Log for #openttdcoop.devzone on 24th December 2009:
Times are UTC Toggle Colours
00:25:36  *** Frankr has quit IRC
00:31:21  *** Frankr has joined #openttdcoop.devzone
01:27:43  *** Frankr has quit IRC
01:34:51  *** PeterT has quit IRC
02:13:35  <Brot6> feed DevZone had 15 updates, showing the latest 10
02:13:35  <Brot6> ::DevZone:: Redmine - Revision 3068: Adds a setting to remove incoming emails body after a delimiter (#4409). @ (by jplang)
02:13:35  <Brot6> ::DevZone:: Redmine - Revision 3069: Adds Indonesian translation by Raden Prabowo (#4399). @ (by jplang)
02:13:35  <Brot6> ::DevZone:: Redmine - Revision 3070: Translation updates. @ (by jplang)
02:13:38  <Brot6> ::DevZone:: Redmine - Revision 3071: Fixes delimiters regexp. @ (by jplang)
02:13:41  <Brot6> ::DevZone:: Redmine - Revision 3072: Remove invalid escaping in version field (#4460). @ (by jplang)
02:13:45  <Brot6> ::DevZone:: Redmine - Revision 3073: Fixed: wiki pages in search results are referenced by project number, no... @ (by jplang)
02:13:49  <Brot6> ::DevZone:: Redmine - Revision 3074: Force TimeEntry#hours default to nil (#3075, #4449). @ (by jplang)
02:13:53  <Brot6> ::DevZone:: Redmine - Revision 3075: Set version to 0.9.0. @ (by jplang)
02:13:58  <Brot6> ::DevZone:: Redmine - Revision 3076: Upgrade to Rails 2.3.5 @ (by jplang)
02:14:01  <Brot6> ::DevZone:: Redmine - Revision 3077: Rails 2.3.5 deprecation. @ (by jplang)
02:33:54  <Brot6> Backup done! (Usage: 111M)
07:15:48  *** Rubidium has quit IRC
07:26:39  *** ODM has joined #openttdcoop.devzone
07:30:04  *** Rubidium has joined #openttdcoop.devzone
07:35:36  *** andythenorth has joined #openttdcoop.devzone
07:44:08  *** welshdragon has quit IRC
08:24:06  *** ODM has quit IRC
08:24:30  *** ODM has joined #openttdcoop.devzone
08:34:59  *** Zuu has joined #openttdcoop.devzone
08:58:17  *** welshdragon has joined #openttdcoop.devzone
09:15:49  *** Mark has joined #openttdcoop.devzone
10:03:01  *** Zuu has quit IRC
10:38:24  <andythenorth> planetmaker: I am looking at the changes.txt in the FISH repo...hasn't updated since r74.
10:38:28  <andythenorth> I thought it was automated?  Anything I need to do?
10:38:41  <planetmaker> it's not automated anywhere
10:39:15  <planetmaker> what I do is run hg log --template='- {desc}\n' -r74:tip >> changes.txt
10:39:16  <andythenorth> ah
10:39:27  <planetmaker> and then hand-edit to make the changes look nice
10:39:48  <planetmaker> or don't pipe it into the file, just have it display on the console and copy&paste to changes.txt
10:41:08  <andythenorth> hmm
10:41:18  <andythenorth> for a 0.1 release, do I need a changelog?
10:41:26  <andythenorth> Changes: tagged 0.1 and released? :P
10:42:15  <planetmaker> Well, but you might want to make a feature list then :-)
11:02:57  <Brot6> ::DevZone:: FISH - Revision 225: Added tag 0.1 for changeset a459ad0c770c @ (by andythenorth)
11:05:56  <andythenorth> planetmaker: ^^
11:06:48  <planetmaker> nice :-)
11:32:32  <Ammler> omg
11:32:46  <Ammler> openttd beta has music replacement support
11:33:55  <Rubidium> and NoMusic is free :)
11:35:34  <planetmaker> :-P
11:35:40  <Ammler> hmm, shall I create new tickets for the 256 icon and the desktop file?
