IRC log for #koha, 2008-12-06

All times shown according to UTC.

Time S Nick Message
15:34 nengard Stats question:
15:34 nengard In the calendar for the statistical reports are the dates inclusive or exclusive?  In other words if I need the circ stats for all of November do I pick:
15:34 nengard Nov 1 to Nov 30 or
15:34 nengard Nov 1 to Dec 1 or
15:34 nengard Oct 31 to Nov 30 or
15:34 nengard Oct 31 to Dec 1 ?
15:36 hdl nengard: was designed as exclusive.
15:36 hdl But user usually think it is/should be inclusive.
15:40 nengard hdl so that means that I should put in  Oct 31 to Dec 1
15:44 nengard ?
15:49 hdl sorry
15:49 hdl yes
19:25 atz hdl: I find special behavior with statistics report date range because the field being queried is DATETIME and not just DATE
19:25 atz mysql> select "2007-12-06 17:25:39" < "2007-12-06";
19:25 atz +--------------------------------------+
19:25 atz | "2007-12-06 17:25:39" < "2007-12-06" |
19:25 atz +--------------------------------------+
19:25 atz |                                    0 |
19:25 atz +--------------------------------------+
19:25 atz 1 row in set (0.00 sec)
19:25 atz mysql> select "2007-12-06 17:25:39" > "2007-12-06";
19:25 atz +--------------------------------------+
19:25 atz | "2007-12-06 17:25:39" > "2007-12-06" |
19:25 atz +--------------------------------------+
19:25 atz |                                    1 |
19:25 atz +--------------------------------------+
19:25 atz 1 row in set (0.02 sec)
19:25 atz so it range will include on lower bound, exclude on upper bound
19:26 atz this would not be true if the field were just DATE
19:29 atz this does what users want in most cases, i think
19:36 atz so you can say   6/1 - 7/1 for June instead of having to remember how many days June has...
20:24 hdl hi atz
20:26 hdl atz: should the DATETIME data be changed to DATE for 3.2 ?
20:26 hdl in your opinion ?
20:28 gmcharlt hdl: not that you asked me, but I would say not :)
20:29 gmcharlt with hourly loans, will need to track circ events down to the minute
20:29 hdl hi gmcharlt
20:30 hdl gmcharlt++
20:30 gmcharlt for reports that only worry about the date
20:30 gmcharlt query can be munged to get expected results
20:31 hdl btw I think that bounds should be >= and <= and not > and < for stats wizard.
20:31 atz hdl: in this case, i would not change.  probably changes will go the other direction with many DATE fields becoming DATETIME
20:31 hdl users are confused.
20:32 atz hdl: it won't matter for comparison of just_date against date_and_time
20:32 atz because 2008
20:32 atz sorry
20:33 atz because "2008-12-05" will NEVER match "2008-12-05 00:00:01"
20:33 atz or any other DATETIME value
20:34 hdl but we could take the date part of it in order to compare
20:34 hdl And not the whole DateTime field.
20:34 atz or we could add 00:00:00 (or 99:99:99) to the query
20:35 atz which would save the recomputation against every line queried
20:35 hdl could be faster yes.
20:35 hdl But would need testing.
20:36 atz (actually we would need to use 23:59:59 or something, because I think you can "overload" the time and get a different day
20:37 atz but the strange thing is that basically, i think users are getting what they want with the date against datetime queries
20:37 hdl gmcharlt: atz Is LL considering using an object framework in developments or not at all ?
20:37 atz the problem is that the behavior is different for date against date queries
20:38 hdl ok atz.
20:42 atz hdl: do you mean something like DBIx::Class ?
20:43 atz i think all of us are fans of using it.
20:43 hdl yes we are.
20:43 hdl But not only
20:43 hdl I was thinking of MOOSE or Class::Std
20:44 atz MOOSE is new to me, but seems worth looking at.  acmoore reviewed it and seemed to prefer DBIx::Class at this time.  He also looked at Rose and Class::DBI amongst others.
20:45 atz not sure about Class::Std.  
20:45 hdl In fact, for me DBIx::Class is rather DB/storage and query oriented
20:45 gmcharlt Moose/Class::Std are for a different problem space than DBIx::Class/Rose::DB
20:46 hdl MOOSE or Class::Std are for OOP and object representation of our programming paradigm.
20:47 hdl In fact, i think that Koha would much benefit from Some object abstractions.
20:47 atz almost certainly
20:51 acmoore I'd love to turn some of our packages into classes. I could see using something like Moose, or even starting out by using no framework at all. I kind of thought I'd start lobbying for that after I got much of the code using an ORM.
20:52 acmoore hdl have you used Moose before? I only used it in one small module to get a feel for it.
20:54 hdl Nope, but mc thinks much good of it.
20:55 acmoore oh, good. An ally! ;)
20:55 hdl And we shall build a POC on importing borrowers data  (and maybe subscription) next week.
20:55 hdl count me too.
20:56 hdl But we have to be cautious and quite.... sane.
20:57 hdl this change should be invisible for users.
20:57 acmoore sure. And, it will be a long process.
20:57 hdl yes.
20:57 acmoore I think we should see how well the DBIx::Class conversion goes before we start on that stuff.
20:57 acmoore hdl: do you think it would be helpful if I gave one or more of the BibLibre guys (or anyone else) some interactive help on getting started with DBIx::Class?
20:58 hdl But parts by parts, the game would take on well.
20:58 acmoore I'm not sure how I envision it, but I would like to do what I can to make it easy for you guys to start using it at some point.
20:58 atz yeah, many things will get easier, including creating OO layers, if we can incorporate DBIx
20:59 acmoore I can write some documentation, but sometimes showing someone some basic steps makes it easier to get going.
21:01 hdl problems is that we are atm overwhelmed by job for 3 Universities.
21:01 hdl But documentation would be helpfull.
21:01 hdl i read your patches on C4::Members.
