IRC log for #koha, 2007-06-25

All times shown according to UTC.

Time S Nick Message
12:11 [K] *** join #koha@FreeNode: jaron
15:18 BoSchafers Haven't used IRC since 1996....lets see
15:53 BoSchafers Hi folks, been in here a while, no visible activity, are you all in private conversations or busy doing Koha stuff :)
15:54 [K] <jaron@FreeNode> If there's a project day today they might not be starting until noon EST.
15:57 BoSchafers ok so are they just parked here? Names only?
15:57 [K] <jaron@FreeNode> yeah, lots of folks are always logged in.
15:57 [K] <jaron@FreeNode> maybe reading the scrollback is easier than reading the logs
15:57 [K] <jaron@FreeNode> or they're around just in case someone pings them
15:58 [K] <jaron@FreeNode> if there's someone you want just put their nick in a message and they might wake up :)
15:58 [K] <jaron@FreeNode> BoSchafers: ping?
15:59 BoSchafers its actually quite late here in Oz and I thought there's be a lot of  *action* here
15:59 BoSchafers i have not used irc in ten its all a bit rusty
15:59 [K] <jaron@FreeNode> I think Sundays can be rather quiet if there's not a project day
15:59 [K] <jaron@FreeNode> many weekdays can be quiet as well :)
16:00 BoSchafers i see and on weekdays there are more people here?
16:00 BoSchafers ok...
16:00 BoSchafers well my name is Bo.....hi there K :)
16:01 BoSchafers thanks for the info
16:01 [K] <jaron@FreeNode> there can be. Actually I'm over on freenode right now so my messages are being relayed by the bot [K].
16:01 BoSchafers no worries...i'll head off for now...many thanks for the info...again
16:01 [K] <jaron@FreeNode> sure. take care.
16:02 [K] * jaron@FreeNode would head off too if he weren't at work :(
16:02 BoSchafers :-)
16:27 ryan hello #koha
17:24 martin hello, got another question about z39.50 searches in koha
17:25 martin if there's anybody awake in NZ right now i'd welcome a few mins of your time :)
17:29 martin no, everybody must be asleep
17:29 martin basically, if anyone sees this, i'm having trouble getting LOC records down
17:29 ryan hi martin
17:29 ryan what sort of troubles ?
17:29 martin hello ryan
17:30 martin i've configured a z39.50 server on my intranet
17:30 martin for library of congress
17:30 martin now, the default installation of koha seems to offer only USMARC and UNIMARC as options for the syntax
17:30 martin but poking around on the koha website i've managed to get an updated z3950servers.tmpl file which i've substituted for the default one
17:31 martin that, however, doesn't include a MARC21 option, and from what i can tell Net::Z3950::ZOOM doesn't mention it either
17:31 martin so i'm wondering what to do basicaly :)
17:31 martin when i do the search, the popup comes up offering me as configured, when i asked to search it the server thinks for a second then gives me a blank box
17:32 martin i've even tried adding two extra entries <option value="MARC21">MARC21</option> and "MARC 21" and have tried changing the syntax to OPAC as i see it offers that too
17:33 martin i'll shut up now :)
17:33 ryan USMARC and MARC21 are interchangeable.
17:33 martin ah, really?
17:33 ryan USMARC should work fine for LOC.
17:33 martin i thought they were a little  bit different, OK i'll try usmarc again
17:34 ryan as far as z39.50 services, differences are negligible
17:34 martin fair enough, let me see what happens
17:36 martin nope, still getting my blank box
17:36 [K] <jaron@FreeNode> LC will give the same results whether you ask for usmarc or marc21 preferred record syntax either one works in the client I use
17:36 martin wonder what i'm doing wrong then
17:37 martin hostname:
17:37 martin port: 7090  Database: Voyager, syntax USMARC
17:38 [K] <jaron@FreeNode> that all looks correct
17:38 martin and yours works fine...?
17:38 [K] <jaron@FreeNode> martin: I don't use the koha z39.50 client
17:38 martin ah right, can i ask what you do?
17:39 [K] <jaron@FreeNode> and I just tried it for the first time in a long time and it gave me a server error.
17:39 martin i don't even get that
17:39 [K] <jaron@FreeNode> I use something I wrote myself
17:39 martin i get the list of z39.50 servers i have configured, ask to search and it comes b ack with a blank page
17:40 ryan martin: did you revert to the released template ?
17:40 martin no i haven't, perhaps I should
17:40 martin but i got exactly the same problem with that template too
17:40 martin and i was using USMARC then
17:41 ryan 2.2.9?
17:41 martin yes
17:41 martin freshly installed this afternoon
17:41 ryan can you post some error logs?
17:42 martin i've not found any let me have another look though
17:42 [K] <jaron@FreeNode> /usr/local/koha/log/koha-error_log if you do a default install on linux
17:42 martin thanks jaron, i was looking at teh apache log
17:43 martin two secs
17:44 martin ah got something, let me do another search so i get some clean logs and i'll cut and paste them
17:44 martin is there somewhere polite i can paste them so i don't bother others here?
17:45 [K] <jaron@FreeNode> lisppaste3 ?
17:45 martin that means nothing to me
17:46 [K] <jaron@FreeNode>
17:46 martin actually it doesn't matter, nothing in the error log
17:46 martin only get this in the koha-access log
17:46 martin - - [24/Jun/2007:18:40:32 +0100] "GET /cgi-bin/koha/z3950/​ldbiblionumber=&isbn=0844237639 HTTP/1.1" 2
17:46 martin 00 806
17:46 martin - - [24/Jun/2007:18:40:35 +0100] "GET /intranet-tmpl/npl/en/includes/intranet.css HTTP/1.1" 304 -
17:46 martin - - [24/Jun/2007:18:40:35 +0100] "GET /intranet-tmpl/npl/en/includes/colors.css HTTP/1.1" 304 -
17:46 martin - - [24/Jun/2007:18:40:47 +0100] "POST /cgi-bin/koha/z3950/ HTTP/1.1" 500 437
17:48 [K] <jaron@FreeNode> don't take my word for it but that doesn't look like the problem
17:48 martin fair enough
17:48 [K] <jaron@FreeNode> have you searched the archives of the list?
