Log for on 14th April 2013:
Times are UTC Toggle Colours
00:21:21  *** Zuu has quit IRC
02:02:50  *** Supercheese has joined
06:45:09  *** Alberth has joined
06:45:10  *** ChanServ sets mode: +v Alberth
07:29:23  *** Supercheese has quit IRC
08:00:40  *** Zuu has joined
08:00:40  *** ChanServ sets mode: +v Zuu
08:08:37  *** fonsinchen has joined
08:08:37  *** ChanServ sets mode: +v fonsinchen
08:22:03  <Zuu> Rubidium: It looks like you have restarted However, I get the SSL error that indicate that the server caught an exception. When you have time, could you see if its the MySQL connection that has died or what the error log is?
09:32:35  <fonsinchen> I've further reduced the code necessary to add a link graph overlay. You can find the result at in branch "frosch2"
09:32:41  <fonsinchen> Specifically:
09:32:42  *** frosch123 has joined
09:32:43  *** ChanServ sets mode: +v frosch123
09:32:55  <fonsinchen> 1. I've removed the unnecessary settings
09:33:07  <fonsinchen> 2. I've removed the flow stat implementation
09:33:32  <fonsinchen> 3. I've added functionality to clear dead links and "compress" old ones.
09:35:20  <fonsinchen> The overlay will never show overloaded links like this. I don't know if that's a problem.
09:38:00  <fonsinchen> It could be hidden by just not showing the "overloaded" part in the legenda.
09:45:45  <Alberth> quak
09:46:04  <frosch123> moin :)
09:46:19  <Alberth> (11:34:59 AM) fonsinchen: I've further reduced the code necessary to add a link graph overlay. You can find the result at in branch "frosch2"   <-- that was just before you entered
09:46:28  <frosch123> i read the logs :)
09:46:35  <Alberth> ah, ok
09:46:55  <frosch123> anyway, is it more clever to clone the whole git thingie? or should i just take the commit-difs and insert them into a proper vcs :p
09:47:32  <Alberth> I wonder whether there is a hg interface to git
09:47:44  <fonsinchen> I'm not going to take part in any hg vs. git flame war here.
09:48:07  <frosch123> Alberth: that would still require the whole cloning
09:48:09  <Alberth> I was not trying to start one
09:48:37  <Alberth> frosch123: ah, that's what you are trying to prevent. Makes sense
09:49:04  <fonsinchen> The new repository at ulfhermann/openttd.git is smaller than the old one.
09:49:21  <fonsinchen> I've ran it through gc --aggressive before uploading.
09:49:22  <frosch123> ok, so i'll just clone it :)
09:50:07  <Alberth> I found recently that dumb applying 'patch' fails with git patches and adding/moving/removing files. We need a smarter 'patch', but I don't know whether it exists
09:50:40  <frosch123> well, 'svn patch' or 'hg import' does it, doesn't it?
09:51:04  <Alberth> svn patch does not, new files end up at  a/<path> :p
09:51:20  <Alberth> or b/<path>, I don't remember :)
09:51:22  <fonsinchen> You need to do "patch -p1"
09:51:55  <Alberth> svn patch does understand git patches with just changes, without -p1
09:52:33  <Alberth> so it does some magic, just not for new files :(
09:52:45  <Alberth> or perhaps better worded, not yet
09:52:45  <fonsinchen> You could first do the raw patching with just "patch" and later add the stuff to svn.
09:52:56  <fonsinchen> Btw, git is a very good svn client.
09:53:08  <fonsinchen> I'm using it all the time like this.
09:53:42  <Alberth> hg is good to. /me and git just don't combine well
09:54:44  <fonsinchen> I was just wondering why you're using the original svn client. That's probably the worst option.
09:55:23  <Alberth> "svn patch" is subversions idea of applying 'patch' to a svn working copy
09:55:40  <Alberth> it's quite new, and not fully operational yet
09:56:46  <Alberth> but indeed, perhaps hg works better here
09:57:21  <fonsinchen> I've messed up the documentation for DeleteStaleLinks, it seems. The version I've committed does not actually reroute cargo.
09:58:38  *** ntoskrnl has joined
10:02:30  <frosch123> hmm, did i clone the wrong thing? i only offers me a cd branch, no frosch2
10:03:38  <fonsinchen> It has only one branch?