21:01 acmoore OK. I'll do what I can to write up some useful documentation. But, you should feel free to ask me lots of questions.
21:02 hdl I also read DBIx::Class documentation.
21:02 acmoore areinmeyer is working on some more patches for C4::Members. We're hoping to have all of the functions in it using DBIx::Class from the beginning. That will serve a bit as a reference example for you, too.
21:03 acmoore yeah, the DBIx::Class documentation is pretty good. the #dbix-class channel on is also quite helpful.
21:03 hdl How soon are you hoping to have creation and modifiction of data ?
21:03 hdl (Members) ?
21:04 acmoore I think that areinmeyer is doing that today, and I'm converting C4::ImportBatch to do that today, too. I've replaced some SQL with DBIx::Class calls to insert and update.
21:04 acmoore Maybe we'll get some example patches out next Friday.
21:04 chris sweet
21:05 acmoore we pretty much just take one day a week to work on these side projects, so progress is sometimes in bursts.
21:05 chris be able to do some nice benchmarking at that point
21:05 acmoore chris: that's the plan!
21:05 hdl well, a little bet too late for me, but will try to start with what you did.
21:06 hdl If I can have creation and modification on my own I will try and commit it on a website.
21:07 hdl I could have a good Benchmark with what i have to do.
21:07 hdl hi chris.
21:07 acmoore OK. please send over anything you like.
21:07 chris hi hdl (and everyone else)
21:08 gmcharlt hi chris
21:34 danny I see a bunch of warn statements in my, however, I can't find where these are actually in the log file, is this file treated differently does anyone know?
21:34 danny they aren't commented out
21:35 atz danny: *perhaps* they are for situations where warning is warranted?
21:35 gmcharlt most of them are
21:36 danny I want to see the warning though, i would use it to debug something
21:36 danny i just dont know where to find it
21:36 gmcharlt Apache error log
21:36 gmcharlt but most of the warns in C4/, as atz says, would not be triggered during normal use
21:36 atz assuming you are using the web interface for a request, Apache log...   on the command line it would be STDERR
21:37 danny normally i find them in koha-opac-error..
21:37 danny hmm
21:37 atz perhaps you are on the staff side?
21:37 atz '
21:38 atz (since auth is more critical to staff interface, used on every page -- well almost every page)
21:40 danny right, i checked both the staff and opac logs
21:40 danny and in the perl script for example in the checkpw function I see a statement like warn "text here";
21:41 danny doesn't really say text here but you know what i mean just a normal warn statement
21:41 danny what would cause that not to be triggered?
21:41 atz .... unless ( $in->{'type'} eq 'opac' );
21:43 atz i only see 11 warns in C4::Auth
21:45 atz 10 actually
21:45 danny our is customized a bit and I was just trying to debug it with warns
21:45 chris using apache2 danny?
21:45 chris check the main error log also
21:45 danny was just finding it odd that they weren't appearing, but i'll keep investigating
21:46 danny yes we are
21:46 chris if you are using mpm-worker it sometimes outputs warns to the main error log rather than the one defined by the virtual host
21:46 atz ah... probably need to trace the code a bit more to make sure you are in the right section.  auth has a lot of copy/paste sections that look similar, unfortunately
21:46 danny ok nevermind, sorry guys, i just tested by placing one after the first line of the sub and it does show up, so it just isn't making it that far into the if statements
21:48 danny which would explain the problem too most likely
21:49 danny i just wasn't sure if there was some like global perl like that was like hide all warn output in this package
21:53 atz danny: you can redirect STDERR, but we don't do that for most of koha since we want the errors to hit the log
21:54 danny ah ok, I understand, thanks
22:27 danny ryan around?
03:45 atz i7uuy5b saccc95x48d9ewz4888878888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​8888888888888888888888888888888888888888888888888​88888888888888888888888888888888888888888888888
03:56 atz .=================================================​==================================================​==================================================​==================================================​==================================================​==================+4444444444444444444444444444444​44444444444444444444444444444444444444444444444444​44444444444444444444444444444444444444444444444444​444444444444444444444444444444444444444440*****
04:21 atz QQ​​​​​​​​
04:24 atz 7O*OO&56666666666666666666666666666666658​884768..............JHHHHHHH9.......​..........................7888888888
04:24 atz 999999999999999999999999999999999999999999999​999999999880777777777000000000000000000000000​0000000000000000000000000000000000000777765
04:25 atz rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr​rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr​rrrrrrrrrrrrrrrr7;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;​;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;56
04:25 atz gmcharlt: ]66666
04:29 atz _______________________________________________J​HRRRRRE;llllllllllllllllllllllllllllllllllllllll​llllllllllllllllllllllllllllllllllllwwwwwh;wjp//​///w;/RWW}_}w-[[[--------[[----------[[--------[​------[---[--[--[--[-------=0jjjjjjjjjjjjjjj
07:38 gelinp Hi, I'm installing koha 3.0 untu Ubuntu Hardy. I've got a problem with the install of the Perl modul "YAML::Syck". The error message is: [ Can't locate object method "install" via package "YAML::Syck" ]
10:37 gelinp Hi...
10:37 gelinp Is there anyone to help me?
10:37 gelinp I try to install koha 3.0b2 under Linux Ubuntu Hardy...

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