IRC log for #koha, 2013-11-28

All times shown according to UTC.

Time S Nick Message
00:04 eythian All the strong passwordness in the world isn't going to do much if you have poor security everywhere else.
00:15 pianohacker dcook: around?
00:18 dcook pianohacker: Ish
00:18 dcook What's up?
00:18 pianohacker dcook: Fair enough :) Updated bug 10486, thought you should know
00:18 huginn Bug[…]_bug.cgi?id=10486 new feature, P5 - low, ---, jweaver, Needs Signoff , Allow external Z39.50 targets to be searched from the OPAC
00:19 mtompset Greetings, the person asking to be removed claims her email was hacked, and she doesn't know the password.
00:20 rangi should could send an unsubscribe message to the koha-request address, but ill just unusb her before she yells again
00:21 rangi gone
00:21 wizzyrea jeez
00:22 dcook pianohacker: Sweet. Thanks for the head's up, pianohacker :).
00:22 pianohacker absolutely
00:22 rangi and if thats the case, she really should be cancelling a bunch of things .. i do find it unlikely someone would hack you and sign you up to a koha list .. but hey, stranger things happen
00:22 eythian in fact, you don't need a password to unsub.
00:22 eythian it sends you a link to click.
00:22 rangi thats what i thought
00:22 eythian so that's probably a lie.
00:22 jcamins rangi: really?
00:22 jcamins Such as?
00:23 rangi sign you up to autocat
00:23 jcamins lol
00:23 rangi :)
00:23 jcamins Touché.
00:23 dcook pianohacker: I have a bit of a back log that I'm trying to get through, but definitely keen to try it out
00:23 wizzyrea though we did have a suspicion that someone is subscribing people without their consent. :P
00:24 rangi yeah but they do need to reply, or click on the link in he email
00:24 rangi so yuo would have to control their email to do that
00:24 rangi and if so, being on a mailing list is the least of your problems
00:24 wizzyrea yea, it's far fetched.
00:24 pianohacker dcook: I understand, just trying to beat my record for bug 10320
00:25 huginn Bug[…]_bug.cgi?id=10320 enhancement, P4, ---, jweaver, Pushed to Master , Show results from library's OverDrive collection in OPAC search
00:25 pianohacker I think I can get 10 signoffs on this one before it gets pushed
00:25 pianohacker gogogo
00:25 wizzyrea hehe
00:26 dcook hehe
00:28 maximep left #koha
01:44 mtompset okay, I have a git question.
01:44 mtompset I've been doing all this testing and patching to koha, and I'd like to push my branch to a remote repository. How?
01:45 eythian git push <reponame> <branchname>
01:45 mtompset where reponame is a URL?
01:46 eythian no, where reponame is the name of the repo. This is managed using git remote
01:46 mtompset So, how do I tell git about this remote repo?
01:46 eythian git remote add [-t <branch>] [-m <master>] [-f] [--tags|--no-tags] [--mirror=<fetch|push>] <name> <url>
01:47 eythian the important parts being the last two
01:53 dcook git remote -v will print out the list of your remotes
01:53 dcook Well git remote will too
01:53 dcook the -v adds the url
02:15 mtompset Well, I'm hoping this combination of git remote and git push I did works. It does look like it is. :)
02:15 mtompset Just so slow to upload. ARG. :(
02:20 eythian is it a new repo?
02:20 eythian because if you're uploading koha to it, it will take a while
02:20 dcook ^^
02:22 papa joined #koha
02:22 eythian hi papa
02:37 mtompset It is a new repo.
02:37 mtompset Greetings, papa.
02:37 eythian then you are uploading a few hundred MB or so.
02:38 mtompset Is there a way to do a link text for 952$u, not just 856$u?
02:38 mtompset Yes, yes I am. Slow uploads. *sigh*
02:39 mtompset is there a way to list branches on a remote repo?
02:39 dcook mtompset: Not by default, me thinks, but we've used other fields for that sort of thing before (with a local mod)
02:39 * dcook is referring to the link text btw
02:40 mtompset That's what I was thinking (re: link text)
02:42 eythian mtompset: I'm not sure that's a meaningful question. What are you trying to do?
02:42 mtompset 856$u has a 856$y to provide pretty link text.