17:48 martin yes i have
17:48 martin but i will search again
17:48 [K] <jaron@FreeNode> hrm.
17:49 ryan martin: i think you should get something in either koha-error or apache error logs
17:49 martin i wouldn't bother but i have a buckletload of stuff to catalogue and if this works it would save me a lot of time
17:50 martin there realy is nothing ryan
17:50 martin do i need to up the debugging level perhaps?
17:51 martin will try the british library in the meantime
17:52 [K] *** part FreeNode!#koha: jaron
21:01 chris yikes its bucketing down here
21:02 russel yikes sarah just left for a walk
21:06 chris hmm seems to have eased off again
21:28 chris hi martin .. must be getting late for you?
21:30 martinmorris 2230 :)
21:30 martinmorris just trying ot get loads of stuff into my new catalogue
21:30 martinmorris think i need a shorter marc input page
21:30 martinmorris !
21:30 chris :)
21:30 chris the thing you can do
21:30 martinmorris the old catalogue is just a text file mostly without isbn's etc etc - terrible
21:30 chris is change which tabs (down the side 0-9)
21:31 chris certain info displays on
21:31 martinmorris ah, that would be useful
21:31 chris so you can go through the framework, and put all the tags you want on tab 0
21:31 chris and the ones you dont often use leave on the other ones
21:31 martinmorris is that under koha administration?
21:32 chris yep
21:32 chris under marc frameworks
21:32 martinmorris ok let me take a look
21:33 martinmorris is that the biblio frameworks (marc structure) heading?
21:33 chris ah yep that sounds right
21:34 martinmorris not immediately seeing how i can do that here
21:34 chris hmm 2 secs
21:35 chris you get to a page that has default framework .. then a MARC Structure button?
21:36 martinmorris no, i'm using the npl template though would that change it?
21:36 martinmorris i get a page headed "MARC tag structure admin for default MARC framework"
21:36 martinmorris drop down box to select a framework with 'default' highlighted
21:37 martinmorris and a tag search button with box
21:37 martinmorris oh i beg you rpardon, that's the next page
21:37 martinmorris on teh previous page i do get that, yes
21:37 chris cool so you clicked marc structure sweet
21:37 martinmorris i click on the 'MARC Structure' link and got this, yes
21:38 chris so you have a big bunch of tags on that page too eh?
21:38 chris starting with 000
21:38 martinmorris yes I do
21:38 chris cool
21:38 chris so lets say we want to shift 100a (personal name)
21:38 chris to tab 0
21:39 martinmorris yep
21:39 chris in the tag seach box, type 100
21:39 chris and hit tag search
21:39 martinmorris i've searched for tag 100 and it's showing on my page
21:39 martinmorris done
21:39 chris ok click view subfields for 100
21:40 martinmorris done that
21:40 chris and you get a page listing all the subfields for it (now i think it might be on 0 already?)
21:40 chris ersonal name   Tab:0, | Koha field:, Not repeatable, Not mandatory, | See Also: '110a', '700a', '710a',
21:40 martinmorris i don't think it is
21:40 chris ahh ok, maybe ive already shifted it on this one :)
21:40 martinmorris it's on tab 1
21:40 chris ok, so what we do
21:40 martinmorris we edit the subfields?
21:40 chris thats the one
21:41 martinmorris and i see the 'Managed in tab' option
21:41 martinmorris just what i'm after, thank you
21:41 chris there ya go
21:41 martinmorris that will save time
21:41 chris[…]1s03.html#d0e1238  
21:41 martinmorris buried away a little there :)
21:41 chris might be helpful too
21:41 martinmorris i think i need to go through the basic set up befor egoing much further
21:41 martinmorris thank you chris
21:44 chris no prob
21:44 chris have you seen the site?
21:44 chris <-- might or might not be useful for you too
22:11 chris hi mason
22:16 mason hiya chris
22:24 slef hi
22:24 chris hi slef
22:27 slef how's it going?
22:27 chris not too bad so far, pretty quiet for a monday
22:27 chris hows things over there?
22:27 slef sleepy sunday evening here
22:28 slef been out with the in-laws
22:28 chris ahh
22:28 chris go well?
22:28 slef seemed to
22:28 slef now trying to dent project emails before sleep
22:29 slef so monday morning isn't too manic
22:29 chris fair enough
22:29 chris im doing a bit more work on unit tests today
22:29 slef just opened wrong mailbox, bit of a pain
22:30 slef is it possible to move the dev meeting to a day later, do you think?
22:30 slef Mondays are often awkward and that one is particularly so
22:30 chris reckon it could be
22:30 slef should be discussing a big contract in London and probably won't be back in time
22:30 chris drop a mail to the devel list
22:31 slef I'll drop a mail RSN
22:31 chris and ill try not to sleep through the meeting this time
22:31 chris :)
22:31 slef yeah, wouldn't later be better?  Or does that scupper paul?
22:32 chris i think yeah it scuppers paul
22:32 russel is better for us early anyway
22:32 chris 7am isnt too bad for me normally, just had a bad nights sleep
22:32 russel i would have thought the baby would have chris up before 7 anyway ;-)
22:33 slef except last week
22:33 chris yeah that day unfortunately he'd about finally gone to sleep
22:33 slef Spanish-speakers walking into the conference accommodation at like 5am
22:33 slef "Hey guys, just because you speak Spanish, it doesn't mean we can't
22:33 slef hear you!"
22:33 chris heh
22:34 slef was good to meet a venezuelan kohaista(?), though
22:35 chris yeah thats cool :)
22:35 russel v cool
22:35 slef his talk's already up online if you want to watch... shall I find the URL?
22:36 chris oh yeah, that would rule
22:36 slef Moment, bitte
22:37 chris slef have you tried #koha on freenode ?
