Log for #openttd on 10th December 2020:
Times are UTC Toggle Colours
00:00:07  <glx> and yes I tested it :)
00:00:14  <TrueBrain> can't believe you fixed it :P
00:00:44  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #8368: Fix: Prevent infinite recursion in commit checker
00:01:06  *** WormnestAndroid has quit IRC
00:01:25  *** WormnestAndroid has joined #openttd
00:01:57  <glx> (the test was comparing to 0 instead of 256 and adding dumb commits to trigger it
00:02:34  *** iSoSyS has quit IRC
00:02:40  <TrueBrain> :D lol
00:04:11  <TrueBrain> hmm, I expected Stadia would have crashed by now, but shocking, the game starts :o
00:04:53  <milek7> looking at useful things in my emscripten pr
00:04:57  <milek7> maybe this?
00:05:04  <milek7>
00:05:33  <milek7> I don't know whether other platforms require it, but missing locking unconditionally is technically wrong
00:05:57  <TrueBrain> I noticed you only did it for the SDL1, so I assumed it was fixed for SDL2?
00:06:44  <TrueBrain> I honestly don't know enough about SDL to judge that commit .. maybe PR it, and we can poke .. who wrote SDL2? :D
00:08:49  <DorpsGek> [OpenTTD/OpenTTD] glx22 merged pull request #8368: Fix: Prevent infinite recursion in commit checker
00:10:30  <milek7> honestly I don't remember how it is in sdl2
00:14:19  <milek7> I guess building 1.10.3 with cmake is a lot of work?
00:18:58  <TrueBrain> we called it 1.11, yes :)
00:35:12  *** sla_ro|master has quit IRC
00:43:04  *** nielsm has quit IRC
02:11:57  *** Wormnest has quit IRC
02:48:46  *** Flygon has joined #openttd
03:09:07  *** glx has quit IRC
03:47:23  *** debdog has joined #openttd
03:50:43  *** D-HUND has quit IRC
05:00:48  *** Speeder_ has joined #openttd
05:06:36  *** Speeder has quit IRC
05:31:55  *** keoz has joined #openttd
06:20:18  *** DasPoseidon has joined #openttd
06:23:51  *** DasPoseidon has quit IRC
06:39:58  *** keoz has quit IRC
07:16:41  *** tokai has joined #openttd
07:16:41  *** ChanServ sets mode: +v tokai
07:19:03  *** tokai|noir has quit IRC
07:21:56  <DorpsGek> [OpenTTD/OpenTTD] orudge commented on pull request #8340: Draft: Feature: Create Universal (x86_64 + Apple Silicon) build on macOS
08:05:28  *** andythenorth has joined #openttd
08:51:19  *** HerzogDeXtEr has joined #openttd
09:01:29  <andythenorth> ha triggered an assert with a heightmap
09:01:31  <andythenorth> good times
09:26:53  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8340: Draft: Feature: Create Universal (x86_64 + Apple Silicon) build on macOS
09:35:23  <TrueBrain> orudge: you don't have to ^^, but at some point someone does, I think :) But we can do it when we merge it in the release pipeline honestly
09:35:35  <TrueBrain> Mac OS always have been the slowest targets :P
09:36:52  <TrueBrain> and I cannot believe vcpkg is useful on anything not Windows .. they did a nice job honestly :)
09:37:41  <andythenorth> oofsies
09:37:51  * andythenorth must not be distracted, work work work to do
09:37:54  *** andythenorth has left #openttd
09:38:42  <TrueBrain> poor andy :(
10:02:21  <DorpsGek> [OpenTTD/OpenTTD] orudge commented on pull request #8340: Draft: Feature: Create Universal (x86_64 + Apple Silicon) build on macOS
10:03:17  <orudge> TrueBrain: I've written a cross-platform library at work and make use of vcpkg for dependency building for that on Windows, macOS and Linux, x86/x86_64 and arm64, it does work rather well.
10:03:21  <orudge> I wonder if I can port it to OS/2 next :D
10:04:13  <TrueBrain> ..... owh boy ....