02:43 mtompset Is there a similar 952 subfield which displays text instead of the super-long URL?
02:43 eythian no, your git question
02:43 mtompset oh, git question, sorry.
02:43 eythian (also, in answer to your 952$u question, I think the answer is currently "no.")
02:43 mtompset How do I know what branches a remote repo has?
02:44 eythian well, you have them all yourself.
02:44 eythian so git branch -a
02:45 eythian they should have the repo name in them
02:47 mtompset ah... -a, okay. That's what I was looking for. :)
02:48 mtompset I'm no longer a git push virgin. :)
02:48 mtompset And from the looks of it, it worked as expected. :)
02:49 mtompset can you "git branch -D remote/.../branchname"?
02:53 eythian Probably, but I doubt that it'd do what you'd expect
02:53 mtompset Hmmm...
02:53 eythian you're looking at a clone of the remote.
02:53 eythian so everything is actually local
02:53 mtompset okay, something not working as I expected.
02:53 dcook Ho ho, I think I found a genuine bug
02:53 mtompset as compared to a fake one? :P
02:54 mtompset I have my nice kohaclone directory.
02:54 mtompset I did the git remote.
02:54 dcook Sometimes, something seems like a bug when it's not :p
02:54 mtompset I did the git push.
02:54 mtompset I figured, "Hey, let's try to clone the sucker," so I went to a different non-git directory and did a git clone command.
02:55 mtompset And it looked like it was working until I got this message.
02:55 mtompset warning: remote HEAD refers to a nonexistent ref, unable to checkout.
02:55 wizzyrea mmm not a bug.
02:56 wizzyrea probably means your push didn't work as you expected.
02:56 dcook What did your git push command look like, mtompset?
02:56 dcook I think it needs to be something like... git push branch:remote
02:56 dcook Or vice versa...
02:56 * dcook hasn't pushed in a while
02:57 eythian git push origin master
02:57 eythian for example
02:57 mtompset AH! I think I know why...
02:57 mtompset the git config says origin, but I used twkoha.
02:57 eythian what does git remote -v say
02:57 eythian that's a goot canonical source
02:57 eythian *good
02:58 mtompset my attempted clone origin, my kohaclone git says twkoha
02:59 mtompset The mismatch is the problem. :)
02:59 mtompset That's my guess.
02:59 eythian they can have different names
02:59 eythian the name is just what your repo is calling that remote
03:00 mtompset oh...
03:00 mtompset Perhaps my URL was bad.
03:06 mtompset I'll try bugging my colleague for some help, since he has been dealing with our git server. Thanks for the assistance thus far.
03:08 dcook Hmm, anonymous purchase suggestions don't appear in the staff client, because they have an undefined branchcode...
03:09 dcook They really should just use the AnonymousPatron's homelibrary..
03:09 dcook Maybe
03:09 dcook Anonymous suggestions from logged in people work, but not non-logged in folk..
03:10 eythian or it should show things with an undefined branchcode
03:11 dcook That's my other thought
03:11 dcook I wonder how much the branchcode makes sense overall anyway..
03:11 dcook It says where the patron is from but not which library branch they want to buy it
03:12 dcook Although some systems do central purchasing so that wouldn't matter (except in terms of a pick up branch which is another syspref all together)
03:12 eythian some do, some done.
03:12 eythian don't
03:12 dcook Indeed
03:12 dcook Classic Koha :p
03:13 eythian well, more that different libraries work in different ways.
03:13 dcook Yes. What I mean is that we need to be many things to many people.
03:14 dcook Oooo
03:14 dcook the branchcode is also used for their desired pickup location
03:14 wahanui okay, dcook.
03:14 dcook Bah
03:14 dcook wahanui: forget the branchcode
03:14 wahanui dcook: I forgot branchcode
03:15 dcook But that one isn't showing up in the UI either..
03:16 dcook Must be locked down by branchcode...even for superlibrarians
03:16 dcook Yep..