22:37 slef nah, can't get a stable connection to freesplit
22:37 chris :)
22:38 slef OFTC and IRCnet are probably my most-used except for here
22:38 slef I suspect my irc relay has a problem.  I'll check it RSN
22:38 slef http://meetings-archive.debian[…]ation_process.ogg
22:38 slef I think that's the talk
22:39 slef yep, that's bureado
22:41 chris 124 meg .. that might take a little while  :)
22:42 slef about an hour of talk+discussion
22:43 chris sweet
22:43 chris nice background listening
22:43 slef mostly about debian, just a quick mention of koha and a call-out to me
22:43 slef it's video btw
22:43 chris oh even better
22:43 slef I was asleep+IRCing at the back of the room, so pretty shocked
22:44 chris heh
22:44 chris yeah i remember you saying something
22:45 slef erm, 404s
22:45 slef that'll be because it's
22:45 slef really must find my login details and add that to the web pages
22:47 slef approx 24 minutes in, I think
22:47 chris 16% downloaded :)
22:49 chris coffee time brb
22:52 slef 1360seconds in
22:55 chris back
22:58 slef chris: thanks for squashing that mysql 5-v-3 bug.  I'll take a quick look for other suspects
22:58 chris no problem
23:05 slef chris: erm, that fix will break installs where mysql host != localhost
23:05 chris good point
23:06 chris at that point have we asked them to enter the host the mysql db will live on .. i think we just assume localhost eh?
23:06 slef but I think if you use % in mysql 5, it doesn't get localhost
23:06 chris yeah
23:06 slef yeah, I guess we put localhost iff mysql host is localhost and % otherwise
23:07 slef or we ask them what mysql thinks the webserver hostname is
23:07 slef ye gods there's some cruft in the misc directory
23:08 chris isnt there just
23:08 chris once we switch to git much easier to tidy/mv stuff around
23:09 slef oh, the serene koha-git tree broke around Tuesday while I was away
23:09 chris bummer
23:09 slef hitting some sort of memory ulimit
23:09 slef will fix RSN
23:09 slef git cvsimport is not light with a cvs the size of koha's, it seems
23:10 chris yeah
23:10 chris at least if we switch over .. we wont have to do that anymore :)
23:10 chris i assume thats the plan eh?
23:12 slef ask the man with the plan
23:12 chris :)
23:12 chris kaitiaki
23:14 rach I don't mind giving it up, but I personally don't know that liblime should be both current release manager and kaitiaki - but I am doing a rubbish job so happy to take suggestions
23:14 chris yeah i agree with rach
23:14 chris (about the first bit)
23:15 chris :)
23:15 slef rach: boss us all around a bit ;-)
23:17 slef who's running kohala?
23:17 chris yeah that might be good
23:17 chris not sure
23:19 rach how is that release coming along :-)
23:19 slef rach: as paul wrote "install from the CVS won't work at all"
23:19 slef which is a bit of a blocker, I suspect
23:20 chris yeah there is a bit of monkeying around to get 3.0 up and running
23:21 chris can i get a quick sanity check slef
23:21 slef I have some install notes
23:21 slef yeah, but I'm sleepy
23:21 chris[…]lopment:git_usage
23:21 chris is that something like you would envisage it working?
23:22 chris (im trying to do a picture to make explaining easier)
23:22 slef I'd have 2.2 and 3.0 talking to each other, maybe through backporters and forwardporters
23:22 chris ahh good idea
23:22 slef and I'd have world talking to everyone
23:23 slef with Master just as a sort of archive service
23:23 slef and reference point
23:23 chris k
23:23 slef after all, the releases will come from the RM trees
23:23 chris true
23:28 slef ooh, kados already scheduled axing rach[…]tingnotes07july02
23:29 chris i think we need to come up with a plan of how to get a new kaitiaki .. ie how thats gonna work
23:30 rach man, you hang about doing nothing for a year or so and you get the boot :-)
23:31 chris lol
23:31 rach better than the bash though
23:31 chris :)
23:32 slef this'll teach me to joke about freenode, won't it?
23:33 slef anyway, see you in 8h
23:47 kados russel: that looks great
23:51 rach do you think it might get confusing that it looks the same as the main koha site?
23:52 kados I think the goal is to make it part of the main koha site
23:52 rach I was wondering wether having the word docs as part of the header might be good?
23:52 rach if that's the case then you might be better to "fake" it being part of that site
23:53 rach so keep the overall nav etc from the main site
23:53 rach and have this like it's a subsite?
23:53 rach or a section
23:53 rach
23:54 russel one thing at a time :-)
23:54 rach at the moment it's ugly but obvious
23:54 rach but russ sticking your usability hat on, now it's pretty but non obvious
23:55 russel ideally i want to move it all back into the community section
23:55 russel eventually - but there are masses of redirects etc that need writing
23:56 rach but you could fake it with the design
23:56 russel hence "indexworking.html"
23:56 kados russel: what do you think about putting everything in plone?
23:56 kados it'd be much easier to manage there, and we could do translations
23:56 rach OR if that's to much right now, change the header image colour, or add the word docs after the logo - like Koha | Docs
23:57 rach so it's clear you've gone to another site in the koha family
23:59 russel i dont think changing the header colour is a good idea
23:59 russel but i will add the text
00:01 russel there are already other subsites
00:01 russel
00:01 russel
00:02 kados yep, I agree
00:02 kados unify the look and feel I think is the goal
00:02 kados !
00:02 kados (not ready yet, but some day)
00:04 kados good idea
00:04 kados so maybe reflect that in the header somehow?
