2014-02-03 GnuCash IRC logs
00:15:33 *** LiamH has joined #gnucash
00:27:22 *** LiamH has quit IRC
01:19:43 *** MechtiIde has joined #gnucash
02:07:23 *** ErKa has quit IRC
02:10:26 *** MechtiIde has quit IRC
02:10:26 *** reaper has quit IRC
02:10:26 *** warlord has quit IRC
02:14:36 *** warlord has joined #gnucash
02:14:36 *** reaper has joined #gnucash
02:14:36 *** irc.acc.umu.se sets mode: +o warlord
02:14:36 *** gncbot sets mode: +o warlord
02:15:32 *** yuriks_ has joined #gnucash
02:18:04 *** fell_ has joined #gnucash
02:18:05 *** gncbot sets mode: +o fell_
02:18:21 *** yuriks has quit IRC
02:22:45 *** vantage has quit IRC
02:23:05 *** fell_ has quit IRC
02:23:38 *** fell_ has joined #gnucash
02:23:38 *** gncbot sets mode: +o fell_
02:23:43 *** vantage has joined #gnucash
02:24:33 *** wiggin has quit IRC
02:24:35 *** wiggin has joined #gnucash
02:26:29 *** fell has quit IRC
02:28:34 *** kim27 has quit IRC
02:31:18 *** fell_ is now known as fell
02:32:07 *** kim27 has joined #gnucash
02:36:13 *** GabrieleV has quit IRC
02:37:03 *** GabrieleV has joined #gnucash
02:40:19 <fell> @tell warlord logs show "Zukünftige" as "Zuk�nftige" (the usual ISO- vs utf8 charset issue)
02:40:19 <gncbot> fell: The operation succeeded.
02:54:07 *** kim27 has quit IRC
02:54:07 *** reaper has quit IRC
02:54:07 *** warlord has quit IRC
02:54:39 *** warlord has joined #gnucash
02:54:39 *** kim27 has joined #gnucash
02:54:39 *** reaper has joined #gnucash
02:54:39 *** irc.acc.umu.se sets mode: +o warlord
02:54:40 *** gncbot sets mode: +o warlord
03:02:44 *** yuriks_ has quit IRC
03:05:45 *** gjanssens has joined #gnucash
03:05:47 *** gncbot sets mode: +o gjanssens
03:05:59 *** yuriks has joined #gnucash
03:27:52 *** MechtiIde has joined #gnucash
03:48:44 *** benoitg has quit IRC
03:49:37 *** benoitg has joined #gnucash
04:20:22 *** jralls has quit IRC
04:34:02 *** xnox has quit IRC
04:35:10 *** xnox has joined #gnucash
04:51:56 *** aqua___ has joined #gnucash
05:28:33 *** aqua___ has quit IRC
05:52:14 *** aqua___ has joined #gnucash
07:49:30 *** gncbot has joined #gnucash
08:00:10 *** warlord has joined #gnucash
08:00:30 <warlord> fell, gjanssens: could one of you /op gncbot please?
08:00:30 <gncbot> warlord: Sent 5 hours and 20 minutes ago: <fell> logs show Zukünftige as Zuk�nftige (the usual ISO- vs utf8 charset issue)
08:00:46 <warlord> fell: which logs?
08:00:52 *** fell sets mode: +o gncbot
08:00:56 <warlord> @op
08:00:57 *** gncbot sets mode: +o warlord
08:01:24 <fell> http://lists.gnucash.org/logs/2014/02/2014-02-03.html#T02:40:19
08:03:54 <warlord> fell: that URL displays correctly for me. Where in the logs do you see the wrong encoding?
08:03:59 <fell> sorry, that is now displayed right. http://lists.gnucash.org/logs/2014/02/2014-02-02.html#T11:27:46
08:04:59 <warlord> fell: that's because it IS an ISO character
08:05:57 <fell> while here in my Konversation window it was ü
08:06:35 <warlord> According to od -c from the logfile, it's 374
08:07:48 <warlord> So maybe your Konversation converted it from ISO?
08:08:18 <fell> May be
08:09:08 <warlord> Strangely my xchat log says it's 303 274
08:10:20 <warlord> Then the one YOU sent is even different.. It has 303 274 and then 357 277 275
08:10:33 <warlord> VERY strange.