03:16 eythian there's also a default filter that is quite broken I think
03:16 dcook me thinks that might be it
03:18 wizzyrea yes, the purchase suggestion filtering is quite broken. especially in 3.8-3.10
03:18 wizzyrea possibly even into 3.12 - but I think there was a patch for that sometime in 3.10
03:18 dcook Any open bug reports on that one? I didn't see any when I went looking
03:18 dcook 3.14 as well it seems
03:19 * dcook is looking at master atm
03:19 wizzyrea yea, there was one, it had a strange name
03:20 wizzyrea bug 10072
03:20 huginn Bug[…]_bug.cgi?id=10072 critical, P5 - low, ---, kyle.m.hall, Pushed to Stable , "my suggestions" always says "There are no pending purchase suggestions."
03:21 wizzyrea bug 8147
03:21 huginn Bug[…]w_bug.cgi?id=8147 normal, P5 - low, ---, henridamien, NEW , suggestions management forgets branch info
03:21 wizzyrea bug 10449
03:22 huginn Bug[…]_bug.cgi?id=10449 minor, P5 - low, ---, fridolyn.somers, Failed QA , Suggestions pending approval should count only suggestions of current branch
03:22 wizzyrea bug 10519
03:22 dcook Ah, wait a tic...
03:22 huginn Bug[…]_bug.cgi?id=10519 normal, P5 - low, ---, katrin.fischer, Failed QA , Suggestions: 'Organize by' and correct display of tab descriptions broken
03:22 dcook It could be that I'm just a dumbass
03:23 wizzyrea mmm I doubt it
03:23 wizzyrea there is weird stuff going on in the suggestions interface, as you can see.
03:23 dcook Acquisition information has a default library filter for that staff member
03:23 dcook You can change it to any and then you see them all
03:23 dcook But yeah, that second one would be interesting to see
03:25 dcook Hmm, I see that second one now. That IS irritating..
03:27 dcook It would almost be better to get rid of the "Acquisition Information" bit and use DataTables more extensively
03:27 dcook Like in "Holds to Pull"
03:30 dcook Hmm
03:30 dcook wizzyrea++
03:30 dcook Awesome bugzilla searching skillz
03:30 wizzyrea pft
03:30 wizzyrea I had looked at that before ;)
03:31 wizzyrea (suggestions being wonky)
03:32 dcook Too modest :p
04:23 dpk1 joined #koha
04:28 mtompset okay, continuing with the git questions...
04:29 mtompset git branch -a on the default koha repository shows remotes/origin/HEAD -> ...
04:29 mtompset but on what I just attempted with my push, there is no such thing when I clone it.
04:29 eythian I don't think it's a requirement that there is a HEAD
04:30 eythian but this is beyond what I've really dealt with, so I'm no expert.
04:30 mtompset but when I created a dummy project and did all the pushing, there was.
04:30 eythian tried the push again in case something odd happened the first time?
04:32 jcamins mtompset: what is the problem you're trying to solve?
04:32 mtompset I'm not sure.
04:32 mtompset I'm still a git novice.
04:53 dcook mtompset: Iirc, you can't push to a branch if it is currently checked out
04:53 mtompset to or from?
04:53 dcook To
04:53 dcook If you're using git on machine X with branch X
04:53 dcook You can't push to machine Y if branch X is checked out
04:54 dcook You would have to have machine Y on branch Y
04:54 eythian that is familiar, but I'd expect it to warn you in that case.
04:54 dcook Indeed, it should show a warning
04:54 * dcook wonders how Bywater manages its prod systems in that respect..
04:56 eythian they probably don't push to them, but pull onto them.
04:56 mtompset okay: remote has b1, local has b1, and b2. How do I push local b2 to remote b1?
04:56 yhager joined #koha
04:57 dcook eythian: Ah, right. Good one.
04:57 dcook mtompset: Good question. I don't know if you can?
04:57 dcook Might be a special refspec..
04:58 dcook I think the branch names have to match
04:58 dcook Oh wait..
04:58 dcook git push remote b2:b1?
04:59 eythian no
04:59 eythian oh
04:59 eythian yeah, I never do that
04:59 dcook me either
05:00 eythian I always make sure the names match, everything's merged, and them push.
05:00 mtompset Okay, so how does your work flow go?
05:00 mtompset remote b1 ... branch it to local b1.
05:00 mtompset make changes, push it back up?
05:00 eythian so I keep a local branch b1 that is up to date with origin/b1.