10:04:19  <TrueBrain> this is me leaving the room :P
10:04:43  <orudge> OpenTTD with cmake does still build on OS/2 by the way
10:04:48  <orudge> I maybe had to make a couple of tweaks
10:04:59  <orudge> but when I tried to run it, it was crashing, but I think that's an SDL bug
10:06:01  <LordAro> ha
10:18:39  *** sla_ro|master has joined #openttd
10:18:44  <TrueBrain> milek7: I DID IT! I think .... :P
10:19:33  <TrueBrain> and you didn't update AUTHORS in your other PRs :P
10:21:06  <TrueBrain> what I still do not like about GitHub, that every time you push a commit with "Fixes" in it, even if it is in a now-dead-branch, it shows that commit in the issue
10:21:17  <TrueBrain> so the issue can be full of spam if you forced-push a few times
10:21:28  <TrueBrain> guess it is better to only mention "Fixes" in the PR, and not the commit
10:21:48  <TrueBrain> as example of the spam ..
10:56:17  *** iSoSyS has joined #openttd
11:06:14  *** iSoSyS has quit IRC
11:09:15  *** Herr_Bert has joined #openttd
11:09:57  <Herr_Bert> Hello folks
11:10:03  <FLHerne> Morning
11:10:45  <Herr_Bert> its lunchtime over here :D
11:11:01  <Herr_Bert> I am looking for some forum admin
11:11:12  <FLHerne> orudge is your forum admin
11:11:40  <FLHerne> (but you should probably just use the 'report' button :p)
11:12:45  <Herr_Bert> issue is - i havent been there ~ 6 years and i forgot which email address i used for password recovery
11:13:06  <Herr_Bert> so i guess i cannot report :(
11:13:46  <orudge> Herr_Bert: do you know what your username was?
11:13:49  <orudge> or is
11:14:22  <Herr_Bert> <- it's me
11:14:58  <Herr_Bert> hi owen. you might not remember me, but nice to see you are still around here
11:16:15  <orudge> Herr_Bert: I've sent you a /query message
11:16:40  <orudge> Yep, still around :)
11:33:08  <TrueBrain> we are as shocked as you are :P
11:38:59  *** Speeder__ has joined #openttd
11:45:28  <Herr_Bert> how shocked are you when you see a one-man-salvation-army?
11:45:36  *** Speeder_ has quit IRC
12:01:28  *** HerzogDeXtEr1 has joined #openttd
12:07:38  *** HerzogDeXtEr has quit IRC
12:24:10  *** Herr_Bert has quit IRC
12:27:06  *** WormnestAndroid has quit IRC
12:27:19  *** WormnestAndroid has joined #openttd
12:54:12  *** juzza1 has quit IRC
13:11:06  *** andythenorth has joined #openttd
13:12:39  *** andythenorth has left #openttd
13:26:23  *** DasPoseidon has joined #openttd
13:40:48  *** DasPoseidon has quit IRC
13:41:04  *** DasPoseidon has joined #openttd
14:04:19  *** glx has joined #openttd
14:04:19  *** ChanServ sets mode: +v glx
14:21:57  *** skrzyp1 has joined #openttd
14:43:35  *** andythenorth has joined #openttd
14:47:03  *** nielsm has joined #openttd
14:58:27  *** jottyfan has joined #openttd
15:17:56  *** keoz has joined #openttd
15:29:29  <TrueBrain> LordAro: holy crap, storing a full checkout in artifacts is ... slow
15:29:44  <TrueBrain> ~1 minute .. damnnnn
15:30:48  <LordAro> TrueBrain: mhmm.