08:11:56 <warlord> Anyways, just wanted to let you know I had a power hit this morning... and an ISP outage... simultaneously. Power came back; Internet came back an hour+ later.
08:12:18 * warlord needs to run again..
08:12:30 <fell> OK
08:21:35 <mikee> @op
08:21:35 <gncbot> mikee: Error: You don't have the #gnucash,op capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified.
08:21:41 <mikee> @op
08:21:41 *** gncbot sets mode: +o mikee
08:32:13 *** MechtiIde has quit IRC
08:38:42 *** pc_new has joined #gnucash
08:39:32 *** aqua___ has quit IRC
08:44:06 *** MechtiIde has joined #gnucash
09:06:55 *** aqua___ has joined #gnucash
09:13:16 *** benoitg has joined #gnucash
09:20:50 *** himaxx has joined #gnucash
09:21:32 *** himaxx has quit IRC
10:25:55 *** ErKa has joined #gnucash
10:35:16 *** kpreid has quit IRC
10:36:21 *** kpreid has joined #gnucash
10:39:54 *** aqua___ has quit IRC
10:40:56 *** gwb has joined #gnucash
10:50:56 *** gwb has quit IRC
10:53:28 *** MechtiIde has quit IRC
11:10:30 <warlord> back
11:15:46 *** benoitg1 has joined #gnucash
11:15:47 *** benoitg has quit IRC
11:17:49 <warlord> fell: Have you had the chance to try out the new wiki features?
11:19:09 *** benoitg1 has quit IRC
11:21:06 <warlord> fell: I was considering deleting all user accounts where user_email_authenticated is NULL, but that includes accounts like, oh, WikiSysop, Jsled, and possibly other "early" accounts.
11:22:12 *** benoitg has joined #gnucash
11:25:19 *** benoitg has joined #gnucash
11:27:46 <warlord> Hmm, looks like if I search on user_email_authenticated IS NULL and user_touched < 20140202010000 then I get a more reasonable list... and it will still knock out 16669 accounts!
11:40:07 *** ErKa has quit IRC
12:04:36 *** fell has quit IRC
12:05:13 *** fell_ has joined #gnucash
12:05:14 *** gncbot sets mode: +o fell_
12:11:45 *** ErKa has joined #gnucash
12:11:46 *** fell_ has quit IRC
12:21:18 *** benoitg has quit IRC
12:24:26 *** jralls has joined #gnucash
12:24:27 *** gncbot sets mode: +o jralls
12:26:40 *** fell_ has joined #gnucash
12:26:41 *** gncbot sets mode: +o fell_
12:27:28 <gjanssens> warlord the gnucash-htdocs-docs repo seems to work fine now
12:27:50 <gjanssens> I can commit to it and the changes show up on the website under www.gnucash.org/docs/
12:28:33 <gjanssens> The only thing left to do now is to replace htdocs itself with a repo that no longer carries docs' history
12:29:10 <gjanssens> If ok with you guys I'll put up an announcement to do that next weekend
12:29:38 <gjanssens> I'm thinking of dropping the "trunk" branch from that repo at the same time
12:30:03 <gjanssens> Htdocs has been git only for some time now
12:30:13 <gjanssens> I'd like to fade out the svn legacy
12:31:30 <jralls> OK with me, on both counts. If you haven't already, can you fiddle the gnucash and gnucash-docs repos to make 'trunk' and 'master' interchangeable?
12:32:23 <gjanssens> Ok, can you quickly remind me how to do that again (or do it yourself if that's faster than explaining)
12:34:15 <jralls> Have to remind myself first...
12:35:17 <gjanssens> :)
12:36:44 *** fell_ has quit IRC
12:37:38 *** fell_ has joined #gnucash
12:37:38 *** gncbot sets mode: +o fell_
12:39:27 *** benoitg has joined #gnucash
12:51:07 <jralls> Looks like it's already there. I'm changing the default on Github to be master instead of trunk.
12:58:51 *** MechtiIde has joined #gnucash
13:00:15 *** aqua___ has joined #gnucash
13:03:15 <warlord> gjanssens: sounds great re: docs. re: master, that's fine, too
13:03:43 <warlord> jralls: I think part of the issue is how we push to github; it might've missed some of that.
