action" link will appear in OPAC search results is really convoluted...
19:30 mtompset     jcamins: !important?! I encountered that issue with Koha. :P
19:18 cait         heh
19:17 pianohacker  #koha is all kinds of good drugs today
19:16 drojf        that's what you get from smoking roast beef hash
19:15 * druthb     dances a lil mambo.
19:13 cait         hm it sounds weird reading it
19:13 cait         oh right thx
19:13 jcamins      Mambo.
19:12 cait         hm.
19:12 cait         m..
19:12 cait         does someoen remember what joomla was called before it was joomla?
19:11 jcamins      If you look up "legacy" in an online dictionary, it says "see Joomla."
19:11 jcamins      oleonard: that's why I said there was a problem with legacy code.
19:10 oleonard     Well, .... Joomla. :P
19:10 jcamins      I think the site I was shaking my head over that had that problem used Joomla.
19:09 jcamins      Where an ID selector from way up overrides the class selector, and the developer worked around it by adding !important instead of fixing the underlying problem, which is that div#content is a rather vague selector, really.
19:09 oleonard     Yeah, I wonder if the specificity conflicts are a problem with different kinds of projects? I dunno.
19:08 pianohacker  Because while I'm no battle-hardened web expert, the lack of reusability, js/css coupling and specificity fighting have never been huge issues
19:08 pianohacker  jcamins: how so?
19:08 jcamins      pianohacker: I've seen the problem he describes in legacy code.
19:07 pianohacker  I... feel like he's inventing problems
19:06 pianohacker  ...
19:04 oleonard     http://csswizardry.com/2013/05/hashed-classes-in-css/
19:04 oleonard     http://csswizardry.com/2011/09/when-using-ids-can-be-a-pain-in-the-class/
19:01 * oleonard   hates remembering reading a link from Twitter and not being able to find it again
19:01 oleonard     Not the one I was thinking of, but: http://oli.jp/2011/ids/
18:58 * cait       is not amused
18:58 cait         the train has to go slower because of a weather warning :(
18:58 cait         ook
18:58 cait         aah
18:58 * oleonard   is trying to find the argument he read
18:58 oleonard     cait: Those who argue for that method are not saying you shouldn't use ids in HTML and with JavaScript, just not as a hook for CSS
18:55 cait         and if you do that.. no validator is going to tell you that you have used your intentionally unique id on more than one element on the page
18:55 cait         why would ids be bad?
18:55 cait         it doesn't make sense to me
18:55 * oleonard   won't be using 'class="#something"' any time soon
18:55 oleonard     I'm not convinced, so I'm not telling anyone to do it any particular way.
18:54 mtompset     However, I'm no styling expert, so I would rather trust oleonard's opinion. :)
18:53 mtompset     But creating that many classes seems overkill. Classes make for nice groupings of ids. Hence my granularity thought.
18:52 jcamins      mtompset: so do classes if they are unique.
18:52 druthb       Good for breakfast, on toast, with a fried egg on top.
18:52 mtompset     I could be wrong, but don't id's give you a kind of granularity to the styling?
18:51 druthb       Roast beef hash is the best kind...
18:51 oleonard     1. They think using ids for styling is bad. 2. They think 'class="#something"' conveys the same sense of importance others use an id to express.
18:50 jcamins      oleonard: really? Why?
18:50 jcamins      mtompset: because you've been writing hundreds of lines of Javascript.
18:50 oleonard     Many are now recommending that people use 'class="#something"' and not use ids for styling at all.
18:50 * tcohen     goes check his latest patch *just in case*
18:49 mtompset     YUCK! Why would you put a hash in front of your id?!
18:49 pianohacker  I'd hate to think you were doing it intentionally :)
18:49 jcamins      pianohacker: only by accident.
18:49 pianohacker  That's a new one... does anyone else also occasionally write id="#something" in their HTML?
18:48 pianohacker  hurr, id != class
18:29 tcohen       i fixed some koha-qa errors too oleonard
18:28 tcohen       i resent the patch
18:22 cait         understandable
18:22 oleonard     Ugh, that page always sends me running away in fear.
18:21 cait         ie version i mean
18:21 cait         but as i am not sure which verion that was, i will check that first
18:21 cait         that was particuluarly annoying
18:21 cait         one thing was that you couldn't order multiple items when ordering, because the item form didn't come back after clicking add for the first time
18:20 cait         but not the drag and drop
18:20 cait         yeah
18:20 oleonard     cait: JavaScript-related?
18:20 cait         the problems were in acq and serials mostly
18:20 cait         oleonard: ok :)
18:19 tcohen       hmm, maybe some cache is hiding it from me
18:18 oleonard     tcohen: The error does not happen in master
18:17 oleonard     I probably wouldn't bother if it was only in IE7 or below
18:17 oleonard     We don't technically support IE at all in the intranet, but I do my best to fix stuff I can.
18:16 cait         if i can figure out the version of ie - how far back do we care?
18:16 cait         should we file those? i forgot to get the ie version number in the training room
18:15 cait         oleonard: during the trianing we noticed a couple of things in intranet not working with ie
18:14 tcohen       not sure is related to the patch itself
18:14 tcohen       it appears on current master i guess
18:12 oleonard     tcohen: It appears in the Firebug console after you have selected one of the autocompleted items and the page redirects
18:11 tcohen       where did y find the error oleonard ?
18:11 jcamins      mtompset: only things that require cardnumbers.
18:09 tcohen       oh
18:09 tcohen       i'll check it oleonard, thanks for the feedback
18:09 oleonard     tcohen: I think you need to wrap that JS in [% UNLESS ( borrowernumber ) %]
18:09 mtompset     Does that cause things to break in Koha elsewhere, though?