15:30:58  <LordAro> probably faster after i remove the .git dir
15:31:32  <TrueBrain> I did that in my run, makes no difference
15:31:48  <TrueBrain>
15:31:52  <TrueBrain> source bundles are created
15:31:56  <TrueBrain> docs bundle .... is acting up :P
15:32:34  <TrueBrain> looking at the errors, it makes a single connection for each file
15:32:39  <TrueBrain> ugh :P
15:35:00  *** rptr_ has joined #openttd
15:37:39  <TrueBrain> 3 seconds vs 60, if sending it over compressed as a single file :D haha :D
15:37:40  <TrueBrain> funny
15:50:59  <TrueBrain> okay, GHA groups are pretty nice .. makes organizing the output a lot easier
15:51:27  <orudge> Is there a parallel uploads option? DevOps Pipelines has one and it does speed things up I find when pushing lots of files
15:52:30  <TrueBrain> I honestly expected the actions to do something clever, especially as it is one made by GitHub
15:53:17  *** tokai|noir has joined #openttd
15:53:17  *** ChanServ sets mode: +v tokai|noir
15:53:34  <orudge> When it comes to the Mac workflow for GitHub Actions though, give me a shout, I can work on that. Might have a play with it myself at some point before then anyway, if I have the time
15:54:20  <TrueBrain> funny, Doxygen still stores information in "objs" :P
15:54:42  <TrueBrain> orudge: will do
15:55:35  <TrueBrain> lol, you cannot delete artifacts within a workflow
15:55:42  <TrueBrain> guess it will just be a low retention time :)
15:56:41  *** Flygon has quit IRC
15:59:56  *** tokai has quit IRC
16:00:39  <TrueBrain> <- open up "Build docs"
16:00:54  <TrueBrain> it now has 4 extra groups (the first one is always there)
16:01:04  <TrueBrain> shows a lot better what each blob of stdout/stderr is for :)
16:03:44  <glx> I think you need to also run the API generation rule
16:04:54  <TrueBrain> I am just copying what it is currently doing ;)
16:05:01  <TrueBrain> which was broken, annoyingly enough
16:06:14  <glx> warning: tag INPUT: input source '/home/runner/work/OpenTTD/OpenTTD/generated/script/api' does not exist <-- that's weird, I think it should have at least script_window.hpp
16:07:17  <glx> oh no it miss the "build" part in the path
16:08:09  <TrueBrain> that is my bad, fixed in next run
16:16:23  <glx> I guess it would be easier to add docs rule in cmake
16:16:42  <TrueBrain> it kinda would
16:16:47  <TrueBrain> it is ... complicated, atm
16:17:00  <TrueBrain> if you wouldn't mind looking into that, that would be great :)
16:17:54  <glx> I can try yes
16:22:15  *** andythenorth has left #openttd
16:26:00  <milek7> TrueBrain: your emscripten PR broke on tests ;P
16:26:27  <TrueBrain> horrible! :P
16:26:37  <TrueBrain> pfft, you need to login to see why
16:26:37  <TrueBrain> wtf
16:27:36  <milek7> _memset not defined
16:28:36  <TrueBrain> like I am going to know how to solve that, lolz :D
16:31:16  <TrueBrain> (was not meant against you, more about the completely unclearness to me how some things work :P)
16:32:36  <milek7> well, I don't know either (why _write_sockaddr is used as __write_sockaddr is mystery to me also) :P
16:35:32  <milek7> oh, maybe it needs _memset in _write_sockaddr__deps?
16:35:43  <TrueBrain> but other places that use memset don't have that either
16:35:46  <TrueBrain> that is what surprises me :P
16:35:55  <TrueBrain> and as this works locally, testing this is going to be a pita :P
16:39:05  <milek7> that is ugly, but.. you could treat sin_zero as i64 and use makeSetValue :P
16:41:27  <TrueBrain> and I can also set it 8 times on each index
16:41:37  <TrueBrain> but it is collecting hacks :P
16:43:50  <TrueBrain> grr, your patch conflicts with mine :! :P
16:46:07  <TrueBrain> guess I should also add a test, tbfh :P
16:47:37  <TrueBrain> lol, xz is by default not installed on ubuntu? Lol ...