05:01 eythian I might fork that to b1_bugfix, but when I'm finished with b1_bugfix, I merge it into b1 and push that to origin/b1
05:01 dcook git push origin HEAD:master
05:01 dcook Push the current branch to the remote ref matching master in the origin repository.
05:01 eythian I think that in general that's a bad thing to do
05:02 dcook I do the same thing as eythian, and it works pretty well
05:02 eythian same reason that rebasing should be treated with caution.
05:02 eythian it's breaking the history of master
05:02 * dcook 's interest is piqued
05:03 eythian in fact, it'll probably require a force-push, not a regular one.
05:03 eythian Well, just that remote branches should form a chain. If you move the off to some other refspec, they no longer do.
05:03 dcook Why would rebasing be bad?
05:04 eythian if I have a branch checked out, and you rebase it, then our branches are totally different.
05:04 eythian but they have the same name
05:04 eythian so if I try to merge or something with your updates, everything will go all messy
05:05 dcook Hmm, I know that they would have a different sha1 hash, but shouldn't the history be the same?
05:05 dcook Since your work is replayed overtop of what you fetched?
05:09 dcook Perhaps I should get one of those git guis to visualize it better someday
05:09 eythian the hash is how everything is referenced
05:09 eythian so the patches would be the same, but the parent of my work is no longer in the branch
05:11 dcook Wouldn't you have a new parent for your work after the rebase?
05:11 dcook Admittedly, I'm not following, but that's all right.
05:11 eythian no, because you did the rebase, I didn't.
05:11 eythian So you can't rebase my work.
05:12 eythian so git can't know that the new top of the branch is almost-but-not-quite the same as mine, as the sha is different.
05:12 eythian so it will become unhappy
05:13 dcook O_o
05:13 dcook unhappy in what way?
05:13 eythian Not sure. Probably a lot of conflicts.
05:14 dcook Conflicts when?
05:14 eythian it'll also depend on what you change by doing the rebasing.
05:14 dcook Mmm, I think I might understand what you're saying now
05:14 eythian if I tried to merge your work into my branch
05:14 dcook It would be nasty
05:14 dcook If we touched anywhere near the same code
05:15 dcook Wouldn't it depend on how often we/I was rebasing?
05:15 papa joined #koha
05:15 eythian even then, it'd try to merge patch 12345 on a branch that already has patch abcde, but they're the same thing. So it can't find the bit it's supposed to change.
05:15 eythian No
05:15 eythian I think
05:15 eythian It's something I've always avoided
05:15 dcook What do you do?
05:16 dcook git pull vs. git fetch + git rebase?
05:16 eythian Depends, if you have one or two patches, reset the new branch head and cherry-pick
05:16 eythian *reset to the new...
05:17 eythian if you have more, I'd make a rebase that would take the patches off the old head and put them on top of the new one (but not rewriting history.)
05:17 dcook In which case you'll possibly have a lot of conflicts
05:18 dcook But it might be easier to resolve those conflicts than cherry-pick lots of patches
05:19 dcook Doesn't rebasing only rewrite history if you do it interactively?
05:19 dcook Actually, now that I think about it, I'm not sure how you'd take a bunch of patches off the old head and put it on top of the new one...
05:20 dcook Except by cherry-picking a range of commits, which I think is usually frowned upon..
05:21 dcook Oh wait...
05:21 dcook rebase --onto...
05:22 dcook Hmm, I think I'm more confused in the end, but I'll stop antagonizing you, eythian :p
05:24 eythian yeah, you can rebase onto. And how many conflict you get depends on how far back history gets rewritten.
05:25 dcook Yeah, that's what I was seeing just now :/
05:25 dcook So how would you not rewrite history?
05:25 eythian use merge instead of rebase, generally.
05:26 dcook Mmm, I thought that might be the case
05:27 dcook Following the wiki instructions, I've always done a fetch + rebase for my patches
05:27 dcook As far as I can tell, they've always gone on top and not rewritten history though
05:27 dcook Mind you, I suppose we wouldn't notice because we post to bugzilla rather than pushing directly
05:28 eythian yeah, that's quite OK
05:28 eythian in fact, for that case, it's recommended as it keeps your patches at the top
05:29 eythian it's only an issue if you're sharing branches with other people, rather than patches.