13:04:05 <warlord> gjanssens: will the htdocs change require a new clone or will just a pull work?
13:04:14 <jralls> No, it's a setting in the github repo control panel.
13:04:27 <warlord> jralls: ah, okay.
13:04:45 <warlord> at some point we should make 'trunk' == 'master' in the main repo, too..
13:04:54 <warlord> (maybe)
13:04:59 <warlord> .. .maybe it doesn't matter
13:05:11 <jralls> It already is. That's what I meant by "it's already there".
13:06:05 <jralls> After everyone has settled in to using git we can lose the trunk branches.
13:06:36 *** TradeBorG113 has joined #gnucash
13:06:37 <gjanssens> warlord: a reclone is best
13:07:08 <gjanssens> a pull will try to merge the new master branch into the old one
13:07:09 <gjanssens> I doubt that will go smoothly
13:07:25 <gjanssens> In addition, pulling will not clean your internal git store
13:07:45 <gjanssens> So you will still be stuck with the huge local repo
13:07:58 <jralls> Or two...
13:08:15 <gjanssens> A fresh clone is less than 30Mb the old repo is around 800Mb
13:09:01 <gjanssens> Of course those who need to push to gnucash-htdocs-docs will need to clone that one as well
13:09:11 <jralls> Though I'll probably be the only one who needs to use gnucash-htdocs-docs, at least as long as I'm the only one doing releases.
13:09:22 <gjanssens> Yep
13:09:48 <gjanssens> But it's interesting to observe that a fresh clone of gnucash-htdocs-docs is only 570Mb
13:10:05 <gjanssens> So together the two repos are still about 200Mb smaller than the combined repo
13:10:17 <gjanssens> And yet we haven't lost commits or branches
13:10:19 <gjanssens> I have double-checked
13:10:49 <gjanssens> I think git can optimize its internal object store better with the two separated
13:17:30 <warlord> Okay, so we'll have to redo the clone on www
13:17:55 <gjanssens> Yes
13:18:01 <warlord> (and I might also want to clone it on code, so code/search/externals actually gets updated properly... (maybe)
13:18:38 <warlord> is there any reason not to reclone htdocs off code?
13:19:04 <gjanssens> Would you want me to push to an intermediate repo on code (via gitolite)
13:19:10 <gjanssens> Would that make things easier ?
13:19:25 <gjanssens> I mean, does gitolite support something like renaming repos?
13:19:35 <warlord> I do not believe it does...
13:19:44 *** aqua_ has joined #gnucash
13:20:23 <gjanssens> Hmm
13:20:37 <gjanssens> What do you mean with reclone htdocs off code ?
13:21:22 <warlord> gjanssens: when we re-clone htdocs, instead of cloning from github, clone from code.
13:21:47 <gjanssens> You mean on www ?
13:21:55 <warlord> http://stackoverflow.com/questions/4708465/how-do-i-rename-a-git-repository-created-with-gitolite
13:22:02 <warlord> gjanssens: yes
13:22:13 <gjanssens> I see no issue with that
13:22:26 <warlord> It keeps a github dependency out of the loop.
13:22:54 <gjanssens> The only reason we use github is for bandwidth considerations and because that's our canonical read only repo
13:23:04 <gjanssens> But we could clone from code just as well
13:23:12 <warlord> And yes, it does look like one can rename a repo in gitolite, given shell access to the server.
13:23:24 <warlord> I figure for www it's a one-time deal.
13:23:47 <warlord> and I'm not starving for b/w conservation.
13:23:54 <gjanssens> Ok, then I think I'll push to a separate gitolite repo and ask you to rename it to gnucash-htdocs when done
13:24:45 <warlord> Oka.
13:24:49 <gjanssens> That seems the easiest way to get a thouroughly cleaned out gnucash-htdocs in gitolite
13:25:17 <gjanssens> Your trigger to push to github, is it name based ?
13:26:01 <gjanssens> or is that in our post-commit hook as well ? I don't remember
13:26:02 <warlord> Um, I think you wrote that code..
13:26:15 <warlord> The push to github should be in the post-commit hook
13:26:28 <warlord> the port-knockers are "private"
13:26:44 <gjanssens> Yes found it
13:27:02 <warlord> FWIW, over the past 24 hours my network has seen an average of 208.9 kb/s outbound.