16:52:16  <TrueBrain> adding "memset" to deps doesn't work .. that fails with: don't know memset! :P
16:53:36  <TrueBrain> nevermind,  I was being silly
16:54:01  <TrueBrain> now how to run tests .. hmmmmmm
16:59:57  *** Progman has joined #openttd
17:00:07  <_dp_> ok, not sure what I expected to find in the company name table but definitely not this xD
17:01:13  <TrueBrain> milek7: solved it by the i64 trick after all
17:02:40  <TrueBrain> and re-used your test, which was mine, so that is all good now :P :P :P
17:07:27  *** jottyfan has quit IRC
17:11:48  *** Wormnest has joined #openttd
17:18:47  <TrueBrain> bah, the Mac OS runs don't want to start on GHA :P At least .. they are slow :P
17:19:45  <TrueBrain> glx: <- if you can spot any other mistake, let me know
17:22:09  <TrueBrain> orudge: I integrated your Mac OS work in that release-flow, while splitting them up over 2 jobs. We can combine them if you like, but this just to keep the ball rolling
17:22:28  <TrueBrain> LordAro: I used your release-workflow as base; I did make a few modifications, as I found out that GHA has new toys :D
17:23:53  *** blathijs has quit IRC
17:24:20  <TrueBrain> I fixed the Windows build locally, but I am now just waiting for the Mac OS to start, so I can see if those work :P
17:25:01  <TrueBrain> 16.04 creates 1 artifact, the others 2 .. that is odd :P
17:26:08  <LordAro> TrueBrain: ooh
17:26:59  <TrueBrain> but boy, GitHub needs the ability to include other workflows ... this becomes ... difficult to maintain :D
17:27:23  *** blathijs has joined #openttd
17:27:29  <LordAro> yup
17:28:01  <TrueBrain> and grrr @ Mac OS ... provision already :P
17:28:19  <TrueBrain> fuck it, I want to see if I fixed Windows :D
17:28:49  <TrueBrain> ah, sha256 file .. right, those should be removed
17:29:02  <LordAro> i already did that!
17:29:40  <TrueBrain> yeah, it seems I removed that!
17:32:28  <TrueBrain> but funny that it doesn't do it on 16.04 :)
17:33:23  <TrueBrain> oops, that moment you make a tarfile named ".tar.gz" , which really is just a tar :P
17:37:02  <TrueBrain> the Mac runners really are out for a piss ... sad :P
17:37:47  <TrueBrain> things left are: recombining all the jobs, creating a manifest, generating shasums, and uploading to AWS :) Seems the rest worked fine, so nice work LordAro  :D
17:37:57  *** rptr_ has quit IRC
17:38:22  <TrueBrain> oeh, there is a mac runner starting up! Good, he can also unpack the source .. gzip it is :)
17:42:58  <LordAro> :)
17:49:39  *** andythenorth has joined #openttd
17:49:51  <andythenorth> so do we all have to buy RHEL now then?
17:49:58  * andythenorth mumble mumble
17:50:13  <LordAro> yes
17:53:15  <DorpsGek> [OpenTTD/OpenTTD] DorpsGek pushed 1 commits to master
17:53:16  <DorpsGek>   - Update: Translations from eints (by translators)
18:06:22  *** Progman has quit IRC
18:07:06  *** Progman has joined #openttd
18:21:28  *** frosch123 has joined #openttd
18:24:28  *** gelignite has joined #openttd
18:26:06  *** qwebirc55173 has joined #openttd
18:29:15  *** qwebirc55173 has quit IRC
18:30:34  <TrueBrain> oeh, found a "bug" in the Mac OS results ... their names are not unique :D
18:32:48  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8340: Draft: Feature: Create Universal (x86_64 + Apple Silicon) build on macOS
18:39:13  *** DasPoseidon has quit IRC
18:45:16  <TrueBrain> funny, if you did the "upload to AWS" a few times, it really is not a lot of work to do it yet-another-time :P
18:50:51  <TrueBrain> <- it sure does look pretty!
18:53:03  <frosch123> funny how the linux builds do not name an architecture
18:53:25  <frosch123> is there no ubuntu for mips or something?
18:56:39  <TrueBrain> old docker thing .. nothing == amd64
19:02:03  <DorpsGek> [OpenTTD/OpenTTD] twpol updated pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
19:07:48  * andythenorth needs to test
19:07:59  * andythenorth needs a nap
19:11:27  <DorpsGek> [OpenTTD/OpenTTD] twpol commented on issue #8095: Feature suggestion: option to automatically space out vehicles in a shared order
19:13:58  <andythenorth> the patch JGR implemented really does work but the UI is sub-optimal
19:17:30  <frosch123> is it the css?