11:35:44  <Ammler> Rubidium: ^
11:36:11  <Rubidium> there isn't one already?
11:36:29  <Ammler> there is one, but wrongly dimensioned
11:37:02  <Rubidium> I meant ticket
11:37:08  <Ammler> Rubidium:
11:37:22  <Ammler> but the ticket is a "WIP", so it might go lost there
11:37:24  <andythenorth> a Bananas release for FISH?
11:38:13  <Rubidium> Ammler: then we'll eventually notice it :)
11:38:23  <Rubidium> I'm only wondering there once every while
11:39:33  <Ammler> nice thing with rpms is, that you got informed about new files
11:39:41  <Ammler> else I wouldn't know about the gm
11:41:30  <planetmaker> andythenorth: it's accessible via the coop account
11:42:22  <Ammler> those memory issues sucks like hell
11:43:25  <Rubidium> yeah, leaking a lot is the new way to write webservices
11:44:29  <Ammler> well, there is almost nothing running on that server
11:45:24  <Rubidium> nevertheless... if it's writting in python, expect huge amounts of memory wastage
11:45:35  <andythenorth> meh
11:45:50  <Rubidium> the fracking OpenTTD frontpage uses 180! MB
11:45:54  <andythenorth> we have a Python CMS using all 8GB on our webserver
11:46:06  <Rubidium> mysql a whopping 30 MB
11:47:07  <Rubidium> ottd_content uses 7MB, but that's because it's currently caching OpenGFX
11:47:31  <Rubidium> if it's idling it's using less than 1.5 MB
12:32:38  *** andythenorth has quit IRC
12:34:11  *** andythenorth has joined #openttdcoop.devzone
12:56:59  *** andythenorth has quit IRC
13:56:55  *** PeterT has joined #openttdcoop.devzone
14:13:31  *** andythenorth has joined #openttdcoop.devzone
14:17:42  *** PeterT has quit IRC
14:26:30  *** PeterT has joined #openttdcoop.devzone
14:47:20  <andythenorth> Ammler / planetmaker devzone is offline ?
14:48:08  <PeterT> andythenorth: It works for me
14:48:19  <andythenorth> just came back
14:48:26  <andythenorth> looked like a 404 from a German ISP?
14:50:34  <PeterT> Something like: "Die Domain wurde über den Domain Registrar regfish GmbH registriert. Bitte haben Sie ein wenig Geduld, bis ein Angebot hinterlegt wird."
14:51:23  <andythenorth> yep
14:52:53  <Hirundo> I'm having the same problem
15:01:22  <Ammler> yeah, we transfered to a new registrar and this somehow failed to take over the name servers.
15:05:13  <andythenorth> Ammler: well it's mostly working
15:05:34  <andythenorth> definitely working just now for me
15:05:56  <Ammler> yeah, one server is there
15:06:23  <Ammler> 3 of the 4 servers are wrong, so if you are lucky, you take the right one :-)
15:06:39  <andythenorth> Ammler: are you busy with the servers?  I could use some release / hg help :)
15:07:40  <Ammler> no, that registrar transfer wasn't my task ;-)
15:08:05  <Ammler> and sadly I can't fix it either
15:08:19  <andythenorth> I am trying to tag and release v0.1 of FISH, but I have some troubles...
15:08:19  <Ammler> so yes, I have time :-)
15:08:51  <Ammler> I saw, you already tagged, what is the issue now
15:09:08  <Ammler> just "hg up 0.1"
15:09:15  <Ammler> make bundle_zip
15:09:22  <Ammler> make bundle_tar
15:09:42  <andythenorth> specific problem went away
15:09:50  <andythenorth> usually the case when someone else comes to help, no?