00:05 rach so you might have slight differences in the headers - but that they would clearly all be the same general thing
00:06 rach Other organisations do it by having the line of sites at the top above the banner
00:06 kados rach: you volunteering to design that? :-)
00:06 russel
00:06 russel they just do it with a word
00:06 russel seems nice and simple to me
00:06 rach I'm thinking the news sites
00:06 rach yep - re the word
00:07 rach and we could have a line like on stuff -
00:07 rach of the various sites
00:07 rach ah although on the main koha sites there are already things up there
00:08 russel
00:09 russel my aim with this exercise was that by wessing around with this page - i'll find all the docs and can start structuring
00:09 russel them
00:10 russel so this is a means to an end
00:11 russel anyway spot of lunch me thinks
00:12 chris kados:[…]lopment:git_usage  changed after some feedback from slef
00:12 kados chris: sweet
00:13 rach cheers russ
00:14 kados chris: can you expand it a bit to show how it would work to allow a company to have their own repo too?
00:14 [K] <rangi@FreeNode> company is just a developer
00:14 kados hmmm
00:14 [K] <rangi@FreeNode> they would just have some branches in their clone
00:15 [K] * rangi@FreeNode talks from freenode  <--- chris
00:15 kados so we have a company-wide 'developer'?
00:15 kados heh
00:15 [K] <rangi@FreeNode> well you would clone one
00:15 [K] <rangi@FreeNode> then individuals would clone from that
00:16 kados clone off the RM's clone?
00:16 [K] <rangi@FreeNode> yep
00:17 [K] <rangi@FreeNode> just add another developer off the developer essentially
00:17 [K] <rangi@FreeNode> or you could all just clone of rm
00:17 [K] <rangi@FreeNode> and pass patches around
00:17 kados so here's what I'm thinking
00:17 kados we have Koha Classic
00:17 [K] <rangi@FreeNode> internally its up the company to decide how they want to do it
00:17 kados which is the LibLime customer version of koha
00:17 kados 2.2
00:17 [K] <rangi@FreeNode> yep
00:17 kados it's not always going to be identical to the 2.2.x RM clone
00:18 [K] <rangi@FreeNode> which is rel_2_2 is cvs and would be rel_2_2 branch in git
00:18 [K] <rangi@FreeNode> thats right
00:18 [K] <rangi@FreeNode> all commits are done locally
00:18 [K] <rangi@FreeNode> its not until you push (or pull)
00:18 [K] <rangi@FreeNode> that anything remote happens
00:18 [K] <rangi@FreeNode> (not like cvs)
00:18 kados right
00:19 kados so the RM's version is pull only
00:19 [K] <rangi@FreeNode> could be
00:19 [K] <rangi@FreeNode> could be push too
00:19 kados the company's Koha Classic will be push/pull for the company's devs, pull for everyone else
00:19 [K] <rangi@FreeNode> yep
00:19 [K] <rangi@FreeNode> the rm version could be push/pull for trusted devs
00:19 [K] <rangi@FreeNode> pull for everyone else
00:19 kados ahh, cool
00:20 kados so how does QA fit in?
00:20 [K] <rangi@FreeNode> qa needs to be done on the rm versions
00:21 [K] <rangi@FreeNode> so the qa manager will have push/pull to them
00:21 [K] <rangi@FreeNode> and will monitor changes
00:21 [K] <rangi@FreeNode> and qa them
00:21 [K] <rangi@FreeNode> and alert the rm to code that fails qa
00:21 [K] <rangi@FreeNode> then maybe fix them, or just revert the change
00:22 [K] <rangi@FreeNode> depending on the complexity
00:22 kados it'd be nice to have an explicit process for that
00:23 [K] <rangi@FreeNode> how so?
00:23 [K] <rangi@FreeNode> i thought that was pretty explicit
00:23 [K] <rangi@FreeNode> its gonna have to on a case by case basis ... theres no getting round someone eyeballing the code
00:23 kados yea, I just mean, we should have docs that explain the process officially
00:23 [K] <rangi@FreeNode> ah right
00:23 [K] <rangi@FreeNode> yep
00:24 [K] <rangi@FreeNode> the easiest way to get ur new code rejected will be if its not documented, and if there are no tests for it :)
00:25 kados heh
00:25 [K] <rangi@FreeNode> thats if i'm qa anyway :) we have to sort that out at the dev meeting
00:25 [K] <rangi@FreeNode> but i think we should have flexible/evolving rules
00:25 kados yea
00:25 kados definitely
00:25 kados some kind of working document
00:25 [K] <rangi@FreeNode> yep
00:25 kados the wiki would be good for that
00:26 [K] <rangi@FreeNode> the beauty with git is merging between branches is very easy
00:27 [K] <rangi@FreeNode> means if someone wants their own set of templates (a client)
00:27 kados yea
00:27 [K] <rangi@FreeNode> its more managable
00:28 [K] <rangi@FreeNode> and bringing out of line installs into line because doable :)
00:31 [K] <rangi@FreeNode> because=becomes
00:39 kados chris:
00:40 chris yep
00:41 chris just think of the rm clones as linus's clones
00:47 kados so who manages our master repo?
00:47 chris the master repo is more an archive/reference point
00:48 kados does it ever changes/
00:48 kados change rather?
00:48 dewey kados: that doesn't look right
00:48 chris when rm's push to it yep
00:48 kados there's some fundamental piece I'm missing I think :-)
00:49 chris it shouldnt need any management
00:49 kados I need to re-wire my brain for git I think :-)
00:49 chris ok the rms clone from the master (once)
00:50 kados once for every new version?
00:50 chris no once ever
00:50 kados k
00:50 chris or once per release manager :)
00:50 kados heh
00:50 chris developers clone from the rm
00:51 chris work and commit locally push ... or email patches .. to the rm when they want to
00:51 kados right
00:52 chris the rm might push to the main repo above them ... maybe when they do a new release
00:52 chris or they might do it on a weekly basis .. or they might do it randomly :)
00:52 kados and it can store 2.2 and 3.0 at the same time?
00:52 kados (the master)
00:52 chris yeah they will just be branches
00:52 kados ahh
00:53 kados will each minor version be a branch?
00:53 kados s/will/should/
00:53 chris if the rm wants to do it that way yep
00:53 chris could just tag
00:53 kados and tagging would also support minor bug fixes to make 2.2.10 to 2.2.10b
00:53 kados right?