19:17:55  *** DasPoseidon has joined #openttd
19:19:10  <andythenorth> the css is fine
19:19:16  <andythenorth> really clean
19:19:30  <andythenorth> hmm I built this patch, ffwd doesn't work at all
19:19:34  <andythenorth> I assume that's the same in trunk
19:19:45  * andythenorth has been playing official binaries recently
19:20:56  <andythenorth> the performance of a self-built JGR is noticeably better, but I didn't measure it
19:21:00  <TrueBrain> well, this is new: "hdiutil: create failed - Resource busy" ... hmm
19:51:07  <andythenorth> ok so the wasm client is faster than the self-built mac client
19:51:18  <andythenorth> 60fps vs 32fps
19:51:40  <andythenorth> I will test the separation patch in wasm then :P
19:52:36  <TrueBrain> ..... wuth?
19:52:42  <TrueBrain> did you build a debug build?
19:52:44  <TrueBrain> I guess you did
19:53:02  <TrueBrain> cd build && cmake -DCMAKE_BUILD_TYPE=Release ..
19:53:06  <TrueBrain> build that .. should have better fps :)
19:53:59  * andythenorth tests
19:54:23  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8340: Draft: Feature: Create Universal (x86_64 + Apple Silicon) build on macOS
19:58:19  *** Gustavo6046 has quit IRC
19:58:39  *** Gustavo6046 has joined #openttd
19:58:41  <andythenorth> 200fps :P
19:58:42  <andythenorth> thanks
19:58:48  <andythenorth> oh an assert :)
20:01:53  <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:03:16  <andythenorth> and again
20:03:19  <andythenorth> so it repros
20:11:24  *** HerzogDeXtEr has joined #openttd
20:11:43  <TrueBrain> Stadia is of the opinion I need a better internet connection
20:11:48  <TrueBrain> I on the other hand think they should upgrade
20:13:09  <andythenorth> hmm no assert in self built master
20:17:45  *** HerzogDeXtEr1 has quit IRC
20:31:29  <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:39:20  <DorpsGek> [OpenTTD/OpenTTD] twpol commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:40:34  <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:43:12  <DorpsGek> [OpenTTD/OpenTTD] twpol updated pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:47:45  <TrueBrain> LordAro: you no longer want to build for debian btw?
20:51:33  *** HerzogDeXtEr1 has joined #openttd
20:51:43  <TrueBrain> oops, all linux deb files are named the same
20:51:44  <TrueBrain> that is baddddd
20:55:06  <DorpsGek> [OpenTTD/OpenTTD] twpol commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:56:58  *** HerzogDeXtEr has quit IRC
20:59:23  <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8342: Feature #8095: Allow automatically separating vehicles in shared orders
20:59:39  *** sla_ro|master has quit IRC
21:00:34  <TrueBrain> and I am missing the debug debs LordAro  .. no clue how to generate those :P
21:03:54  <TrueBrain> not even sure if anyone ever uses them, and what is in it :D
21:06:47  <TrueBrain> CPack does seem to offer a setting to enable it .. ddeb? No clue ... the part I didn't want to go through again, so ... yeah ... walks away slowly :P
21:08:17  <andythenorth> that's how I feel about random triggers on dual-headed trains :P
21:08:42  <andythenorth> disable the random....walk away
21:16:52  *** HerzogDeXtEr1 has quit IRC
21:17:59  *** Delzur has joined #openttd
21:18:29  *** HerzogDeXtEr has joined #openttd
21:19:50  *** virtualrandomnumber has joined #openttd
21:20:06  <andythenorth> should separation patch try and warn that full load etc is incompatible with auto separation?
21:20:15  <andythenorth> or do we just let players discover that/
21:20:17  <andythenorth> ?
21:20:32  *** virtualrandomnumber has quit IRC
21:21:09  <frosch123> is it more incompatible with separation than with regular timetables?
21:23:09  <andythenorth> I have NFI
21:23:13  <andythenorth> I understand separation
21:23:20  <andythenorth> I don't understand timetables
21:23:51  <andythenorth> did I miss out the smiley?