10:03:47  <frosch123> "git branch" only lists "cd"
10:03:51  <fonsinchen> That is weird as it should have about 50
10:03:58  <fonsinchen> maybe "git branch -a"
10:04:13  <fonsinchen> Then git checkout -b something origin/frosch2
10:04:14  <frosch123> yeah, that gives me a lot remote ones
10:04:31  <fonsinchen> I think it only auto-tracks the default branch
10:09:58  <Alberth> magic options :)
10:22:56  <frosch123> i suppose all cdist players transport pax :)
10:23:28  <frosch123> for cargos the linkgraph always displays "unused" in one direction :p
10:24:43  <Alberth> :)
10:24:47  <fonsinchen> sure, because the link is ununsed in the "back" direction most of the time
10:24:59  <fonsinchen> I mean, most cargo is only transported in one direction, isn't it?
10:25:11  <frosch123> yeah, it's just the first time that i see it :)
10:25:27  <frosch123> what smallmap land colour are you using btw?
10:25:32  <Alberth> use auto refit :)
10:25:59  <fonsinchen> various. Is there a combination that looks bad with the overlay?
10:26:40  <fonsinchen> You can also use a "leave empty" order. Then the back link won't be created.
10:26:59  <frosch123> well, green lines on green map are barely visible, but i guess red lines are more interesting anyway
10:27:31  <frosch123> ah, leave empty makes sense
10:28:10  <fonsinchen> That's true. We could add different link color schemes for different land colors but that may be confusing.
10:29:11  <fonsinchen> Or we could find a different link color scheme that fits all land colors better. I've been experimenting with different colors and that white->green->yellow->red thing seemed to be generally accepted by the players.
10:29:46  <frosch123> am i right, that it will only display white (empty vehicles) or green (full vehicles) in this branch? while overloaded only happens with destinations?
10:29:56  <fonsinchen> True
10:30:45  <frosch123> well, i guess minimap and colours is quite hopeless in general :p
10:31:17  <frosch123> there's just too much information. though wrt. green maps one might try blue for non-saturated links
10:31:23  <fonsinchen> I'm using the same colors also for the main viewport overlay.
10:31:37  <fonsinchen> (that's not in your branch, though)
10:32:07  <fonsinchen> blue instead of white?
10:32:16  <frosch123> blue instead of green
10:32:33  <fonsinchen> Can be done but looks bad on whater of course
10:32:44  <fonsinchen> s/whater/water/
10:32:47  <frosch123> well, this will do for now
10:33:16  <frosch123> we can leave adding more colour schemes to andy :)
10:42:58  *** Ristovski has joined
12:22:05  *** Ristovski has quit IRC
12:22:27  *** Ristovski has joined
12:58:41  *** fonsinchen has quit IRC
16:46:46  *** ntoskrnl has quit IRC
16:46:46  *** shametd_ has quit IRC
16:46:46  *** michi_cc has quit IRC
16:46:46  *** Rubidium has quit IRC
16:46:47  *** ntoskrnl has joined
16:46:47  *** shametd_ has joined
16:46:47  *** michi_cc has joined
16:46:47  *** Rubidium has joined
16:46:47  *** sets mode: +vv michi_cc Rubidium
16:48:03  *** ChanServ sets mode: +v orudge
17:42:42  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r25187 || Logs: || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
18:26:33  *** Alberth has left
18:38:16  *** Supercheese has joined
19:43:14  <Rubidium> Zuu: I've done nothing like restarting or updating lately
19:45:00  <Rubidium> the last error was "MySQL server gone away"
19:45:36  <Rubidium> updated and restarted it though
19:50:40  *** Zuu has quit IRC
19:50:59  *** Zuu has joined
19:50:59  *** ChanServ sets mode: +v Zuu
19:51:50  <Zuu> Thanks Rubidium I saw your message (and then my xchat crashed :-) )
20:14:34  *** ntoskrnl has quit IRC
20:42:00  *** Ristovski has quit IRC
21:00:08  *** frosch123 has quit IRC
21:47:43  *** Zuu has quit IRC

Powered by YARRSTE version: svn-trunk