18:09 oleonard     tcohen: Your holds autocomplete works well, but once you've selected a patron and the page redirects there is a javascript error
18:08 jcamins      Or you could even not have a cardnumber, yes.
18:08 mtompset     Well, actually, do we need a cardnumber is a question that has been on my mind too.
18:07 mtompset     yes, very similar to what gmcharlt suggested. His link got me thinking that way.
18:06 druthb       hi, drojf! :)
18:06 jcamins      *from the
18:06 drojf        and hi druthb
18:06 cait         hi drojf
18:06 drojf        hi cait
18:05 druthb       hi, mtompset.
18:05 jcamins      mtompset: that sounds like what gmcharlt suggested, but are you sure you wouldn't be better off using a non-autogenerated cardnumber as a way to actually move forward with your development?
18:05 cait         hm wonder if ids does it like we do
18:05 jcamins      drojf: fromt he wiki page.
18:05 mtompset     Greetings, druthb cait. :)
18:05 drojf        jcamins: did you get the DRUCK_ORT question from the new wiki page about old prints (in german) or is that a strange coincidence?
18:05 mtompset     virtual duct tape isn't as good as the real thing. :P
18:04 cait         we are going through all those tunnels soon
18:04 cait         not sure that will work
18:04 mtompset     values don't duplicate because everything is serialized, and the processes won't deadlock, because even if two processes calculate a value for autoincrement value X, it should be the same value.
18:03 mtompset     Creating a table to serialize (autoincrement, ID, value)... the process that needs a value generates a random string for an ID and does an insert into the table, but leaves value blank. It then calculates all the missing values, and grabs the last one for its own ID.
18:03 * druthb     gets the duct tape, and tapes cait to her chair
18:01 mtompset     And it won't deadlock and should work.
18:01 mtompset     Ooooo.... I think I have an idea to deal with my critical section problem.
17:59 mtompset     That be true.
17:58 drojf        up to 38°C in the coming days. time to switch from "it's way too cold for summer" to "omg it's way too hot"
17:57 drojf        hi mtompset
17:57 mtompset     Greetings, drojf jcamins.
17:56 jcamins      drojf: yeah, I had the same problem.
17:55 drojf        fun fact: localstorage does not work if you have your cookie settings set to "ask every time". probably not magnuses problem, but i spent quite some time figuring that out
17:40 * magnuse    has to get off the train now, see y'all later!
17:39 magnuse      jcamins: nope, same thing with chromium
17:37 mtompset     Greetings, tcohen.
17:37 tcohen       hi mtompset
17:37 magnuse      hiya mtompset
17:36 mtompset     Greetings, magnuse. :)
17:36 mtompset     Greetings, #koha.
17:36 magnuse      like the offline html5 stuff is not working at all
17:36 magnuse      no, i turn off apache and then when i try to access offline.pl i just get server not found
17:34 jcamins      If you are I can try to add some handling for the error, at least.
17:34 jcamins      Are you getting any sort of error?
17:34 magnuse      ah, i'll try that then
17:34 magnuse      jcamins: firefox
17:33 jcamins      It looks like possibly it doesn't work under Firefox, but might under Chrome.
17:33 huginn       04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10464 minor, P5 - low, ---, oleonard, Needs Signoff , Patron search on placing a hold doesn't use jqueryui.autocomplete
17:33 tcohen       bug 10464
17:33 jcamins      Which browser are you using?
17:32 * magnuse    is on a train
17:32 jcamins      Yeah, I read something about that as well, but I don't remember exactly where.
17:31 magnuse      i have vague feeling i have read something about local storage not working for localhost, but i'm not sure
17:31 jcamins      I suspect it would not work on localhost.
17:31 jcamins      that is a good question.
17:31 jcamins      ...
17:31 magnuse      or does it have to be over the network to work?
17:31 magnuse      do you know if it is supposed to work when koha runs on localhost?
17:30 jcamins      Yay!
17:30 huginn       04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10240 new feature, P5 - low, ---, jcamins, Needs Signoff , Offline circulation using HTML5 and IndexedDB
17:30 magnuse      i'm looking at bug 10240
17:30 magnuse      good evening, sir
17:30 magnuse      ah
17:30 jcamins      I am.
17:27 magnuse      jcamins around?
17:14 cait         hmmmmm i can imagine that tastes yum? )
17:13 tcohen       i don't have cookies for you, but how about patron autocomplete for placing holds?
17:12 cait         hi tcohen
17:06 tcohen       hi cait
17:05 * cait       waves again
16:57 oleonard     I re-learn jqueryui autocomplete every time I do it again.
16:57 oleonard     I don't remember offhand tcohen, you'd have to give it a try.
16:56 tcohen       i'm not used to the autocomplete jquery plugin, just asking for learning purposes
16:56 tcohen       oleonard: do we really need to provide a render function for the autocomplete? isn't the fact that ysearch.pl retrieves a JSON enough for rendering?
16:40 gmcharlt     heh; yeah, I'll be using more than just grep to evaluate those
16:34 jcamins      oleonard: I've seen a few of those.
16:33 oleonard     "TODO: Make this not broken."
16:27 * gmcharlt   will push patches with TODOs, but is not inclined to push patches that introduce new FIXMEs unless absolutely necessary
16:25 * gmcharlt   hands oleonard a paper bag to breath into
16:24 oleonard     So. Many. FIXMEs.