13:27:16 <gjanssens> If the repo has an upstream named origin configured, it will push
13:27:31 <warlord> Over the past week it's been 244.4 kb/s
13:27:46 <warlord> ok
13:27:47 <gjanssens> So next to renaming the repo, you'll also have to add github as "origin" remote"
13:28:28 <warlord> right, there's no way to push a 'remote' config, is there?
13:28:37 *** aqua___ has quit IRC
13:29:08 <gjanssens> no
13:31:13 <gjanssens> Putting all pieces together, this sequence may have less waiting for each other:
13:31:45 <gjanssens> - (on server) move old repo out of the way and create empty repo
13:31:55 <gjanssens> - (on server) add origin to empty repo
13:32:14 <gjanssens> - (on github) move old repo out of the way
13:32:31 <gjanssens> - (on github) create empty new repo
13:32:47 <gjanssens> - (at kobalt) push pruned htdocs to new repo
13:32:53 *** TradeBorG113 has quit IRC
13:33:04 <gjanssens> => will get pushed through to github in one go
13:33:30 <gjanssens> (intermediary step: move old repo out of the way on www, put new empty repo in place)
13:36:17 <gjanssens> Hang on, I'll put it in the wiki somewhere
13:36:27 <gjanssens> irc is not very practical for editing :(
13:41:23 <warlord> the problem is that on www we need to clone the new-htdocs repo after step 2, but then move it "into place" on www after step 5/6
13:41:53 <warlord> otherwise www will pull wrongly at step 5/6
13:42:02 <warlord> This means we need linas to coordinate too
13:55:29 *** ErKa has quit IRC
13:56:52 <gjanssens> http://wiki.gnucash.org/wiki/Gitolite
13:57:41 <gjanssens> That procedure should work ok on www as well
13:57:46 <warlord> gjanssens: I just sent email to linas..
13:57:48 <gjanssens> I can clone on www by the way
13:57:56 <warlord> You can? From 'code'?
13:58:15 <gjanssens> No, but the automatic pull will do that for us
13:58:27 <gjanssens> I only need to set up an empty repo
13:59:29 <gjanssens> But in worst case, in fact I could even clone with my own ssh keys
13:59:34 <gjanssens> Didn't think of that before
14:00:07 <warlord> Ah, true, you could set up an empty repo. But... Don't we need Linas' help to make sure the scripts perform the pull from the correct directory?
14:00:22 <warlord> .. or that apache points to the right place?
14:00:35 <gjanssens> It depends
14:01:11 <gjanssens> My steps will mean that during the process http access to the gnucash will error out
14:01:32 <gjanssens> I suspect the whole operation to take less than half an hour
14:01:47 <gjanssens> If that's a planned downtime that should be acceptable
14:01:58 <gjanssens> If we really want www to remain online all of the time
14:02:08 <gjanssens> then yes linas will have to get involved
14:03:00 <gjanssens> ...or not if I first manually clone into a separate repo
14:03:16 <gjanssens> and only swap them when the clone is successful
14:03:56 <gjanssens> By the way git allows for some clever tricks with upstreams
14:04:30 <gjanssens> For htdocs, I could begin with a repo on www that pulls from github
14:04:38 <gjanssens> I don't need ssh keys for that
14:05:06 <gjanssens> Once the whole process is done I could simple set the upstream to code for any future automated pulls
14:05:08 <warlord> true, we could swap the 'origin' remote after it is set up.
14:05:25 <warlord> That might not be a bad approach to keep linas out of it.
14:05:35 <warlord> I think we'd rather not have website downtime if we can avoid it.
14:05:39 <gjanssens> Yeah, I believe he's rather busy
14:06:12 <gjanssens> Sure I'll update the steps a bit to make sure www remains online
14:06:38 <gjanssens> That means I'll have to temporarily suspend tickling www
14:06:49 <gjanssens> Not that difficult
14:11:01 <gjanssens> I have updated http://wiki.gnucash.org/wiki/Gitolite
14:11:32 <gjanssens> warlord I'll need your intervention two times
14:11:41 <gjanssens> once to move a repo out of the way on code
14:11:49 <gjanssens> and once to set the proper upstream for the new repo
14:12:28 <gjanssens> So we should coordinate when you can be available (for 5 mins or so)
14:16:09 <gjanssens> I'm planning this somewhere next Saturday afternoon (your morning)
14:26:19 <warlord> Saturday I am home, but I know we have someone coming to the house early (8:30am). I suspect I'll need to be at that session. I don't know what we're doing after that.
