IRC logs for #openttd on OFTC at 2020-10-10
⏴ go to previous day
01:33:34 *** jottyfan has joined #openttd
08:59:22 *** longtomjr_ has joined #openttd
09:15:21 *** sla_ro|master has joined #openttd
09:46:32 *** Progman has joined #openttd
09:54:09 *** HerzogDeXtEr has joined #openttd
09:56:56 *** andythenorth has joined #openttd
10:07:22 *** jottyfan has joined #openttd
10:31:20 <TrueBrain> lol, I know why webster always lags behind 1 log line ..
10:31:33 <TrueBrain> Supybut flushes its logs before write, not after :D
10:34:49 *** tokai|noir has joined #openttd
10:34:49 *** ChanServ sets mode: +v tokai|noir
11:19:59 *** WormnestAndroid has quit IRC
11:20:12 *** WormnestAndroid has joined #openttd
12:28:00 *** arikover has joined #openttd
14:05:02 *** frosch123 has joined #openttd
14:21:53 <LordAro> eints can't change "ownname", can it?
14:24:53 <frosch123> it can't. only strings, no meta ##
14:26:02 <frosch123> but eints is supposed to automatically restart if you change the ## via git
14:26:16 <frosch123> so you can be the first one to test that :)
14:26:52 <LordAro> is it worth the bother of getting them to rebase to fix the commit message?
14:26:58 <LordAro> or shall i just force merge it?
14:27:34 <frosch123> it's a squash anyway
14:29:09 <frosch123> ah, it's noun vs adjective
14:31:33 <LordAro> frosch123: definitely correct though?
14:34:33 * LordAro uses Administrator privileges
14:34:57 <frosch123> does dorpsgek know you?
14:37:36 <LordAro> looks like msys is happier now
14:41:59 <andythenorth> I might go to the shop
14:42:09 <andythenorth> buy cheese and so on
14:50:36 <LordAro> andythenorth: i am also perilously low on food
14:51:05 <frosch123> meh, so many choices... mediawiki capitalises sites, so "foobar" becomes "Foobar". gollum does not. both are case sensitive otherwise
14:51:29 <frosch123> should i just make everything case insensitive?
14:52:43 <LordAro> probably not a bad idea, tbh
14:53:07 <andythenorth> I don't have fancy winter-is-coming type food
14:53:21 <andythenorth> basic survival looks likely though
15:22:54 *** ChanServ sets mode: +v tokai
15:47:28 *** matt21347 has joined #openttd
16:00:13 <TrueBrain> frosch123: eints no longer does anything on git commit :P It does it once a day ;) Does it process meta-data too?
16:00:27 <TrueBrain> the lang-sync stuff is no longer executed on a schedule
16:02:27 <TrueBrain> no clue why you wanted me to review that PR; seems anyone could have seen that was obviously correct :P
16:02:49 <LordAro> we spoke a bit about it before
16:03:00 <TrueBrain> just surprised nobody else gave it a review :)
16:03:05 <LordAro> i guess the next step is the release CI...
16:03:11 <TrueBrain> owh boy .. the big one!
16:03:19 <TrueBrain> with the container-stuff you found, that can be pretty interesting
16:03:32 <TrueBrain> it is really cool you can execute stuff inside a container :)
16:03:36 <glx> oh I could have approved it, but I was testing stuff :)
16:04:06 <LordAro> TrueBrain: i've forgotten already, what was it that i found? :p
16:04:47 <TrueBrain> you can execute steps inside a container
16:05:01 <TrueBrain> so you can have an upstream ubuntu-16.04, ubuntu-18.04, debian, fedora-core, etc
16:05:04 <TrueBrain> and execute stuff in there
16:05:11 <TrueBrain> meaning we no longer need the CompileFarm stuff, at all
16:05:25 <TrueBrain> and yes, it takes a bit longer to install the dependencies, but it will be much more readable
16:06:53 <TrueBrain> also means supporting things like rpm will be easier :)
16:07:16 <TrueBrain> really curious how the 1st of November will go with the Docker Hub libraries .. like "ubuntu:16.04"
16:07:22 <TrueBrain> so far it seems they will rate limit those too
16:07:32 <TrueBrain> which will grind many many many GitHub Actions to a halt, I am afraid :P
16:10:49 <TrueBrain> LordAro: the part I can help with, with the release-stuff, is after the binaries are produced, how they show up for download. This is now part of "Releases" in Azure DevOps, and because of the way that works, not part of the OpenTTD repository
16:10:59 <TrueBrain> with GitHub Actions they can be moved into proper workflows, like the other repositories we have
16:11:16 <TrueBrain> so once you/who-ever build the part that produces binaries, I can pick it up from there, basically
16:11:52 <TrueBrain> the workflows already exist in different repositories; just a matter of picking the right one, and assigning the right permissions in AWS :D
16:12:48 <TrueBrain> and, I would suggest we publish the binaries in the Releases on GitHub, like with NML
16:12:59 <TrueBrain> so people can also download the release via GitHub, in binary form, if they like
16:13:55 <TrueBrain> so that might be a nice first step, being able to do that :D
16:14:26 <glx> anyway the idea behind run-vcpkg is to pin dependencies, so it makes sense to not auto update
16:15:11 <TrueBrain> but he accepts PRs for this; sounds you already have the solution, so might be worth making that PR?