00:54 chris yep
00:55 kados so lets run through a few scenerios
00:55 chris the RM manager has a bit of freedom over how they will run their release
00:55 kados and I'll document this discussion and turn it into a git for dummy RMs :-)
00:55 chris because what they do, doesnt make a mess for the other RM
00:55 kados 'Got for Dummy RMs' :-)
00:56 chris so its really up to them to figure out what is going to work best for them
00:56 kados /Got/Git/
00:56 chris ie they might just want to give a bunch of ppl push access to their clone
00:56 chris or they might be more anal and want all changes as patches
00:56 chris whatever works for them
00:56 kados right, or give some people push access
00:56 chris yep
00:57 chris it'll be another one that will evolve
00:57 chris as well all get more used to git
00:58 chris "The ever changing Git for Dummy RM's"
00:59 chris :)
00:59 chris eg
00:59 chris at liblime, we have kohaclients.git
00:59 chris the master of which lives on denethor
01:00 chris now it contains all the code from the koha clients who came over from katipo
01:00 kados heh
01:00 kados right
01:00 chris i have a clone on arwen
01:00 chris and mason (and now ryan i think) have a clone too
01:01 chris or maybe ryan has a clone of mason
01:01 chris :)
01:01 kados ok, so I'm on arewn
01:01 kados arwen
01:01 chris but mason and I definitely have push rights back to denethor
01:01 chris so i work in my clone
01:02 chris the dev sites run out of mason's clone (symlinked up)
01:02 kados right
01:02 chris i work in mine, commit .. debug test a bit .. then push
01:02 chris then i either get mason to do a pull or i do a pull in his repo and the changes show up on our test site
01:03 chris ideally, we would have another layer in here
01:03 chris denethor -> test sites clone -> chris clone
01:03 chris denethor -> test sites clone -> mason clone
01:03 chris so when i push it shows up on the test site
01:04 chris and when im happy with that, can push it back to denethor
01:04 chris does taht make sense?
01:04 kados yes
01:04 kados very much so
01:04 chris then there might be a new programmer come along
01:04 kados I think if we could explain that process to paul/hdl/toins they would appreciate it
01:04 russel kados: can you please give me privs so i can edit the headers on contribs and translate sites - just want to add text to the headers
01:04 chris lets call her alice
01:05 kados russel: yea, I'll check it out
01:05 chris now she might just be learning
01:05 chris kados: if you make it group writable by group staff
01:05 chris that would do it
01:05 russel cheers
01:05 chris (i put russ in the staff group)
01:05 chris ok back to git
01:05 chris so we have alice, and i want to check her code
01:06 chris before it makes it to the test sites
01:06 kados russel: try now
01:06 chris denethor -> test sites clone -> chris' clone -> alice's clone
01:06 kados right
01:06 kados in this scenerio, alice 'works' for you
01:06 chris so when alice pushes i can eyeball the changes
01:06 kados (say)
01:06 chris and push them further up
01:06 kados right
01:07 chris yeah she might a member of my team
01:07 kados yea, this is making sense
01:07 kados so lets break it down further to the actual commands
01:07 chris and then after 2 months i realise she is brilliant
01:07 kados hehe
01:07 chris and she clones direct from the test sites
01:07 chris and no longer pushes thru me
01:07 kados i figure, we can propose a workflow for the project, given our current size, etc.
01:07 kados and we can change that workflow as needed
01:08 chris yep
01:08 russel kados: nah still the same
01:08 kados russel: hmmm, sec
01:08 russel says it is owned by staff though - interesting
01:08 chris group staff?
01:09 kados well contribs is in rangi's home dir
01:09 chris remind me to fix that at some point
01:09 russel ahh i am in translate
01:10 russel .//kartouche/includes
01:11 chris needs a chmod -R g+w staff
01:11 kados yea, just did that
01:11 chris cool
01:11 kados for includes dir anyway
01:11 kados russel: lemme know if that works
01:11 russel yeah will do two secs
01:12 chris brb
01:12 chris craft night here tonight so i have to hold the baby for 10 mins while laurel vacuums
01:13 kados heh
01:13 russel chris when you get back - can you change the privs on this dir
01:13 russel ./
01:13 russel in /home/rangi/
01:13 kados sec
01:14 kados russel: try that
01:14 kados mason: so whatcha workin on?
01:14 mason ping
01:14 kados hey :-)
01:15 mason yep, just ticking thru the bugs at the mo, commited a fix for BUG-807
01:15 russel kados: super
01:15 kados mason: one thing we need to do is make sure all the changes we added to rel_2_2 the npl templates, also get made to the default templates
01:16 kados mason: you up for looking into that at some point?
01:16 kados mason: I think paul would be willing to release 2.2.10 off of rel_2_2 if we did that
01:16 mason yep, sure thing
01:17 mason ill come up with a plan for testing that
01:17 kados excellent!
01:18 russel is kartouche all in cvs somewhere?
01:19 kados yep
01:19 russel ahh
01:19 kados somewhere :-)
01:19 kados but yea, you can hack on it
01:19 kados it's not our CVS
01:19 russel so i can hack the templates directly?
01:19 kados though the owner said we could take over as he's not got time to manage it
01:19 kados yep
01:19 russel sweet as
01:19 kados some day I'll have enough to hire someone to work on kartouch full time
01:20 kados it's a cool little project, just needs some luving
01:21 russel ok well that is the headers hacked a little - not super pretty but at least it is consistent
01:22 kados sweet russ
01:22 kados look good
01:23 kados guys, have we committed corporate serials to head yet?
01:23 chris who knows anymore :)
01:23 kados heh
01:23 chris someone will have to go look
01:23 kados that shoudl be on the list I think for 3.0
01:24 chris (back btw)
01:24 kados cool, did you do the diff_patch doc?
01:24 chris paul did
01:25 kados and unit testing too?
01:25 kados cool
01:25 chris[…]opment:diff_patch <-- that one?