15:10:13  <Ammler> well, I am happy being able to help :-)
15:11:31  <andythenorth> so I now have:
15:11:31  <andythenorth>
15:11:39  <andythenorth> fish-nightly
15:11:41  <andythenorth> (dir)
15:11:51  <andythenorth> fish-nightly.tar
15:11:59  <Ammler> looks like you didn't update to the tag
15:12:18  <andythenorth> that's what I thought
15:12:22  <andythenorth> I did do hg up 0.1
15:12:29  <Ammler> hmm, you might need to run "make release"
15:12:40  <Ammler> and make release_zip
15:12:56  <Ammler> let me clone fish
15:13:11  <andythenorth> ok
15:15:53  <Ammler> yes, make release_zip
15:16:29  <andythenorth> gives me
15:16:50  <andythenorth> containing 2 dirs
15:16:53  <andythenorth> fish and fish-nightly
15:17:09  <Ammler> what does hg parent tell?
15:17:22  <Ammler> or hg id
15:17:38  <Ammler> it works here
15:17:55  <Ammler> I do upload the zip to the bundles dir
15:18:26  <andythenorth> pdq2s-macbook-3:fish_build andy$ hg parent
15:18:26  <andythenorth> changeset:   224:a459ad0c770c
15:18:27  <andythenorth> tag:         0.1
15:18:27  <andythenorth> user:        andythenorth
15:18:27  <andythenorth> date:        Mon Dec 21 21:58:41 2009 +0000
15:18:28  <andythenorth> summary:     Change: updated readme.txt
15:18:38  <andythenorth> pdq2s-macbook-3:fish_build andy$ hg id
15:18:38  <andythenorth> a459ad0c770c 0.1
15:21:34  <andythenorth> Ammler: think it worked now
15:22:05  <andythenorth> I thought make would fill in the tag number in the zip filename?
15:22:14  <andythenorth> never mind if it doesn't :)
15:22:45  *** PeterT has quit IRC
15:22:50  *** PeterT has joined #openttdcoop.devzone
15:23:35  <Ammler> andythenorth: it does
15:23:41  <Ammler> at least here
15:23:43  <andythenorth> oh
15:23:49  <andythenorth> maybe my hg is sulking
15:25:34  <andythenorth> which is better?  Fix my hg problem, or you do the bananas release :P
15:35:25  <andythenorth> I'll do the bananas release then :D
15:35:27  <Ammler> andythenorth: I have updated bananas, feel free to edit it
15:35:50  <andythenorth> nice
15:37:28  <Ammler> add some tags
15:43:07  <andythenorth> done
17:12:16  *** Hirundo_ has joined #openttdcoop.devzone
17:12:59  <Hirundo_> openttdcoop bouncer is down?
17:18:02  <Brot6> 2cctrainset: update from r<html><head><title>Domain wurde soeben registriert</title>
17:18:03  <Brot6> <link rel="stylesheet" href="" type="text/css">
17:18:06  <Brot6> </head><body align=center>
17:18:08  <Brot6> <p>&nbsp;</p><p>&nbsp;</p><table border=0 cellpadding=0 cellspacing=0 width="600" align=center><tr><td align=left style="font-family:Verdana,Arial; font-size: 17px; line-height:32px;color: rgb(80,80,80); padding: 10px; text-align: justify;">Die Domain <strong style="color:black;"></strong> wurde über den Domain Registrar <a href="" target=_top>regfish GmbH</a> registriert. Bitte haben Sie ein wenig
17:18:14  <Brot6> Geduld, bis ein Angebot hinterlegt wird.</td></tr></table>
17:18:16  <Brot6> <hr noshade size="1" width="650" color="#c9c9c9"><div class='footer' align="center"><a href="">regfish GmbH - Ihr Domain Registrar</a></div></body></html> to r413, starting nightly compile
17:40:41  <Brot6> 2cctrainset: compile done (1 errors) -
17:40:41  <Brot6> bros: nightly compile not needed. (r10)
17:40:41  <Brot6> comic-houses: nightly compile not needed. (r68)
17:40:41  <Brot6> firs: nightly compile not needed. (r366)
17:40:41  <Brot6> fish: update from r224 to r225, starting nightly compile
17:41:08  <Brot6> fish: compile done (0 errors) -
17:41:09  <Brot6> heqs: nightly compile not needed. (r172)
17:41:09  <Brot6> nmts: nightly compile not needed. (r15)
17:41:09  <Brot6> opengfx: nightly compile not needed. (r293)
17:41:09  <Brot6> opensfx: update from r69 to r70, starting nightly compile
17:42:11  <Brot6> opensfx: compile done (0 errors) -