16:15:16 <TrueBrain> (I might be overly optimistic :D)
16:15:38 *** Speeder has joined #openttd
16:15:42 <TrueBrain> I do not see how run-vcpkg pins dependencies btw; cache can be wiped on a moments notice ..
16:16:17 <glx> the prefered way is to have vcpkg as a submodule
16:16:36 <TrueBrain> ugh .. that sucks balls :P
16:17:15 <TrueBrain> I think for OpenTTD "latest release" / "latest tag" works fine .. the chances of it breaking are pretty small, I would think
16:18:20 <glx> but it can also be used to download vcpkg in an arbitrary location (I do it over the github runner vcpkg) and with "do not upgrade" it was like using directly the given vcpkg install
16:19:02 <glx> but if you want a different version you need to give the commit id
16:20:53 <glx> latest vcpkg release is from before the msys2 fix :)
16:21:11 <TrueBrain> yeah, but running "latest" has other issues
16:21:58 <glx> yeah using current master is not fail proof either
16:22:19 <TrueBrain> if msys was still down, I would expect there to be already another vcpkg release :)
16:22:30 <TrueBrain> but, given that is fixed for now, the pressure is off
16:23:00 <TrueBrain> anyway, you could also debate that GitHub would have upgraded its runners by now, if it was still down and vcpkg released
16:23:05 <TrueBrain> so possibly this is a mute point anyway :D
16:24:13 <glx> yeah so probably the "best" way is to hardcode a commit id in the workflow (like using a submodule) and update it from time to time
16:24:39 <TrueBrain> if you really want to pin dependencies, sure .. but .. sounds a bit irrelevant for OpenTTD :)
16:24:54 <TrueBrain> not enough packages are used from vcpkg to really break anything, I would think
16:25:16 <glx> anyway it works now, so no hurry to touch it
16:26:44 <glx> and I think vcpkg will change again in this area, as current repo.msys2.org is not dimensionned to be asked directly
16:41:49 *** snail_UES_ has joined #openttd
16:48:15 <TrueBrain> well, can't blame them for not using a proper CDN for their master, as it is expensive as fuck (just looking at OpenTTD< which does run on a proper CDN atm .. it is expensive as fuck :P). Using free mirrors is always better
16:48:21 <TrueBrain> so moving the master out of that rotation is a good thing
16:48:43 <TrueBrain> sorry, forgot the new "no more master" rule :P
16:49:27 <glx> and I think they are looking to get some more mirrors
16:50:09 *** ramesh is now known as shudharson
16:50:36 <TrueBrain> I still need to offload some of the bigger files to a cheap alternative .. like a OVH VPS :P
16:50:46 <TrueBrain> before our free credits run out :D
16:52:35 <shudharson> I joined translation team for tamil through github. When I try to edit lang files it shows 'access denied'.
17:02:42 <glx> I see you are in the right team
17:04:01 <glx> are you logged in translator ?
17:06:13 <shudharson> i am logged in with my github profile name in translator
17:06:47 <frosch123> did you login before you joined the team? you may need a logout/login cycle then
17:07:13 <shudharson> ok i will try that then
17:07:27 <frosch123> just press "logout" on eints site
17:07:31 <frosch123> no need to logout from github
17:19:33 *** jottyfan has joined #openttd
17:46:53 <LordAro> i feel like that shouldn't have been necessary
17:47:02 <LordAro> why can't it just take the header from the existing file?