01:25 kados yea
01:25 chris i shifted the unit testing one, and edited a bit
01:25 chris ive been tidying/shifting things into the namespace en:development
01:26 kados sweet
01:26 chris finished the test script today too
01:26 kados nice
01:26 chris (easy module)
01:26 kados that's great chris!
01:26 kados gonna make our releases solid
01:26 kados I'm gonna write up a summary of our git conversation
01:27 kados then you can double-check I got it right :-)
01:27 russel
01:27 russel not too much an improvment today
01:27 russel but at least you can use the links down the side to jump down the page
01:27 kados yea, looks good russ!
01:27 chris :)
01:28 chris yo russ
01:29 russel yep
01:29 chris the links in the banner bit seem wrong
01:29 russel yeah onto it
01:29 chris sweet
01:33 tnb hi
01:33 dewey what's up, tnb
01:33 tnb ok, just seing if people could see me ;)
01:33 kados hi tnb
01:37 russel kados: what is the easiest way to work in the 2.2 userguide?
01:37 russel there are some empty sections that would be pretty easy to flesh out
01:37 kados stephen wrote that using freemind I think
01:37 kados the xml
01:37 kados then converted it to html
01:38 kados I think we should just throw it all in a plone site
01:38 kados where it could be translated side by side
01:38 russel right
01:39 russel ah XMLMind i think you mean
01:39 russel FreeMind is some kind of Java mindmapping tool
01:40 kados oops
01:40 kados yea
01:40 russel
01:41 chris heya philip
01:41 philip hi
01:42 philip russ around?
01:42 chris he was a minute ago
01:42 russel hiya philip
01:43 philip I'm looking to clear a bit of space on wilbur and noticed you have over a Gig of stuff.
01:43 russel i have copies of that
01:43 russel you can nuke it
01:43 philip excellent.
01:43 philip Is your email still there?
01:44 russel i guess so
01:44 russel well katipo email
01:44 russel ahh or that might be on shelob?
01:48 mason hey chris, u about?
01:48 chris yep
01:49 mason could you change the perms on  /koha/etc/ on arwen?
01:49 mason perhaps make em the same as /nzkoha/etc ?
02:01 russel join #koha
02:01 russel doh
02:01 russel nevermind
02:01 chris heh
02:14 russel kados you still there?
02:15 chris i think he might be writing stuff up
02:15 chris seen kados?
02:15 dewey kados was last seen on #koha 35 minutes and 28 seconds ago, saying: yea [Mon Jun 25 13:40:04 2007]
02:19 kados russel: sorry
02:20 kados chris: still around?
02:20 chris yup
02:20 russel no worries watching a screen cast i'll be a while :-)
02:20 kados chris: wondering if we can pick a method for RMing 3.0
02:20 kados with git
02:20 kados and QA
02:21 chris yep, we could do a proposed method
02:21 chris and take feedback on it
02:21 kados would you want to 'approve' all pushes?
02:21 kados yea, that's waht I mean
02:21 kados proposed method
02:22 chris hmm conversely
02:22 chris id unnaprove some
02:22 kados right
02:22 chris approved is the default state
02:23 chris what my plan would be
02:23 kados then we'd have to just be sure to check every commit
02:23 kados or push
02:23 kados :-)
02:23 chris yeah every push
02:23 chris what id do
02:23 chris is edit the .git/hooks/update
02:24 chris file such that it emails me (and whoever else, maybe a mailing list)
02:24 chris and have gitweb set up
02:24 chris so i can easily look at diffs
02:24 kados right
02:24 chris then if i spot something i dont like the look of, or am unsure about
02:25 kados in that case, everyone can have push on the RM's repo
02:25 chris id mail the pusher :) and/or the devel list
02:25 kados I think anyway
02:25 chris everyone who currently has cvs write access yet
02:25 chris yep even
02:27 chris the RM manager has to wathc the push's from another angle
02:27 chris ie if he/she have called a feature freeze
02:28 chris they will have to watch out for new features etc
02:28 chris ie the QA person is watching for coding practices
02:28 chris RM is at a higher level
02:28 kados right
02:29 chris so something might pass qa just fine, but the RM might not want it in the next release so may decide not to merge it in
02:29 chris i think the key will be being communicative on the mailing list, so ppl can see whats happening and why
02:30 chris (the devel list that is)
02:30 kados yea, that makesw sense
02:31 kados ok, so commands
02:31 chris and keeping this in mind
02:31 chris[…]hp?g2_itemId=6121
02:31 kados step one is to clone the RM
02:31 kados yea, that's a good motto
02:32 chris step one, you will need to give the rm your public ssh key
02:32 chris so they can set you up with access (using the git shell probably, not a full shell)
02:32 chris then you can clone
02:34 kados ok, just sent you mine I think
02:34 kados mailx -s "my ssh key for git" <
02:34 kados (for the denathor one)
02:35 chris right and if i was the RM id use that to set up access for ya
02:35 chris yep, ill do it for that now
02:37 chris k ssh
02:38 chris that work for ya?
02:38 chris if so then
02:39 chris git clone whateveryourwanttocallyourclone
02:39 kados yep
02:40 kados well, the shell works
02:40 kados git clone gave me a error:
02:40 kados ssh_exchange_identification: Connection closed by remote host
02:40 kados fatal: unexpected EOF
02:40 kados fetch-pack from '' failed.
02:41 chris hmm
02:42 chris you are trying that from aule eh? (or another machine)
02:42 kados from arwen
02:43 chris odd
02:43 chris same key on arwen?
02:43 chris ie, thats were you ssh'd from and that worked eh?
02:43 kados yep
02:43 chris i only see the one succesful attemp
02:43 chris t
02:44 kados hmmm
02:45 kados git clone kohaclone
02:45 kados same error
02:45 chris you ran that from arwen just now
02:45 chris ?