17:42:12  <Brot6> worldairlineset: nightly compile not needed. (r584)
18:10:08  <Brot6> ::DevZone:: Infrastructure Sharing - Feature #670: Basic version @ (by Hirundo)
18:16:46  *** Hirundo_ has quit IRC
18:18:52  <Brot6> ::DevZone:: Infrastructure Sharing - Feature #670: Basic version @ (by Hirundo)
18:38:48  <welshdragon>
18:38:49  <Webster> Title: Does Visual Studio Rot the Mind? (at
19:39:11  *** PeterT has quit IRC
19:42:12  *** PeterT has joined #openttdcoop.devzone
20:26:55  *** PeterT has joined #openttdcoop.devzone
21:21:43  *** PeterT has quit IRC
21:29:03  <planetmaker> Hirundo: moving to bigger server
21:31:27  <Hirundo> bigger as in more cpu power and/or bandwidth?
21:36:24  <planetmaker> more cpu, more space, more RAM, more expensive :-P
21:36:26  *** PeterT has joined #openttdcoop.devzone
21:36:50  <planetmaker> same bandwidth (it's 5TB montly total I think before we pay extra)
21:37:25  <planetmaker> and we're at like 66GB for this month. So plenty of space left there :-)
21:38:09  <Rubidium> maybe there's some spiking today :)
21:41:09  <planetmaker> :-) Bundles is already on the new server. So that should not be affected.
21:41:44  <planetmaker> and yes, I don't mind that spike ;-) Christmas spike
21:42:46  <planetmaker> but stats show no spike today. Rather on the contrary. The far biggest transfers were on 14th
21:42:47  <Hirundo> planetmaker: <- opinions?
21:44:14  <planetmaker> let me look...
21:48:45  *** Mark has quit IRC
21:49:48  <planetmaker> looks like a sane and good separation from a quick scan
21:51:36  <planetmaker> but the main question now is: how is that going to be maintained?
21:51:39  *** Zuu has joined #openttdcoop.devzone
21:52:32  <planetmaker> I have no idea how to maintain a repo on top of a patch queue
21:52:48  <Hirundo> my preference would be to maintain the 'big IS' as a (set of ) patch(es) on top of it
21:52:52  <planetmaker> using a patch queue for these patches seems fine to me - especially, if they don#t interfere with eachother
21:53:22  <Hirundo> so there is only one repo to maintain with the same underlying core
21:54:14  <planetmaker> The only problem I personally found when working with a patch queue is, that it's in my experience more work to maintain than a repo like we have now.
21:54:28  <planetmaker> But then, now we have ONE patch, while a patch queue has n patches
21:54:35  *** PeterT has quit IRC
21:54:47  <planetmaker> The real fun begins, if patches start to depend on eachother
21:55:16  <planetmaker> and it's not too difficult to mess with your (local) repo during updates of the underlying repo
21:56:27  <Zuu> I though that the idea with a patch queue was that later patches should depend on patches in the beginning of the queue.
21:56:27  <planetmaker> dunno, how much you recall the procedure to kind of re-base the whole patch queue. If one doesn't call some of those hg commands at the wrong time or at the wrong base it's hard to get it reverted
21:56:58  <planetmaker> Zuu: yes... that's how it works. But not always what you want ;-)
21:59:14  <Hirundo> hg qpop -a; hg pull -u; hg qpush -a should rebase correctly
22:00:05  <planetmaker> yes. And then start fixing patches as they need it for a rebase... and forget where you are ;-)
22:00:45  <planetmaker> I started to make a copy of my whole repo just in case I completely f**ed up ;-) I don't say it's bad. Not at all
22:01:04  <planetmaker> I just say that it need greater care with updates than a simple trunk merge as we know now
22:01:30  <planetmaker> I used the same procedure for months for my local client patch queue
22:04:19  <planetmaker> the added complexity basically is in the fact that you need to call these commands in that sequence. While hg merge is one command.