05:29 dcook Interesting side note: in 3.8.0, if you have a first overdue notice trigger without a delay, none of your second or third notices get processed...
05:29 eythian that seems less than ideal.
05:30 dcook Agreed. I might have to look into that...
05:30 eythian you shouldn't be running 3.8.0 unless you've backported the security patches to it.
05:30 eythian damn, I just realised I wrote my scripts in the wrong order. I wrote one to process the converted data before writing the one to convert the data. So now I have no converted data to test it on.
05:30 dcook I have backported the security patches, but...I don't think we should be running 3.8.0 anyway
05:30 dcook Definitely behind the times..
05:31 dcook Dinger. That sucks :(
05:31 eythian yeah. even upgrading within the 3.8 series would be preferable.
05:31 dcook Agreed.
05:31 eythian it's not the end of the world, but I wanted to test my script :)
05:31 dcook hehe
05:31 dcook I know the feeling
05:31 dcook Yeah, 3.8.5 seems pretty nice actually
05:31 dcook Too many customizations for that though, I think
05:32 dcook Death by a couple hundred Koha localizations
05:32 eythian oh well, I might brave the rain now and do the first half tomorrow.
05:32 eythian that could be a problem.
05:32 dcook Beer o'clock?
05:32 wahanui i guess Beer o'clock is a vital part of the Koha development process.
05:32 eythian it's only thursday :)
05:33 eythian (so that's a maybe ;)
05:33 dcook hehe
05:33 dcook Hope the rain isn't too bad in any case
05:34 cait joined #koha
05:35 eythian @wunder nzwn
05:35 huginn eythian: The current temperature in Wellington, New Zealand is 14.0°C (6:00 PM NZDT on November 28, 2013). Conditions: Light Rain. Humidity: 94%. Dew Point: 13.0°C. Pressure: 29.44 in 997 hPa (Steady).
05:35 eythian it says light rain, but there's a fair bit forecast for the next couple of hours
05:36 dcook Hmm. Hurry home?
05:36 dcook I always seem to get caught in the worst rain here
05:36 dcook As soon as I'm home and drenched, it clears
05:36 dcook @wunder sydney, australia
05:36 huginn dcook: The current temperature in Sydney, New South Wales is 27.0°C (4:00 PM EST on November 28, 2013). Conditions: Clear. Humidity: 45%. Dew Point: 14.0°C. Pressure: 29.62 in 1003 hPa (Steady).
05:36 eythian heh
05:36 dcook Hmm, I'm not sure how reliable that is
05:36 * dcook looks at the slightly overcast sky
05:36 eythian oh well, lets see what happens. I don't expect I'll drown :)
05:36 dcook If you do, it's been nice knowing you
05:37 dcook But hopefully I'll see you back around
05:37 dcook You have another script to write :p
05:38 eythian it's true :)
05:38 eythian later
05:38 dcook ta
05:40 dcook Aha! delay1 totally makes the sql select fail if it's null
05:41 dcook I guess why would you have 2nd and 3rd level triggers if you don't have the 1st set?
05:47 cait dcook: yep
05:47 cait but it could fail nicer
05:47 cait and tell you
05:47 dcook Mmm, agreed
05:48 * dcook will write up a quick report
05:51 dcook bug 11318
05:51 huginn Bug[…]_bug.cgi?id=11318 minor, P5 - low, ---, koha-bugs, NEW , Overdues won't send if first trigger lacks a delay value
05:51 dcook For the record if nothing else
05:54 cait :)
05:56 mtompset Have a great day. I need to get sleep.
05:56 mtompset Bye, #koha.
05:57 busla joined #koha
06:00 dcook Time to go get the cat from the dentist!
06:00 dcook Ciao
06:47 laurence joined #koha
06:56 yhager joined #koha
07:24 cait joined #koha
07:39 paul_p joined #koha
07:42 reiveune joined #koha
07:43 reiveune hello
07:49 cait hi reiveune
07:49 cait bbl
07:49 cait left #koha
07:50 sophie_m joined #koha
07:55 alex_a joined #koha
07:56 alex_a bonjour
08:05 gaetan_B joined #koha
08:06 gaetan_B hello
08:14 dpavlin joined #koha
08:48 cait joined #koha
09:05 ashimema joined #koha
09:28 marcelr joined #koha
09:28 marcelr hi #koha
09:33 cait hi mar
09:33 cait hi marcelr
09:33 marcelr hi cait
09:33 marcelr did you give up on kf?