02:45 kados yea
02:45 kados maybe it's jmf@
02:45 chris naw the user is kados
02:46 chris im not seeing anythng in the logs on denethor for it
02:46 chris 2 secsc
02:50 chris try
02:50 chris git clone ssh:// kohaclone
02:51 kados ssh: Name or service not known
02:51 kados fatal: unexpected EOF
02:51 kados fetch-pack from 'ssh://' failed.
02:51 mason_ perms on the dest. dir?
02:52 kados shoot
02:52 kados i was on denethor :-)
02:52 kados hehe
02:52 chris that would do it
02:52 chris ok back on arwen
02:52 mason_ yow
02:52 dewey I'm having an EMOTIONAL OUTBURST!!  But, uh, WHY is there a WAFFLE in my PAJAMA POCKET??
02:52 chris the command you ran before should work
02:53 kados yep, seems to be working
02:53 chris cool
02:56 kados ok, great
02:56 kados so there are a bunch of dirs in there, those are branches?
02:58 chris nope those are just dirs :)
02:58 mason_ nah, they are actual directories
02:58 chris type git branch
02:59 chris ok lets do some pretend work
02:59 kados yea, that's master
02:59 kados I see liblime/moo
02:59 kados I'll edit it
02:59 chris 2 secs
02:59 chris first lets do
02:59 chris git branch kados_test
02:59 chris then
03:00 chris git checkout -f kados_test
03:00 chris now you are in the the kados_test branch
03:00 chris we branched from master, so everythign will be exactly the same at this point
03:00 chris now edit a file
03:01 chris and git commit file
03:01 kados k, sec, I'm writing this down as I go
03:01 chris righto, yell out when ur ready
03:03 chris it might grumble at you on arwen when you try to commit you can get round that by exporting the variables its grizzling about
03:03 kados fatal: empty ident  <> not allowed
03:03 chris export GIT_AUTHOR_NAME="whatever"
03:03 chris export GIT_COMMITER_NAME="whatever"
03:04 chris arwen is running an oldish version of git
03:04 chris with the new versions, you can run git config
03:04 chris to set all that up
03:05 chris (arwen is running the stock debian stable one .. figure its handy for us to be familair with that version as well as the lastest version .. which i run on the laptop)
03:05 kados yea
03:05 kados ok, commit worked now
03:05 chris cool
03:05 chris so git log
03:05 chris will show the log, including your commit
03:06 kados sweet
03:06 chris now say we have finished doing our work, and we are pretty happy with it
03:06 chris and we wanted to get pull it back into the master branch
03:07 chris then we could
03:07 kados pull? or push?
03:07 chris checkout -f master
03:07 kados won't that wipe out the changes we made?
03:07 chris nope
03:07 chris they are all still in that branch
03:07 chris now we are in master
03:08 chris we could go
03:08 chris git merge kados_test
03:08 chris (try it out0
03:08 kados k
03:08 chris and it will suck in and merge our changes
03:08 chris with whatever changes have been made in this branch (since we branched .. in this case none :))
03:09 chris and now if we do git log
03:09 kados git merge kados_test
03:09 kados fatal: Needed a single revision
03:09 kados Usage: /usr/bin/git-merge [-n] [--no-commit] [--squash] [-s <strategy>]... <merge-message> <head> <remote>+
03:09 chris ahh old git again
03:10 chris taht would have worked on the newer gits :)
03:10 chris try
03:11 chris git merge master kados_test
03:11 kados Usage: /usr/bin/git-merge [-n] [--no-commit] [--squash] [-s <strategy>]... <merge-message> <head> <remote>+
03:12 chris ahh git merge "Merging in changes" master kados_test
03:12 chris that might be it for this version
03:12 kados sweet
03:14 chris its more sane on the new version :)
03:15 chris ie it will just spark and editor up so you can enter your message
03:15 kados huh
03:15 chris unless you do -m="message"
03:15 kados yea, done
03:15 chris (in the new version)
03:15 kados so shouldn't it have sent a mail?
03:15 chris nope
03:15 chris this is all still locally
03:15 kados still local to this repo?
03:15 chris yep
03:15 kados got it
03:16 chris if you do a git push now
03:16 chris it will send out a mail
03:16 chris this may or may not be how we want to do it, we might want ppl to push their branch
03:16 chris and we merge the changes in
03:17 chris rather than pushing the master branch
03:17 chris does that make sense?
03:18 chris you can do git diff between branches too
03:18 kados right
03:18 kados so which do we want to try first?
03:18 kados giving people push to their branch? or push to the master branch?
03:18 kados I guess the wouldn't have clone ability if they just had push to their branch?
03:19 chris they would still be pushing to the rm's clone
03:19 chris just into a different branch
03:20 kados way I see it
03:20 chris and the rm would have to do the merge .. it would mean they woudl also have to do a git pull on the master
03:20 kados you'd want to create a new branch every time you were working on a new feature
03:20 chris yes
03:20 chris the question is
03:21 chris do we want them to merge to master and push when the feature is done
03:21 chris or do we want to do that merge
03:21 kados yea
03:21 kados I suspect we want to do the merge
03:21 chris yeah im inclined to try it that way first
03:21 kados cool
03:22 kados so that's gonna be slightly different then
03:23 kados they wouldn't do a checkout -f master
03:23 kados right?
03:23 chris thats right
03:23 kados they'd just push the branch to the RM?
03:23 chris yep im not sure of the syntax to do that
03:23 chris lets try
03:23 kados lets try
03:23 kados snap
03:23 kados :-)
03:23 chris checkout ur branch again
03:24 kados git checkout -f kados_test
03:24 kados fatal: Not a git repository
03:24 chris you in the right dir?
03:24 chris you need to be inside your clone
03:24 kados ahh
03:24 kados right
03:25 chris (see how easy its gonna be to merge from rel_3 (when we create the branch when we go stable) to the dev branch)
03:25 kados yea
03:25 kados rock!
03:25 kados ok, so I check it out
03:25 chris much easier to do the backport, forwardport stuff
03:25 chris yep
03:25 kados now I branch?