16:22 oleonard     gmcharlt++ # No FIXMEs
15:29 reiveune     bye
15:27 cait         :)
15:21 * druthb     wriggles happily
15:21 libsysguy    o/
15:20 * cait       waves
15:13 jcamins      I'll just generate IDs.
15:12 jcamins      Text <input>s.
15:12 oleonard     Which controls?
15:09 jcamins      Ah... that messes up the layout.
15:02 jcamins      oleonard: is wrapping controls in the label tag something that we are actively trying to avoid in Koha, or something that we just happen to not do?
14:59 wahanui      de nada jcamins
14:59 jcamins      wahanui: why thank you.
14:59 pianohacker  good morning
14:59 wahanui      I like *your* figure
14:59 jcamins      Figures.
14:58 oleonard     Yes
14:58 jcamins      oleonard: does the for attribute always have to refer to an id?
14:44 gaetan_B     bye !
14:42 oleonard     That was a tough patch to get in.
14:41 huginn       New commit(s) kohagit: Bug 8662 - Remove unused famfamfam icons <http://git.koha-community.org/gitweb/?p=koha.git;a=commitdiff;h=32960372225163b76fd214813b9a4e10f39b162f>
14:33 libsysguy    forms in general have their own special set of rules
14:32 oleonard     Yeah for an entry form I don't think it's appropriate. Better for display.
14:32 jcamins      I'm not 100% sold on using a collapse control, though.
14:32 oleonard     We have the jQueryUI version already included in the staff client.
14:31 jcamins      libsysguy: yeah, I considered that.
14:31 jcamins      I just tried it on the right, and it doesn't really work.
14:31 jcamins      I think I prefer the button on the left.
14:31 libsysguy    http://twitter.github.io/bootstrap/javascript.html#collapse
14:31 jcamins      True.
14:30 oleonard     On a long form like this I don't think vertical space is such a big deal. No way to avoid scrolling no matter what.
14:30 libsysguy    less_vert_space++
14:29 oleonard     I think I could go either way on the delete button but think the add button should stay left-aligned.
14:29 libsysguy    jcamins opinions have noted and marked as incorrect :p
14:29 oleonard     :)
14:29 jcamins      Unless, of course, you decide you agree with him, in which case I think it's a great idea.
14:29 jcamins      I disagree with libsysguy about moving the Add contact button.
14:28 jcamins      oleonard: I was thinking of putting the button in the top right so that it used less vertical space.
14:28 oleonard     Why put buttons on the right?
14:27 oleonard     Also, use 'icon-remove' for delete
14:26 oleonard     Within each individual fieldset is the delete button
14:25 jcamins      I was thinking of a pull-right on the Delete contact, and possibly putting it up at the top of the fieldset as well.
14:25 oleonard     My instinct is to say that the contact details for each contact should be in their own fieldset inside a grouping fieldset
14:25 libsysguy    imo
14:25 libsysguy    add a pull-right class to add contact button
14:24 oleonard     Hmm...
14:22 jcamins      oleonard: any suggestions on making these buttons fit in to the page and general Koha styling better? http://screencast.com/t/lpiLrlD6x1
14:22 khall        you'd think I'd demonstrated a super power or something
14:21 libsysguy    oh my
14:21 khall        lol
14:21 khall        it reminds me of when I showed a friend you could open a browser link with a middle click, he literally freaked out!
14:21 * jcamins    should've chosen zjcamins as a nick.
14:21 jcamins      Alas, I am the only one with jc, but one of many with j.
14:20 khall        yeah, I just discovered that from your previous comment. Never knew about it!
14:20 jcamins      khall: actually, you're the only one with k<tab>
14:19 khall        ; )
14:19 khall        but is *two* characters shorter! That's like a 40% increase in efficiency, right?
14:18 jcamins      khall: why? You're the only one with kh<tab>
14:18 jcamins      libsysguy: the problem is that you're not the only one with li<tab>.
14:17 * khall      should switch to kmh
14:17 khall        it's more efficient!
14:17 druthb       "li<tab>"…  how hard can it be?
14:16 nengard      heh
14:16 libsysguy    first nengard didn't want to type library_systems_guy so I shortened it to libsysguy, then khall wanted it to be lsg :p
14:15 druthb       me!
14:15 libsysguy    anybody remember when I was library_systems_guy
14:15 druthb       jcamins: I may shift to a different space shortly—the learning room, maybe.  The meeting is kind of immovable, since there's a bunch of folks in on it.
14:15 libsysguy    heh dean_and_draper
14:15 libsysguy    if we can hear it on 5 I know they are getting it hard
14:15 libsysguy    druthb David said D&D were mega complaining
14:14 druthb       Yeah, the Dean and Draper folks are probably not happy right now.
14:14 druthb       libsysguy:  no, I wasn't in on that.
14:13 jcamins      You cPanelians have the weirdest IRC nicks. "dean_and_draper"?
14:12 libsysguy    I bet Dean and Draper is pissed
14:12 libsysguy    David was talking about it last night. Were you in on that convo?
14:12 jcamins      Or reschedule the meeting?
14:12 jcamins      druthb: ask if you can use a different office?
14:11 druthb       If I didn't have a 12:30 meeting, I'd be out of here.  It's freaking *intense* down here.
14:11 libsysguy    :p
14:11 libsysguy    email david
14:10 druthb       They are *right above* my office.
14:10 druthb       duuuude.
14:10 libsysguy    but we are fume free
14:10 libsysguy    we can hear some banging if we turn down the heavy metal
14:10 druthb       hey, libsysguy—are the banging and paint fumes intense for you guys up on 5?
14:09 * druthb     thanks Galen that it's Thursday, though.