17:50:38 *** jottyfan has joined #openttd
17:52:39 <TrueBrain> I feel an eints v2 being created :p
18:02:00 <frosch123> LordAro: because it is a newgrf translation service, abused into an openttd translation service
18:02:53 <frosch123> anyway, eints needs the metadata to validate things, so it needs a restart either way
18:03:04 <frosch123> just previously it did that automatically when ottd pushed
18:07:58 <frosch123> thanks, one more migrated :)
18:09:04 <frosch123> migrated: GitHub packages 536/1648 (32.5%), GitHub authors 75/469 (16.0%)
18:09:32 <frosch123> likely many of the "one hit wonders" won't migrate
18:09:49 <glx> yeah probably many inactive authors
18:12:45 <frosch123> there is one yet-unmigrated author, who could contribute 5% to the migrated packages on their own
18:15:09 *** Wormnest has joined #openttd
18:18:56 *** HerzogDeXtEr has joined #openttd
18:33:55 *** andythenorth has joined #openttd
18:58:31 <TrueBrain> frosch123: why waste time on a GH action if this is a very unlikely event to ever happen again? :)
19:05:36 <TrueBrain> so .. where did I leave the BaNaNaS migration script .. eeuuuuhhhhh
19:08:08 <TrueBrain> frosch123: I found a really old migration request, that was only directed at me .. and I completely missed it till today .. I also completely forgot how to do migrations ... so I forwarded it to you .. so sorry :(
19:14:04 <frosch123> no idea what you forwarded, maybe mails are stuck somewhere
19:14:48 <TrueBrain> cannot believe I do not have that script on my disk anymore .. I mean .. lol?
19:15:09 <TrueBrain> ah, I do, but in WSL1, not WSL2
19:15:13 <TrueBrain> let me migrate him :)
19:17:18 *** arikover has joined #openttd
19:17:51 <TrueBrain> frosch123: owh, it seems I did not understand tt-forums enough to really forward it to you
19:18:01 <TrueBrain> turns out only filling in your name is not enough .. you also have to press the weird "Add" button
19:21:00 <TrueBrain> so another one for the stats!
19:49:32 <andythenorth> license.md or copying.md?
19:49:38 * andythenorth tidying up some newgrf repos
19:50:10 <TrueBrain> there is a stack overflow for EVERYTHING
19:50:44 <andythenorth> github says LICENSE, I wondered why we don't do that for main OpenTTD :)
19:51:32 <TrueBrain> OpenTTD repositories has more LICENSE than COPYING, if you care about those stats
19:51:58 <DorpsGek_III> - Update: Translations from eints (by translators)
19:52:17 <andythenorth> the newgrfs have to be multi-license
19:53:12 <andythenorth> hmm maybe today is not a legal day :)
19:53:15 <andythenorth> have a day off andythenorth
19:53:33 <andythenorth> what tomorrow as well?
19:53:56 <andythenorth> I went to the shops earlier, how much more living would I need?
19:54:03 <TrueBrain> nah, that is about it :P
20:27:52 *** Speeder_ has joined #openttd
20:37:54 *** Gustavo6046 has joined #openttd
20:53:24 *** snail_UES_ has joined #openttd
21:25:21 *** longtomjr_ has joined #openttd
21:30:01 *** gelignite has joined #openttd
21:39:55 <Afdal> What is the actual breakdown probability of a vehicle at any given tick?
21:40:16 <Afdal> as a function of current reliability
21:44:21 <Afdal> Trying to calculate the expected uptime for a logic gate with optimal reliability/servicing
21:45:02 <Afdal> And what optimal servicing actually looks like
21:45:28 <Afdal> considering the tradeoff of time spent going in and out of a depot versus time spent idle while broken down
22:01:01 *** Progman has joined #openttd
22:04:54 <Afdal> Where in the source is the Next() function defined?
22:12:23 <Eddi|zuHause> Afdal: afair a vehicle rolls a random value that is decreased until 0, then it breaks down, and rolls a new value
22:13:25 <Afdal> and current reliability % modulates that random number?
22:13:50 <Eddi|zuHause> i think reliability determins the maximum value that can be rolled
22:14:40 <Afdal> so it's a time series function then
22:15:08 <Afdal> I'd like to know the actual math going on at each step
22:15:28 <Afdal> but trying to interpret this code is hard -_-
22:18:15 *** tokai|noir has joined #openttd
22:18:15 *** ChanServ sets mode: +v tokai|noir
22:23:01 *** michi_cc has joined #openttd
22:23:02 *** ChanServ sets mode: +v michi_cc
continue to next day ⏵