14:26:32 <warlord> frankly, week days are better for me ;)
14:34:13 *** lmat has joined #gnucash
14:34:28 <linas> hey
14:37:08 <linas> gjanssens you have setgid perms for the svn-mirror group, and so should be able to manipulate stuff there. ping me for changes to apache config
14:37:56 <gjanssens> linas thanks
14:38:08 <gjanssens> I don't think we need apache changes for our next step
14:38:12 <linas> BTW, I noticed git-clone took several tens of minutes to run; that was a surprise ...
14:38:41 <gjanssens> It's a relatively large repo you cloned
14:38:57 <gjanssens> That's why we're splitting it off from the ordinary website pages
14:39:48 <linas> Heh. Makes sense
14:45:18 <gjanssens> warlord: for me during the week works fine as well
14:46:34 <warlord> Okay.
14:46:44 <warlord> I have a little more control over my schedule during the week.
14:46:52 <warlord> Although if linas is around now.... ;)
14:48:57 <gjanssens> Shouldn't we first inform our developer base ?
14:49:11 <gjanssens> Anyway, this week is a little tight for me
14:49:24 <gjanssens> Next week can work on Tuesday or Thursday
14:50:18 *** gwb has joined #gnucash
14:50:23 <linas> uh what? huh?
14:50:30 <warlord> Okay. Next week I'll be in California, but that shouldn't really affect me (except perhaps what time of day I have free)
14:50:54 <warlord> linas: see http://wiki.gnucash.org/wiki/Gitolite
14:51:06 <warlord> (there's a bit of coordination on your end)
14:51:10 *** gwb has quit IRC
14:52:11 <gjanssens> warlord: no you're mistaken
14:52:16 <gjanssens> linas is not needed
14:52:38 <gjanssens> you're welcome to help out if you want to of course linas ;)
14:53:45 <gjanssens> let me rephrase "linas is not needed" it sounds a bit ...extreme
14:53:47 <gjanssens> there's no action required by linas
14:53:54 <gjanssens> there that sounds better :)
14:54:06 <warlord> Ah, I thought we would need his help to deal with the port-knocker scripts on ww
14:54:13 <gjanssens> these won't change
14:54:30 <gjanssens> I can disable them on the gitolite side
14:54:37 <gjanssens> and re-enable them afterwards
14:54:53 <gjanssens> as you said I wrote that code :)
14:55:07 <warlord> Ah, right, because we're going to clone from github first.
14:55:51 <warlord> So we disable the push to www, then migrate code/github, then we pull from github, make sure it works, repoint origin to code, and then turn back on the port-knocker
14:56:28 <gjanssens> exactly
14:57:05 <warlord> Gotcha
14:58:12 <warlord> Okay, well from my side I'm generally available. Next week I'll be best from 6-8am PT (9-11am ET, 1400-1600 UTC)
15:04:08 <gjanssens> If I'm counting well that would be 1500-1700 at my place (I keep forgetting UTC=GMT)
15:04:18 *** jmd has joined #gnucash
15:05:48 <gjanssens> Let's aim for Thursday that time (I have some more room later on that day in case something doesn't go as planned)
15:05:56 <lmat> gjanssens: Where are you staying ?
15:06:00 <lmat> oh yeah, probably Germany ;)
15:06:11 <gjanssens> Nope Belgium :)
15:06:52 <warlord> Okay, we can target Thursday..
15:07:02 <warlord> (this will be as soon as I wake up in Cali)
15:07:17 <gjanssens> I'll be on IRC, just give me a shout
15:09:24 <warlord> Okay, I'll shout out thursday as soon as I wake up
15:10:20 * gjanssens will write up an announcement now
15:10:51 <warlord> Sounds like a plan.
15:11:07 <warlord> I wonder if we can easily see who has a clone ?
15:13:50 <jralls> No, but we only care about the few people who have commit anyway. There's no reason for anyone else to use gnucash-htdocs.