21:24:25  *** HerzogDeXtEr has quit IRC
21:25:42  *** HerzogDeXtEr has joined #openttd
21:29:05  <andythenorth> I wondered about just extending the tool tip :)
21:29:18  <frosch123> i have no idea either
21:29:46  <frosch123> i guess you can mix fullload with timetabling some other orders
21:30:03  <frosch123> but timetabling the full load order is pretty pointless
21:30:21  <andythenorth> I am sure somebody has a carefully crafted savegame proving it works :)
21:30:23  <andythenorth> but eh
21:30:55  <andythenorth> the small-but-friendly-crowd on discord basically said "dunno how all that works I just use auto separate"
21:31:02  <andythenorth> when I asked about timetables
21:31:20  <frosch123> pretty sure they would say the same about transfer orders, if they used cargodist :)
21:32:22  <andythenorth> I use transfers with cdist :)
21:32:24  <andythenorth> am I wrong?
21:32:39  <frosch123> no idea, i used transfers only without cdist
21:32:43  <andythenorth> prevents weird backlinks getting formed when there are a lot of nodes and edges doing same cargo
21:33:10  <andythenorth> many-sources-many-destinations networks get weird in cdist without explicit orders
21:33:38  <frosch123> but well, if you want to know how stuff works, you need to find someone who exploits it
21:34:00  <frosch123> speedrunners may not be able to tell you how to have fun with a game, but they surely know the mechanics
21:34:43  <frosch123> ideally you have someone opinionated like V
21:35:07  <frosch123> V loves orders, refits and signals. but hates pbs and station-refit
21:35:21  <frosch123> probably also perfers transfers over cdist
21:35:29  <frosch123> i have no idea whether coop made anything with timetables
21:35:42  <frosch123> but they likely used signals to time trains :p
21:37:59  <frosch123> maybe display a popup when starting ottd the first time
21:38:58  <frosch123> choose: to you want to play the game? or do you want the game to play itself? this affects the availability of pbs, cdist, auto-separation, vs refitting, transfers, timetables and presignals
21:40:21  <frosch123> you have to uninstall ottd and delete all your savegames to be able to make the decision again
21:41:26  *** tokai has joined #openttd
21:41:26  *** ChanServ sets mode: +v tokai
21:48:17  *** tokai|noir has quit IRC
21:49:20  *** tokai|noir has joined #openttd
21:49:20  *** ChanServ sets mode: +v tokai|noir
21:51:57  <andythenorth> :)
21:52:25  <andythenorth> there are people sequencing entire models of countries with timetables
21:55:12  *** DasPoseidon has quit IRC
21:56:19  *** tokai has quit IRC
21:56:41  <andythenorth> having tested this patch a bit more
21:56:49  <andythenorth> I can see why JGR players say it's essential
21:57:36  *** jottyfan has joined #openttd
21:57:50  <andythenorth> strong words eh :)
22:02:39  *** frosch123 has quit IRC
22:17:12  *** jottyfan has quit IRC
22:27:12  <milek7> TrueBrain:
22:27:16  <milek7> looks like proper debugging :P
22:27:32  <TrueBrain> funny :D
22:30:19  *** gelignite has quit IRC
22:33:51  <_dp_> what I definitely wouldn't want to see is having to mindlessly enable autoseparation or whatever for every damn new route
22:34:01  <_dp_> if it ends up being more effecient than just full load
22:34:28  <andythenorth> what's the alternative? :)
22:35:09  <_dp_> idk, option to disable timetabling?
22:39:22  <_dp_> also did anyone check performance? implementation looks kinda fishy
22:40:00  <andythenorth> do we have a standard baseline for performance?
22:40:47  <_dp_> not much slower than it already is? :p
22:41:09  <TrueBrain> if people don't mind it being slower, what is the hurt?
22:41:25  <TrueBrain> I mean, Cyberpunk 2077 is not running on any modern PC at 144 fps ... does it mean it is bad? :P
22:41:48  <TrueBrain> <- LordAro  :D
22:41:49  <andythenorth> we have a jenkins / selenium job that tests branches against a baseline
22:41:52  <andythenorth> (web app)
22:41:53  <TrueBrain> it is not really uploading yet, but that is minor
22:42:23  <andythenorth> can we dump the fps to a log from the wasm build? o_O
22:42:31  <_dp_> TrueBrain, it would be bad if it lags the game into unplayble state in certain scenarios
22:42:40  <TrueBrain> "in certain scenarios" it already does
22:42:58  <_dp_> afaict autoseparation iterates every shared vehicle for every loading vehicle every tick which is kinda a lot
22:43:10  <TrueBrain> but don't you rather mean, that you don't want performance regression with it disabled?