22:04:42  <Zuu> I've found it easiest to for the base case when all patches are un-applied I use SVN for pulling the last OpenTTD trunk and then make a hg commit of that. Then I use hg queues and apply my patches one by one and resolve conflicts with trunk.
22:04:43  <planetmaker> and you need to fix things - iirc - between hg pull -u and hg qpush
22:05:13  <planetmaker> doesn't sound easier, Zuu :-)
22:05:21  <Zuu> For me it is easiest.
22:05:31  <planetmaker> the advantage of what Hirundo just posted is the 3-way merge. You have more info for the merge.
22:05:53  <planetmaker> old conflicts which you resolved don't need resolution again.
22:06:53  <Zuu> old as in a previous patch, but after the last update or earlier in time for another pull?
22:06:57  <Hirundo> The advantage of patches to branches is that patches don't require a dozen merge changesets per trunk merge
22:09:56  <planetmaker> Zuu: the advantage is bascially that you update the patched repo - and not the repo and then update each patch
22:11:14  <Zuu> Okay, and you resolve the conflicts automatically in the patch that affects the part that has the conflict.
22:11:30  <Rubidium> and hopefully it's too complex for petert :)
22:11:52  <Zuu> I mean, you sometimes need to resolve manually, but it somehow detects which patch it should update?
22:12:20  <Zuu> PeterT isn't here right now. :-)
22:12:34  <planetmaker> Zuu: yes, it tells you which patch fails
22:13:19  <Hirundo> if you hg qpush -a it stops when it fails to update, then you can fix the patch and continue updating
22:13:23  <Zuu> PeterT is quite interesting, sometimes he is quite clever and sometimes he makes noob statements. But maybe I shouldn't say too much or you would say the same about me :-)
22:13:46  <planetmaker> :-P Don't worry Zuu ;-)
22:14:16  <planetmaker> He's a kid. He's not stupid. And he wants recognition
22:14:44  <planetmaker> All three make for a somtimes painful or strainful experience
22:15:00  <planetmaker> (my personal 2ct)
22:16:58  <Rubidium> he is annoying as hell
22:18:18  <Zuu> pm: At least I'll have a masters degree in communications and transport systems sometime next year. So I'm not totaly useless in traffic engineering hehe :-)
22:18:32  <planetmaker> he adds his 2ct to too many things where he has no clue. And it's emberrassing that he walks the forums with out #openttdcoop banner
22:18:49  <planetmaker> hehe @ Zuu :-)
22:19:48  <Rubidium> ask him to remove it, or if he doesn't react as a mod to do it :)
22:19:51  <planetmaker> Rubidium: could you give advise? Would you rather recommend to use a patch queue or kind stay with some repo as we have now, use a repo with xxx heads like cargodist...? How do you develop bigger features?
22:20:22  <Zuu> pm: you mean, you don't like that he has the #openttdcoop banner?
22:20:41  <planetmaker> Zuu: yes... he's not good for the reputation of it ;-)
22:20:50  <Zuu> hehe :-)
22:20:56  <Rubidium> usually I just start and see where it ends; when I got something working I start bothering how it can be split in logical chunks and commit those
22:22:25  <planetmaker> I probably could also ask more directly: how should we? Hirundo? best sell you infrastructure sharing? ;-)
22:23:14  <Rubidium> not :)
22:23:47  <Rubidium> I'm personally not interested in the feature
22:24:42  <planetmaker> he :-) fair enough
22:24:47  <Rubidium> that doesn't mean that there isn't another developer that is interested in pushing it, or that I would veto it
22:25:27  <Hirundo> Rephrasing: I probably could also ask more directly: how should we? Hirundo? best sell infrastructure sharing? to a developer ;-)
22:25:38  <planetmaker> :-P
22:25:54  <Zuu> Perhaps finding that developer who is interested in the feature?