09:35 ashimema morning #koha
09:37 cait marcelr: someone else is using it - it was too short to register :(
09:37 kf ah
09:37 kf maybe somene gave up on using it :)
09:37 marcelr hey kf
09:37 marcelr well, register it now !
09:37 marcelr oh that is too short
09:37 kf marcelr: it gave me an error in the past about the nick being too short, but i will try
09:38 marcelr too bad
09:38 marcelr i registered marcelr
09:38 kf "The nick you are trying to register is invalid, please choose another nick to..."
09:39 marcelr not nice
09:40 kf nope
09:40 kf discriminating short nicks
09:40 kf kfischer is already registered and not by me it seems
09:40 kf katrinf looks not right somehow
09:42 cait t
09:53 paul_p joined #koha
10:26 ashimema cait is friendly.. we like cait..
10:26 ashimema dunno about that kf person.. seems scarier to me ;)
11:14 gaetan_B joined #koha
12:17 cait ashimema: :P
12:23 Barrc1 joined #koha
12:35 ashimema :)
13:13 francharb joined #koha
13:37 francharb Good morning #Koha
13:57 cait hi francharb :)
13:57 francharb hi cait
13:57 francharb ;)
14:26 ashimema joined #koha
14:34 ebegin joined #koha
14:50 mtompset joined #koha
14:51 mtompset Greetings, #koha.
15:06 maximep joined #koha
15:33 ebegin joined #koha
16:09 ashimema Are the OPACMainUserBlockMobile and other Mobile specific sysprefs used at all whne using the bootstrap theme?
16:11 dpk1 joined #koha
16:12 ashimema joined #koha
16:21 ashimema joined #koha
16:34 yhager joined #koha
16:39 Barrc1 left #koha
16:42 reiveune bye
16:42 reiveune left #koha
16:54 cait left #koha
17:01 wajasu joined #koha
17:02 laurence left #koha
17:14 gmcharlt @quote random
17:14 huginn gmcharlt: Quote #43: "< Ropuch> Uh - beware of tea: it can burn your throat" (added by chris at 08:35 AM, October 30, 2009)
17:22 yhager joined #koha
17:24 gaetan_B bye
18:17 rangi @quote random
18:17 huginn rangi: Quote #166: "jcamins: this is IRC. There's a several decade long tradition of lurking." (added by wizzyrea at 03:06 PM, November 17, 2011)
18:17 wahanui i already had it that way, huginn.
18:19 rangi @quote random
18:19 huginn rangi: Quote #195: "jcamins: libsysguy's test plans all involve sandwich-eating. I think he only programs when hungry." (added by wizzyrea at 06:12 PM, March 30, 2012)
18:19 rangi its a valid development method
18:19 wajasu cookie eating as well
18:19 rangi joubu++ #good testing of the forgot password bug
18:19 rangi wajasu: it's true
18:19 wahanui I read it on the internet!
18:19 rangi heh
18:21 blou joined #koha
18:23 wajasu when attempting to change xslt template for staff that show wrapped text with pip chars between, I wonder if I should supply an m880 template implementation that take a "list" param so it one can choose to emit an html list.
18:25 wajasu with that last release are we technically plack enabled? or is that still work in progress?
18:28 rangi the opac is plack safe
18:28 rangi the intranet still has a few little gotchas
18:29 rangi basically we just need some people running the staff side under plack and testing most everything :)
18:29 rangi before i would feel good about saying "you should run under plack"
18:30 rangi however the public side is fine under plack, thats in production in a few places
18:31 rangi wajasu: that might be quite a nice feature, the html list one
19:06 yhager_ joined #koha
19:09 ebegin joined #koha
21:22 eythian hi
21:39 wizzyrea sup
21:51 trea o/
22:11 papa joined #koha
22:11 maximep left #koha
22:17 francharb see ya
22:24 busla joined #koha
22:31 cait joined #koha
22:37 kathryn joined #koha
23:00 sophie_m left #koha
23:23 cait left #koha
23:38 ebegin joined #koha

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