22:45:02  <andythenorth> free M1 with every download :P
22:45:05  <_dp_> TrueBrain, I don't want anyone using it to dos servers :p
22:45:09  <andythenorth> problem solved, for next few PRs at least
22:45:13  <TrueBrain> so disable it on your server?
22:45:20  <TrueBrain> I mean, if we check back in OpenTTD's history
22:45:24  <_dp_> well, if there is an option, sure
22:45:25  <TrueBrain> YAPF is pretty cool
22:45:31  <TrueBrain> but do you remember the version before that?
22:45:35  <TrueBrain> it was HORRIBLE
22:45:44  <TrueBrain> but I would claim, that without that version, we would never had this
22:46:17  <TrueBrain> innovation and "perfect" rarely go side-by-side
22:46:18  <andythenorth> I keep playing, even though the mac performance is sometimes lolzy
22:46:29  <andythenorth> it's not unplayable, just annoying
22:46:48  <glx> ok it seems I can generate source doc via cmake, now trying api doc
22:47:08  <andythenorth> ha opening the train window still dumps the fps
22:47:11  <andythenorth> oh well
22:47:18  <andythenorth> stuff and things
22:47:34  <andythenorth> I don't know what JGR has actually done, but he has a list, and it is a shitload faster on mac
22:47:48  <glx> will be harder in msvc because input filter is currently bash+gawk
22:53:57  <LordAro> TrueBrain: nice.
22:57:06  <_dp_> ok, nvm, it caches stuff so performance should be alright
22:57:36  <andythenorth> \o/
22:57:40  <TrueBrain> LordAro: in Upload you can see all the files created
22:57:40  <andythenorth> let's merge it :)
23:00:00  *** keoz has quit IRC
23:03:23  <_dp_> omg, it uses floats :/
23:03:50  * andythenorth very bedtime
23:03:53  *** andythenorth has quit IRC
23:23:39  *** Progman has quit IRC
23:36:24  *** HerzogDeXtEr has quit IRC
23:42:44  <orudge> TrueBrain: nice, will look at the Mac stuff tomorrow.
23:42:56  <TrueBrain> orudge: in my PR is a suggestion for how to solve bundle names
23:43:00  <TrueBrain> not sure that is correct
23:43:05  <TrueBrain> I just needed them deduplicated :P
23:43:10  <TrueBrain> well, unique
23:43:14  <orudge> :)
23:43:40  <TrueBrain> I btw always build the build-host tools, even on x86_64 .. as .. that was easier
23:43:54  <TrueBrain> I am guessing something similar will be created for Windows too :D
23:46:23  <TrueBrain> owh, and the bundle name for Windows will break with ARM .. it is either win32 or win64 :D Ghehe
23:46:25  <TrueBrain> details ......
23:48:13  <orudge> Windows arm64 would just need the host tools thing, dependencies installed via vcpkg - arm64-windows-static
23:48:26  <orudge> I did test it in an emulator
23:48:43  <orudge> It worked, albeit very slowly under emu
23:48:53  <TrueBrain> Sounds easy enough :)
23:49:09  <TrueBrain> Well, in my branch that should be easy to add
23:49:18  <TrueBrain> Hope to finish it this weekend or so
23:50:34  <orudge> Nice one
23:50:44  <orudge> Will hopefully play tomorrow
23:51:03  <TrueBrain> Well, finish the current stuff; arm64 I leave to you
23:51:04  <orudge> Also hopefully my vcpkg PRs will be merged
23:51:37  <TrueBrain> (I will remove Mac OS arm tomorrow, just so we are not blocked by vcpkg:) l
23:52:50  <TrueBrain> I hope my emscripten PR gets accepted and they make a new release .. less hacks in OpenTTD :D
23:59:04  <_dp_> or they'll take example from openttd and wait a year :p

Powered by YARRSTE version: svn-trunk