14:09 libsysguy    premature friday alert
14:09 druthb       Indeed.
14:09 gmcharlt     too soon!
14:08 huginn       druthb: Quote #98: "drulm: TGIF=Thank Galen It's Friday" (added by wizzyrea at 07:15 PM, October 08, 2010)
14:08 druthb       @quote random
14:02 huginn       gmcharlt: Quote #146: "rangi: when I was at brendans house we went to the farmers market and bought lots of almonds which I had to eat on the plane, cos can't bringthem into nz, so I was offering ppl nuts on the plane. cait: makes sense rangi: if its sensible for a crazy nzer to be waving almonds in ur face at 3am, then yes" (added by wizzyrea at 08:03 PM, July 28, 2011)
14:02 gmcharlt     @quote random
13:57 druthb       :P
13:47 oleonard     Heh... Lappy. http://www.homestarrunner.com/sbemail119.html
13:41 druthb       If I did not have a 12:30 meeting, I'd be grabbing the lappy and working from home.
13:40 jcamins      druthb: yay! Day off!
13:39 druthb       well, this is…annoying.  The construction on the third floor is apparently *right above my office*.
13:35 huginn       New commit(s) needsignoff: [Bug 6536] Z3950 Enhancements: SRU targets, MARC conversion, additional XSLT processing (umbrella report) <http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=6536>
13:32 marcelr      have to go; bye all
13:31 tcohen       you are right
13:31 oleonard     It's fine with me, but I would tie it to the circautocomplete system preference
13:30 tcohen       exactly
13:30 oleonard     Patron autocomplete?
13:30 tcohen       ?
13:30 tcohen       what do u think of adding the autocomplete feature on placing a hold
13:30 oleonard     Yes
13:30 wahanui      i heard oleonard was still here, if you just wish hard enough. or Koha's master UI designer
13:30 tcohen       oleonard?
13:26 tcohen       hi druthb
13:21 druthb       o/
13:02 tcohen       a
13:02 tcohen       morning #koh
12:46 oleonard     The commenters on that Facebook post sounded like a lynch mob.
12:44 nengard      yikes
12:38 oleonard     The sheriff said on Facebook they were surprised they didn't find a charred body on the scene, given the kind of equipment they were messing with.
12:37 jcamins      oleonard: really? You read about that, but I didn't realize people actually did that.
12:36 oleonard     Took 9 hours to repair, for some reason.
12:36 oleonard     Someone stole copper wire from the substation
12:28 Oak          pakistan?
12:27 nengard      storms?
12:24 * oleonard   hopes to be around more than a half hour today, electricity willing.
12:20 oleonard     Hi #koha
12:04 samueld      hi everybody :-)
09:38 marcelr      well could be worth a try now..
09:38 Joubu        marcelr: Yes I think we should have a Logger.pm module. log4perl is maybe a better solution but I never used it.
09:36 marcelr      i would leave it in a module
09:36 marcelr      if log4perl is syslog compliant, that is a plus
09:36 marcelr      could that not resolve the discussion?
09:35 Joubu        marcelr: yes, maybe
09:32 marcelr      Joubu: wouldn't it be easy to just replace LogLite with Log4Perl ?
09:31 marcelr      Joubu around?
08:51 marcelr      hi gaetan_B and rangi
08:40 rangi        hi gaetan_B
08:37 gaetan_B     hello
07:21 dcook        good night #koha
07:20 dcook        Dumplings ^_^
07:20 dcook        That said, I actually have to run in a moment...
07:19 dcook        I'm keen to chat to someone who can explain why that code is necessary :S
07:19 dcook        Do you use UNIMARC at all?
07:19 dcook        Besides wrapping it (which is already done), I suppose it would be an idea to add some checks for the variables...but I wonder why the code is necessary at all
07:18 dcook        Pretty much :/
07:18 marcelr      looks like you are abandoning it already
07:18 dcook        I'm not really attached to the component
07:18 dcook        Feel free to change it if you like
07:18 dcook        true that :)
07:18 marcelr      choosing the component is sometimes very arbitrary
07:16 dcook        But it looks like the unimarc-specific code is wrapped with the marcflavour preference in master (I was looking in 3.8. I thought I checked master, but I must have had my git branch wrong.)
07:15 dcook        Which should only be there if one is importing via z39.50
07:15 dcook        Yeah, that sub is only called if there is a breedingid
07:14 dcook        It's been a crazy day
07:14 * dcook      checks
07:14 dcook        I think...
07:14 dcook        Except that it's irrelevant without z3950
07:13 dcook        marcelr: Sort of
07:13 huginn       04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10456 normal, P5 - low, ---, gmcharlt, NEW , z3950NormalizeAuthor and z3950AuthorAuthFields cause problems for MARC21 (and are sloppily written)
07:13 marcelr      dcook: your bug 10456 is more about cataloguing, i suppose
07:12 marcelr      mtompset left quite fast
07:12 wizzyrea     whaddup homies
07:06 mtompset     Good day, everyone.
07:03 marcelr      hi dcook
07:03 dcook        hey ya marcelr
07:03 marcelr      hi #koha
06:55 huginn       04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10457 minor, P5 - low, ---, dcook, Needs Signoff , Export selected items data doesn't work when using checkall checkbox
06:55 dcook        If anyone wants an easy sign off... bug 10457
06:53 dcook        ta bag
06:53 dcook        I'm about to head home and I'm on the other side of the world...
06:53 dcook        About time, bag!
06:52 bag          ok paul_p is in - that means it's way to late to be at the office - night everyone I'm heading home
06:50 paul_p       hi bag !