03:25 chris naw
03:26 chris lets try and push that branch
03:26 kados OK
03:26 chris what happens when you do git push ?
03:27 kados Everything up-to-date
03:27 chris right, lets make a change, commit it
03:27 russel kados you going to be around for much longer?
03:27 chris and then try our push again
03:27 kados russel: 30 mins or so
03:28 kados chris: same deal
03:28 kados I edited a file, did git commit, and git push
03:28 kados Everything up-to-date
03:28 chris gonna have to figure out how we push the branch, ill go have a read
03:29 russel the plone idea sounds good - but i was thinking perhaps we set up a seperate - set up all the content in there and then make the change
03:29 kados russel: I think better to maybe do a :-)
03:29 kados and put docs in
03:29 kados and we can redirect kohadocs when it goes live
03:30 russel ok so well that is something i can punt around on on these koha days
03:30 russel but i will need someone to set it up for me
03:31 kados yea
03:31 kados what I was thinking
03:31 russel i mean i assume that the look and feel is all controlled by templates and can be done later
03:31 kados was we could do a contest
03:31 kados for the design
03:31 russel of what?
03:31 kados say $1,000 or something
03:31 kados koha.orguthe new
03:31 kados the new
03:32 kados we could get plone designers to compete
03:32 kados it'd be good for both projects
03:32 russel well the is cool
03:32 russel but the look and feel
03:32 russel is seperate from the content
03:33 russel and i can start doing the content
03:33 kados yea
03:33 russel a contest will take a while to run
03:33 russel and if this does what it says it does
03:33 kados yea, we can do the content separately
03:33 chris gonna have to have a read figure out the best way to do it
03:33 russel then i should be able to start on the content now
03:33 kados chris: sounds good
03:33 kados russel: exactly
03:33 kados russel: ok, I won't get to it today, but I'll have it ready for you by next koha project day
03:34 russel sure the homepage layout etc will be different - but there is a bunch of content that needs working on
03:34 kados I imagine the first job is tjust to get the organization going
03:34 kados copy/paste what docs we have into that organization
03:34 kados s/organization/taxonomy/
03:34 kados yea
03:35 russel yep
03:46 chris cool
03:56 kados chris: ok, I'm done up to the part about how to push branches :-)
03:57 chris im gonna have to play, ill email what i find
04:03 kados chris:[…]lopment:git_usage
04:03 kados I'm gonna get to bed
04:04 chris yep busy day tomorrow
04:04 chris ill edit it with whatever i find
04:04 mason_ yep, cya later josh
04:04 kados mason_: b4 I go, how's the hunting going?
04:04 kados finding the stuff paul wanted committed to default templates?
04:05 mason_ i hadnt switched over to that yet, was still picking thru bugs
04:06 mason_ is it urgent-er?
04:06 kados I guess in some sense it _is_ a bug :-)
04:07 kados I think it's the only thing holding back a 2.2.10 release based on rel_2_2
04:07 kados so maybe sorta urgent
04:07 kados if you're int he middle of a bug fix
04:07 kados I'd finish that first
04:07 mason_ yeah, that sounds urgenter that the bugs ive been looking at today, for sure
04:07 kados then maybe move on
04:07 kados anyway, I should get some rest
04:07 kados ttyl
04:08 mason_ yep, great work at ALA too
05:08 russel[…]w_bug.cgi?id=1360
05:09 russel pity it was a simple bug
05:09 chris nice proof of concept
05:09 russel but i think even reading it is useful
05:09 russel[…]chment.cgi?id=109
05:09 chris yep
05:10 russel i was trying to find a bug that i could use it for
05:10 chris you dont miss steps that way
05:10 russel yeah
05:10 russel i have recorded a bunch of tests
05:10 russel but not sure what to do with them now i have them
05:11 chris we have to get our automated tester going
05:11 chris so it can run those tests
05:11 russel yeah - but it would be nice to have a place to start collecting these
05:11 chris we should put them in git
05:12 russel i might do some more reading
05:13 russel about relative vs absolute links
05:13 russel so people can run them against their own installs
05:13 russel i think there must be some way that someone can write the tests on one site, but use them on another
05:13 chris right that one is all relative
05:14 chris so shoudl work on any koha
05:14 chris ie it says
05:14 russel ah true it does too
05:14 chris no hostname .. so should work anywhere
05:14 russel <td>/cgi-bin/koha/</td>
05:15 chris yep
05:21 chris ok im done for the day i think
05:21 chris ppl will be arriving for craft night soon, so i best go help
05:26 russel yep i am done too
05:26 russel laters
07:15 hdl chris around ?
09:21 chris am now hdl
09:21 hdl that late ???
09:21 chris its only 9.18pm
09:21 chris be awake for a little bit longer
09:22 hdl was just to ask you if you investiguated perl modules for selenium ?
09:22 chris not really
09:24 chris we were planning to use selenium server
09:26 chris sorry twisted server
09:27 chris but i actually i think doing it with perl scripts might be nicer
09:30 chris hmm
09:34 chris with the IDE you can export as perl
09:37 chris i just did one now
09:38 chris $sel->open_ok("/cgi-bin/koha/");
09:38 chris $sel->click_ok("//img[\@alt='KohaZOOM  Enterprise ILS']");
09:38 chris sel->wait_for_page_to_load_ok("30000");
09:38 chris $sel->type_ok("q", "fish");
09:39 chris very cool, the selenium ide will write your perl test for you, then you can just run them
09:47 chris so i have my test (which is perl) and im just getting a selenium server up and running so i can run against it
09:48 chris but even with recording it in html, its useful
09:48 chris for example[…]chment.cgi?id=109
09:52 chris wow cool
09:53 chris i installed this
09:53 chris and got the server up and going
09:54 chris and now i can run the tests ive recorded with the ide
09:57 chris oh wow cool
09:58 chris i might have to do some screen shots
10:06 chris but now i migt have to go to sleep :)
11:42 kados g'morning #koha
11:55 paul hello kados

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