22:26:07  <planetmaker> I guess
22:26:38  <Zuu> I have two guesses, but both are rather bussy I think. However, tell me one dev that is not bussy?
22:27:50  <Rubidium> as a set of patches, unless it's a very small feature. Although having the full diff available at the same time would be useful (easier for doing a full review)
22:27:56  <planetmaker> you might consider to drop one "s" from bussy ;-) I also have exactly two guesses... Y and S ;-)
22:28:13  <Rubidium> unless it's incredibly complex, when we'd generally review the patches themselves
22:28:42  <planetmaker> hm... so a patch queue sounds like what is best reviewed
22:29:47  <planetmaker> even in its really, really basic form an IS patch is never small.
22:30:22  <Zuu> planetmaker: my two guesses for who might be interested in the feature are Y and B.
22:30:36  <Hirundo> currently it's around 50kb total, roughly two-thirds of which is pretty trivial
22:30:56  <planetmaker> 50kB is a lot less than last time I checked...
22:31:11  <Rubidium> oh, and never tell lies; that would instantaniously drop your rating
22:31:19  <Hirundo> <- is 52k
22:32:11  <planetmaker> Lies usually and naturally do... if discovered ;-)
22:32:20  <planetmaker> But why and where would you lie about a patch?
22:32:48  <Rubidium> saying things like: I've done this and that and it's faster, when it's actually slower
22:33:45  <planetmaker> which boils down to "be able to prove your claims".
22:33:46  <Rubidium> FS#3304 someone saying he changes some condition because he thinks it's more correct
22:37:41  <Rubidium> about the former especially when the person later finds out that his testing was 'flawed'
22:38:37  <Rubidium> only testing some things and losing the big picture
22:39:37  <Rubidium> ofcourse not merging cargopackets is faster than merging them, or merging only the last... but in most cases you'll end up with way more packets which means you have to handle more packets and it ends up being slower
22:40:31  <planetmaker> well... therefor it should be clearly stated *what* was tested...
22:41:06  <planetmaker> which then reminds me of this one link - I guess you gave it to me once - which had 10(?) points, and one was that good testers are hard to come by.
22:41:47  *** andythenorth has quit IRC
22:41:58  <Rubidium> probably something of Joel
22:43:13  <planetmaker> yes, indeed
22:44:44  <Rubidium> anyhow, that's why I wanted to do the first beta early as then much more people start using it and thus more semi-proper testers are using it
22:45:03  <Rubidium> it amazes me how many things are reported for 1.0.0-beta1 that were already present in 0.7.0-beta1
22:46:48  <Rubidium> anyhow, today I'm not going to fix any of them
22:46:56  <planetmaker> hm, yes :-) It makes people actually try a few things they commonly might not do. I don't exclude myself from that list
22:47:28  <planetmaker> or maybe look again at something one might have wondered somewhen but couldn't be bothered right back then.
22:48:34  <Rubidium> something else...
22:48:52  * Rubidium wonders when petert figures out he needs some custom build dll to make the installer
22:50:22  <planetmaker> he... does he post useless binaries?
22:51:09  <Rubidium> no, he just can't build them
22:51:24  <Zuu> Why would he want to build the installer?
22:51:40  <Zuu> Oh, of course for his binary imperium. :-)
22:51:47  <Rubidium> because 20 minutes after the tagging of 1.0.0-beta1 there was no windows binary/installer yet
22:52:10  <Hirundo> Imperium Binarium ;)
22:55:57  <planetmaker> lol
22:57:41  * Hirundo heads off to bed
22:57:50  <Rubidium> Hirundo: good idea
22:58:53  <planetmaker> indeed.
22:58:59  <planetmaker> good night :-)
23:02:19  *** welshdragon has quit IRC
23:19:49  *** PeterT has joined #openttdcoop.devzone
23:25:48  *** Zuu has quit IRC
23:37:57  *** ODM has quit IRC

Powered by YARRSTE version: svn-trunk