06:50 bag          heya paul_p
06:50 paul_p       hello everyone !
06:45 drojf        hi alex_a and dcook
06:45 dcook        Guten morgen, drojf :)
06:44 alex_a       good morning drojf
06:44 christophe_c hi alex_a
06:44 drojf        good morning #koha
06:44 alex_a       et salut christophe_c
06:44 alex_a       salut dcook
06:43 christophe_c hello dcook ;-)
06:43 dcook        salut alex_a, christophe_c
06:43 christophe_c hello #koha
06:42 dcook        While I have confidence in the current volunteers, it would be tough to make sure that current and future volunteers are all able to maintain and upgrade Koha
06:41 dcook        cjh: Alas, I've suggested that Koha might not be the right solution for them
06:41 dcook        hey ya paul_p, reiveune
06:36 wahanui      hi, reiveune
06:36 reiveune     hello
06:24 cjh          awesome
06:21 dcook        Eeee...a student from my alma mater wants some help setting Koha up at a volunteer library :)
06:03 dcook        And by ever...I mean in my short experience recently
06:03 dcook        I think this training session has yielded the most bugs ever
05:44 dcook        But fixing bugs...which is good...
05:44 dcook        All the bugs..
05:30 cait         gmcharlt++ for being an awesome rm :)
05:28 dcook        Oh so the day of bugs...
05:22 dcook        Good call on the renaming of system preferences...
05:22 dcook        gmcharlt++
05:13 eythian      (I assumed there was a public one I hadn't heard of :)
05:13 dcook        No one tells me nothin' :p
05:13 wizzyrea     i might have heard of that
05:13 eythian      ah right :)
05:13 wizzyrea     :)
05:13 wizzyrea     dcook was
05:12 eythian      wizzyrea: you're thinking of the schools one perhaps?
05:12 dcook        That seems more plausible
05:12 dcook        Ah, that's a lot slower than I originally was led to believe
05:11 wizzyrea     well over like 3 years
05:11 * wizzyrea   too
05:11 dcook        I'm actually really interested to hear how that goes...
05:10 dcook        Well, in NSW, the public library system wanted to put 2,000+ libraries up rather quickly
05:10 wizzyrea     from not-koha to koha
05:10 dcook        O_O
05:10 wizzyrea     a 42 library consortium in like 3 days
05:10 wizzyrea     yea, that was pretty epic
05:09 dcook        But yay ByWater!
05:09 dcook        It sounds...sad
05:09 dcook        Yep, lol
05:09 wizzyrea     :P
05:09 wizzyrea     which means we really need to know the problem that he's trying to solve
05:08 wizzyrea     that probably means their ztarget changed
05:08 wizzyrea     so
05:08 wizzyrea     big country
05:08 wizzyrea     they were absorbed
05:08 wizzyrea     yep
05:08 dcook        "In 2013 the Big Country Library System and the Big Country Library System combined to form Texas Library Consortium Catalog or TLCcat. The consortium operates through a partnership between the Texas State Library and Archives Commission and e-Libris Technologies."
05:08 dcook        Is that what you were linking to?
05:08 dcook        http://www.librarytechnology.org/lwc-displaylibrary.pl?RC=12102
05:07 wizzyrea     ^ this just happened
05:07 wizzyrea     http://bywatersolutions.com/2013/04/05/tlc-koha-support/
05:07 dcook        Yeah, I noticed the click through took you to koha.org, but that Bywater seems to support some of the former libraries?
05:06 wizzyrea     that, peeps, is a library consortium - TLC
05:06 wizzyrea     lookit that list of libraries
05:05 wizzyrea     http://anspl.tlccat.bywatersolutions.com/cgi-bin/koha/opac-search.pl?q=fish&limit=branch%3AANSPL
05:04 wizzyrea     clicking through to their online catalogue sends you to koha.org
05:04 wizzyrea     that is interesting
05:03 wizzyrea     http://www.librarytechnology.org/diglib-fulldisplay.pl?SID=20130612290649414&code=lwc&RC=24240&Row=3
05:03 eythian      It apparently last responded with a blank page in May.
05:02 wizzyrea     I think it has gone away
05:01 dcook        Although that error message was still weird..
05:01 dcook        My curiosity is satisfied for now.
05:00 dcook        Someone should do a study of Google research speeds
05:00 wizzyrea     which isn't thaaaaat long
05:00 dcook        That's the one I'm looking at now
05:00 wizzyrea     so a year old
05:00 wizzyrea     for the tsl.state.tx.us site listing for bcls
05:00 wizzyrea     This page last modified on June 25, 2012.
04:59 dcook        Or maybe these are all just really out of date websites
04:59 dcook        But just as a loose affiliation
04:59 dcook        I think "Big Country Library System" might still exist in name..
04:58 dcook        Which is a shame since it's a beautiful system in many ways (although not so much in others)
04:58 dcook        There was some talk in Vancouver
04:58 wizzyrea     yea, and ohio
04:58 dcook        Didn't California have huge cuts too a while ago?
04:57 dcook        Mmm, good call
04:57 wizzyrea     ask her ;)
04:57 wizzyrea     druthb
04:57 wizzyrea     you know who would know
04:57 wizzyrea     they could have been a victim
04:57 wizzyrea     texas had some pretty massive library cuts
04:57 dcook        http://nottypical.org/downloads/FY2013/Overview-NTLPeReaderLabsProject.pdf refers to libraries "formerly" served by them as well
04:57 dcook        Yeah, I saw that
04:57 dcook        I think so too
04:57 wizzyrea     www.bclstx.org -> not responding
04:56 wizzyrea     yea I think they have gone away
04:56 dcook        My curiosity is insatiable :S
04:55 dcook        [off] Wow...I'd heard of PDFs in place of HTML web pages, but really? http://www.wtc.edu/SACS/documents/BigCountryLibrarySystems.pdf
04:54 wahanui      the fine folks at Northeast Kansas Library System
04:54 wizzyrea     NEKLS?