15:14:12 * jralls wonders if we should take down the Github mirror, for that reason.
15:15:57 <warlord> jralls: people need access for website translation
15:16:13 <gjanssens> That makes sense
15:16:16 <jralls> Ah, good point.
15:16:27 <gjanssens> On github there are 3 forks of gnucash-htdoce
15:16:34 <gjanssens> s/e/s/
15:16:43 <gjanssens> one is cmarchi
15:17:06 <gjanssens> the other two I don't know (user fellen and kleopatra99)
15:17:26 <gjanssens> I think cmarchi is the only active website maintainer currently
15:18:14 *** MechtiIde has quit IRC
15:20:09 <jralls> Isn't fellen Frank Ellenberger (a.k.a fell_ )?
15:20:16 *** aqua_ has quit IRC
15:25:53 *** lmat has quit IRC
15:26:49 <jralls> Kleoppatra999 seems to be someone random who's been on GH for 2 months, has forked more than 30 repos (including gnucash-git-helpers) and hasn't pushed a single commit. I wouldn't worry about him.
15:28:41 *** lmat has joined #gnucash
15:45:29 *** aqua___ has joined #gnucash
15:58:17 *** lmat has quit IRC
16:01:55 <gjanssens> jralls warlord I have resurected http://wiki.gnucash.org/wiki/Git#Conversion_Notice and adapted it slightly for the gnucash-htdocs rebuild
16:02:07 <warlord> ok
16:02:59 <gjanssens> Can you proof read it please and report or fix any issues in it
16:03:00 <gjanssens> grammar or technical ;)
16:07:20 <warlord> *looking*
16:16:58 *** lmat has joined #gnucash
16:20:57 *** lmat has quit IRC
16:27:06 <gjanssens> There, notification is on the website
16:27:49 *** lmat has joined #gnucash
16:29:50 <warlord> gjanssens: you might want to give people a window to do the final pull/rebase
16:32:01 <gjanssens> Shouldn't that be until right before we start working on this ?
16:32:10 *** TradeBorG113 has joined #gnucash
16:33:04 <gjanssens> warlord I'm not sure I understand what you mean
16:33:44 <gjanssens> The website article states date and time and ends with
16:33:46 <gjanssens> So avoid that time frame to push or pull any changes.
16:33:50 *** jmd has quit IRC
16:35:36 *** fell__ has joined #gnucash
16:35:37 *** gncbot sets mode: +o fell__
16:41:33 *** aqua___ has quit IRC
16:41:33 *** fell_ has quit IRC
16:41:33 *** warlord has quit IRC
16:41:33 *** reaper has quit IRC
16:41:33 *** kim27 has quit IRC
16:41:44 *** kim27 has joined #gnucash
16:41:53 *** warlord has joined #gnucash
16:41:53 *** gncbot sets mode: +o warlord
16:42:00 *** lmat has quit IRC
16:47:27 *** lmat has joined #gnucash
16:51:49 *** gjanssens has quit IRC
16:52:35 *** lmat has joined #gnucash
17:08:56 *** jralls has quit IRC
17:12:57 *** lmat has quit IRC
17:41:16 *** lmat has joined #gnucash
17:52:19 *** benoitg has quit IRC
18:03:12 *** lmat_ has joined #gnucash
18:13:04 *** fell__ is now known as fell
18:13:36 <fell> github:fellen is /me
18:17:57 *** warlord has quit IRC
18:17:57 *** kim27 has quit IRC
18:19:28 *** warlord has joined #gnucash
18:19:28 *** kim27 has joined #gnucash
18:19:28 *** gncbot sets mode: +o warlord
18:27:54 *** reaper has joined #gnucash
19:05:16 *** lmat_ has quit IRC
19:05:26 *** lmat has quit IRC
19:33:41 *** benoitg has joined #gnucash
19:44:07 *** LiamH has joined #gnucash
20:07:47 *** vantage has quit IRC
20:45:05 *** ErKa has joined #gnucash
21:07:06 *** ErKa has quit IRC
22:04:15 *** kpreid has quit IRC
22:04:57 *** kpreid has joined #gnucash
22:18:27 *** TradeBorG114 has joined #gnucash
22:25:28 *** TradeBorG113 has quit IRC
23:32:32 *** LiamH has quit IRC