04:54 wizzyrea     NEKLS is <reply> the fine folks at Northeast Kansas Library System
04:53 wahanui      wizzyrea: I forgot nekls
04:53 wizzyrea     forget nekls
04:53 wahanui      i think NEKLS is North East Kansas Library Service
04:53 dcook        NEKLS?
04:53 wizzyrea     now that I think about it
04:53 dcook        [off] I saw the release but that's the last thing for a while
04:53 wizzyrea     also, it may have been disbanded
04:53 wizzyrea     looks like it's a NEKLS type
04:52 wizzyrea     [off] http://liswire.com/node/199 < the target's press release
04:52 dcook        I think maybe it amalgamated with another..
04:51 dcook        It's a weird situation. I can't even find that actual library system or Koha instance
04:51 dcook        No worries
04:51 dcook        lol
04:51 wizzyrea     he's afk, but I will tell him to check :)
04:51 dcook        Yeah, I was just thinking that..
04:51 wizzyrea     def @later him
04:49 dcook        trea: Just thinking...if it used to work for you...maybe it's an issue on their end (or the information you have available)? It looks like they're using Metaproxy rather than Zebra as their z39.50 server...
04:33 dcook        I think that would present earlier though..
04:33 dcook        Unless maybe it's an authentication issue?
04:32 dcook        I'm surprised that using the db "bcls_koha" gives an error with '' rather than 'bcls_koha'
04:31 * dcook      is just curious.
04:31 wizzyrea     probably use it as a target from another koha, if I had to guess
04:31 dcook        Sorry. I know you ran out of time.
04:30 wahanui      "What are you trying to do?" or "What is the goal?"
04:30 dcook        first question?
04:30 trea         it's a remote target, that we don't admin
04:30 trea         if i had that level of access, but alas
04:29 dcook        np. Check their koha-conf.xml next time.
04:29 trea         sorry to query and run
04:29 dcook        Definitely
04:29 trea         probably so. and i'm out of time. will have to pick it up again later o/
04:29 dcook        yep
04:29 wizzyrea     it's gotta be something weird they did with the z db names
04:29 trea         or latter
04:28 trea         in this case, it's the former
04:28 wahanui      Interesting is probably sometimes good and sometimes bad
04:28 dcook        Interesting
04:28 pastebot     "trea" at 127.0.0.1 pasted "Z> base biblios Z> f laser Sen" (11 lines) at http://paste.koha-community.org/68
04:28 trea         oh, is it?
04:28 wizzyrea     that's a former they-who-shall-not-be-named customer eh?
04:27 dcook        Can you try searching again?
04:27 trea         right
04:27 dcook        You did base biblios and that's what you got?
04:27 trea         oh, that's set based on the current "base" command
04:26 dcook        Looks good
04:26 wahanui      interesting is sometimes good and sometimes bad
04:26 dcook        Interesting...
04:25 pastebot     "trea" at 127.0.0.1 pasted "Z> list_all Connected to : tcp" (18 lines) at http://paste.koha-community.org/67
04:25 dcook        trea: What do you get when you do "list_all"?
04:23 pastebot     "wizzyrea" at 127.0.0.1 pasted "Connecting...OK. Sent initrequ" (8 lines) at http://paste.koha-community.org/66
04:23 trea         doing it the way listed above has always worked correctly for me in the past
04:22 pastebot     "trea" at 127.0.0.1 pasted "Z> f laser Sent searchRequest." (9 lines) at http://paste.koha-community.org/65
04:22 wizzyrea     your command gives me
04:22 dcook        No, it's probably the basing..
04:22 dcook        It might relate to your query
04:22 dcook        What's the rest of the error?
04:22 cait         but i think lizis right, there might be something you have to do before you can actually start searching
04:22 dcook        I've used yaz quite a bit..
04:21 cait         hm trying to remember what I did last time
04:21 trea         it's already based
04:21 trea         yaz-client tcp:bcls.kohalibrary.com:210/bcls_koha
04:21 trea         if i run
04:21 wizzyrea     or base bcls_koha
04:21 dcook        Not base <kohadatabasename>
04:21 wizzyrea     base biblios
04:21 dcook        It should be base biblios
04:21 wizzyrea     wait you base using the dbname?
04:20 dcook        Sorry, I should've read more
04:20 dcook        Oh*
04:20 dcook        Pj
04:20 dcook        base biblios
04:20 dcook        trea: yep
04:19 wizzyrea     hum.
04:19 trea         unless having a db with an underscore in the name causes a problem
04:19 trea         just says unavail when doing the search
04:19 trea         yep, i base it using the dbname, and it doesn't throw an error
04:19 wizzyrea     ^ mailing list
04:19 wizzyrea     Jared
04:19 wizzyrea     Regards,
04:19 wizzyrea     Don't forget to choose the database before running a query: base biblios
04:19 trea         yeah it connects fine. it just doesn't return results
04:18 bag          trea is that the right URL?
04:18 trea         none specified
04:17 trea         http://irspy.indexdata.com/full.html?id=Z39.50%3Abcls.kohalibrary.com%3A210%2Fbcls_koha
04:17 wizzyrea     user/pass?
04:17 trea         any idea why yaz-client would respond with "Database Unavailable" errors despite specifying the correct dbname from irspy?
04:15 dcook        oleonard++
04:15 dcook        Really glad though. Very useful work!
04:15 dcook        The code certainly has changed a lot between 3.8 and master. That's for sure...
04:15 wizzyrea     though i recall oleonard tying himself in knots working on virtualshelves.
04:15 dcook        hehe
04:15 wizzyrea     ^ sounds familiar
04:15 dcook        I must be blind or...something
04:14 wizzyrea     probably he's brilliant.
04:14 dcook        It's...mind-boggling
04:14 cjh          hahaha
04:14 dcook        Maybe I should enlist your help to figure out this virtualshelves issue :p
04:14 dcook        Must be those keen 20s eyes
04:14 cjh          dcook: :) just spotted a missing </script> tag
04:13 mtompset     does anyone remember where the multi-select branch limit is?
04:12 cait         hi all :)
04:12 cjh          hi cait :)
04:12 dcook        Nice one, cjh. Whatever it is that you did :)
04:12 dcook        hey ya cait
04:12 * cait       waves
04:10 wizzyrea     PFFFF
04:10 cjh          luck :)
04:10 wizzyrea     cjh++ I'm blind and stuff, and he is not
04:07 dcook        mtj++
04:07 * dcook      waves to Oak
04:06 * Oak        waves
03:33 bag          hello
03:32 cjh          hey bag :)
03:32 mtompset     Greetings, bag.
03:20 bag          evening all
03:13 huginn       04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10455 normal, P5 - low, ---, gmcharlt, NEW , remove redundant 'biblioitems.marc' field
03:13 mtj          bug 10455, there we go...
03:04 mtj          (hopefully that will provoke someone to attempt to justify it's existence)
03:01 mtj          i might log a bug to remove the bi.marc field, just as a start...
03:01 dcook        Bbiab
03:01 dcook        Anyway, lunch time
03:01 dcook        mtj: Np. I think there are a few other instances of that scattered through Koha as well. If we do get rid of the serialization, we'll want to be pretty thorough about checking where it's still being used...
03:01 mtj          hey dcook, thanks for the info ^^
03:00 dcook        gmcharlt: Sounds good. I'll file a bug and post it to the koha-devel listserv
03:00 gmcharlt     gotcha
03:00 dcook        (I mean Stephen King)
03:00 dcook        It's like he wasn't even trying
03:00 gmcharlt     dcook: it's pretty ancient code, though; worth asking the UNIMARC users if it still servers a purpose
02:59 dcook        I'll never understand the end of "Needful Things"
02:59 dcook        gmcharlt: Too true.
02:59 dcook        So perhaps damage control is best...:/
02:59 dcook        But I'm sure "someone" uses it
02:59 dcook        Personally, I rather remove the whole block and the system preferences...
02:59 gmcharlt     I mean, just look at Stephen King
02:59 dcook        gmcharlt: Agreed.
02:58 gmcharlt     dcook: I don't know, some authors are *very* strange
02:58 dcook        I don't know why one would want to "normalize" the author...
02:55 dcook        And filing a bug to tell someone who knows unimarc to probably look at it
02:55 dcook        I suppose wrapping it in a check "MarcFlavour" might be enough for now...
02:54 dcook        I'm not even sure where to start with re-writing it. Since it's for UNIMARC, it looks completely alien
02:54 dcook        hehe
02:53 wizzyrea     "hissy fit" is a term that is not used often enough in relation to Koha
02:52 dcook        Koha has a bit of a hissy fit..
02:52 dcook        It deletes things and tries adding them before even checking that the variable has been set
02:52 dcook        Even if you were a UNIMARC user...
02:51 dcook        But if you do...yeesh
02:51 dcook        As a MARC21 user, there's no reason to have them turned on
02:51 dcook        The code controlled by "z3950NormalizeAuthor" and "z3950AuthorAuthFields"...
02:50 mtompset     Yes there is, dcook. Sadly, there is.
02:49 dcook        Man...there is some weird code in Koha...
02:48 mtompset     I suppose gmcharlt's table with autoincrement for ID is good for that, but the problem is cardnumber isn't always numeric. :(
02:47 mtompset     This is what locking does create. But since it isn't recommended on Inno DB's, that really isn't a good solution.
02:46 mtompset     transactions don't create critical sections. Though, they do make things atomic.
02:46 mtompset     transactions don't create critical sections. Though, the do make things atomic.
02:45 mtompset     That's the phrase I was trying to remember... critical section.
02:37 dcook        GetImportRecordMarc. I imagine it could be GetImportRecordMarcXML instead. I think the MARC module can create a MARC object from MARCXML and not just the serialization...
02:36 dcook        sub MARCfindbreeding in addbiblio.pl
02:36 dcook        mtj: Just found an instance of when usmarc is used over marcxml
01:57 dcook        Today is the day of bugs...
01:45 dcook        Because pages can equal undef but it doesn't come out as 33 or 34..
01:45 dcook        But...that can't be quite right...
01:42 dcook        And somehow...Template Toolkit is translating undef as 33 (or 34)...
01:42 dcook        Hmm, so Koha is generating an itemloop where size = undef
01:06 mtompset     And that would solve my problem. I'm going to give that a whirl shortly. Bug report title to change shortly. ;)
01:04 mtompset     I do think a transaction patch in AddMembers would work.
01:01 mtompset     But how will that prevent a second process from starting the same transaction? Does MySQL only run one transaction at a time?
01:00 mtompset     so START TRANSACTION fixup_cardnumber AddMember COMMIT END TRANSACTION?
00:57 rangi        thats what they are for
00:56 rangi        if you want atomic, use transactions
00:54 mtompset     then references to fixup_cardnumber could be removed.
00:53 mtompset     Okay... so then really AddMember could be tweaked to add either the hash directly, or the hash plus a generated number in a single SQL statement, which should not need locking.
00:51 gmcharlt     mtompset: sure -- following the example, one would just grab the results of "select nextval('abc');"
00:50 mtompset     Can you do stored procedures with DBI, gmcharlt?
00:50 dcook        When I run the SQL, my results are showing the size column as a primary key as well...no idea why that's happening...
00:50 dcook        size
00:50 pastebot     "dcook" at 127.0.0.1 pasted "virtual shelves null issue" (7 lines) at http://paste.koha-community.org/64
00:49 gmcharlt     dcook: which column?
00:49 * dcook      's brain explodes
00:49 dcook        jcamins: When I look on my dev system, it shows 34
00:48 jcamins      dcook: e^0 would have been harder to figure out.
00:48 dcook        But there does have to be a logic to it...
00:47 gmcharlt     particular comment #8
00:47 gmcharlt     mtompset: http://www.microshell.com/database/mysql/emulating-nextval-function-to-get-sequence-in-mysql/
00:47 dcook        jcamins: I caught on eventually :p
00:47 jcamins      mtompset: fixup_cardnumber is not okay if it doesn't work. I was suggesting a way to have something that works now.
00:47 jcamins      dcook: that was nonsense, BTW. :P
00:47 dcook        wth...
00:46 jcamins      dcook: it's shelf e^-1. That's why.
00:46 dcook        I just noticed it in 3.8
00:46 * dcook      is checking out master atm
00:45 dcook        Although I don't know why it would be happening in virtualshelves and nowhere else..
00:45 dcook        jcamins: I figured there had to be some sort of logic to it :p
00:44 jcamins      dcook: base e logic.
00:44 dcook        gmcharlt: I know, right?
00:44 gmcharlt     rather, weird, not weird?
00:44 gmcharlt     dcook: '33'?  weird?  converting null to '', I would understand
00:44 mtompset     that makes the locking tables a bad idea.
00:44 mtompset     that makes the lockink tables a bad idea.
00:43 mtompset     Okay, the "not recommended"
00:42 gmcharlt     mtompset: it is not functional unless one assumes that there will never be simultaneous writes to the borrowers table; it doesn't matter how short the average lock period is
00:39 * dcook      would love to know how Koha turning a null in Mysql into "33"...
00:37 mtompset     Which, I do admit is ugly, but it is functional.
00:37 mtompset     Locking tables makes it atomic.
00:37 mtompset     fixup_cardnumber is okay, but it does no locking... and a second process could grab the same calculated number before AddMember is called.
00:33 mtompset     Yes. That is the problem that I am seeing.
00:33 jcamins      mtompset: the problem you are trying to deal with is generating cardnumbers automatically, right?
00:32 mtompset     A hidden syspref? With row locking?
00:31 gmcharlt     mtompset: table locks are *specifically* not recommend for innodb; see http://dev.mysql.com/doc/refman/5.5/en/table-locking.html; one possible alternative might be creating a table with an auto_increment key whose sole purpose is to emulatie a sequence
00:31 mtompset     automatic card number generator? What function are you thinking of? Or is this something you are suggesting I build?
00:30 jcamins      Something like SAML domain + ID, or whatever SAML exposes.
00:28 jcamins      If you're still working on SAML, why don't you just not use the built in automatic card number generator, which has always been not-best-practice?
00:26 dcook        O_o
00:25 mtompset     I can conceptually agree to that. The problem is that I want the assurance of something that works now.
00:24 gmcharlt     which is something that IMO calls for a solution that just about anything other than diong table locks
00:23 mtompset     if you don't then the fixup_cardnumber doesn't generate a unique number in some cases.
00:23 gmcharlt     innodb has row-level locks
00:23 mtompset     It generates an error log entry otherwise.
00:23 gmcharlt     we shouldn't need to lock *anything*
00:22 jcamins      Do you need to lock action_logs?
00:22 mtompset     I put in a 90 second lock to see how it affected other accesses.
00:22 mtompset     But the time is under a second for the lock, so I don't see a problem.
00:21 mtompset     The others prevent silent errors in the error log file.
00:21 pastebot     "mtompset" at 127.0.0.1 pasted "AddMember_Opac testing" (7 lines) at http://paste.koha-community.org/63
00:20 dcook        Oh, Koha. Why you so crazy...
00:19 mtompset     Ah... good point... forgot about the random string mess.
00:17 gmcharlt     also, upon reading scrollback, are you proposing to add code to actually lock the borrowers table?
00:10 gmcharlt     er, and thereby remove the ability for patron imports to set patron passwords?
00:08 mtompset     Because I'm about to submit a locking patch for AddMember_Opac, which would then remove the FIXME.
00:08 mtompset     Because I'm thinking of taking a line from tools/import_borrowers.pl and changing the logic to use AddMember_Opac if the cardnumber is not set and AddMember if it is set, and then just use if ($borrowernumber) instead of if ($borrowernumber = AddMember(%borrower)) { ... }
00:05 gmcharlt     definitely the sort of contruct that if you used it (and please don't) woudl require a comment saying "yes, I really mean it that way"
00:04 jcamins      Because it will not be understood.
00:04 jcamins      But it is wrong.
00:04 jcamins      It means the same thing, yes.
00:02 mtompset     I'd like to confirm... if ($a=$b) { ... } is the same as $a=$b; if ($a) { ... } correct?