Config
Log for #openttd on 3rd October 2021:
Times are UTC Toggle Colours
01:08:50  *** sim-al2 has joined #openttd
01:13:09  *** sim-al2 has quit IRC
01:32:36  *** Progman has quit IRC
02:11:04  *** Wuzzy has quit IRC
02:50:36  *** glx has quit IRC
03:12:33  *** WormnestAndroid has quit IRC
03:13:37  *** WormnestAndroid has joined #openttd
03:30:25  *** Wormnest has quit IRC
04:46:12  *** Speeder_ has joined #openttd
04:53:05  *** Speeder has quit IRC
05:58:36  *** nielsm has joined #openttd
06:31:56  *** andythenorth has joined #openttd
07:30:45  <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9598: Fix: Incorrect Romanian own name https://git.io/J2HKq
07:30:56  <DorpsGek> [OpenTTD/eints] LordAro approved pull request #61: Fix: Incorrect Romanian own name https://git.io/J2HKu
07:36:51  <DorpsGek> [OpenTTD/OpenTTD] James103 commented on pull request #9598: Fix: Incorrect Romanian own name https://git.io/J2HQf
07:57:46  *** sla_ro|master has joined #openttd
08:00:59  *** roadt__ has joined #openttd
08:07:40  *** roadt_ has quit IRC
08:08:50  <TrueBrain> right, because of a new Sentry error, I can indeed safely conclude that "Abort" mostly happens when people hammer the Refresh Server button
08:09:04  <TrueBrain> I wonder if we should disable it for the time the refresh is pending
08:09:16  <TrueBrain> basically what the GC does: cancel your last refresh request, and allow the new one
08:14:08  <LordAro> TrueBrain: probably, but GC should be as solid as possible as well
08:14:24  <LordAro> people might use a hacked client to 'break' the GC :p
08:15:35  <TrueBrain> it is solid in that sense
08:15:41  <TrueBrain> it just give a very high rate of "Aborted"
08:15:53  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #69: Fix: only delete the client token if we are the owner https://git.io/J27fw
08:16:37  <TrueBrain> and I now finally understand why .. people who love to press refresh a lot
08:16:55  <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #9599: [Crash]: An AI fills up the maximum memory allocation almost exactly (just a few bytes left) and crashes the game. https://git.io/J27UX
08:17:15  <LordAro> TrueBrain: fair
08:17:17  <LordAro> lol james
08:17:30  <TrueBrain> and I wonder what the best counter is against refresh-pressers
08:17:38  <TrueBrain> do I ignore the new attempt, or close the old
08:17:49  <LordAro> ignore, imo
08:17:59  <TrueBrain> the problem with that is: if you press refresh, and after that join
08:18:01  <TrueBrain> the join would be ignored
08:18:32  <LordAro> the client knows that it's waiting for the refresh to complete, right? just keep the button depressed until its completed
08:18:34  <TrueBrain> so I reasoned, as join give a popup showing the join, it is better to close the old .. but it is more resource intensive
08:18:46  <LordAro> though that doesn't help with refresh+join, i suppose
08:18:50  <TrueBrain> depressed? Not sure that is possible
08:18:56  <TrueBrain> but I can disable it while refresh is pending
08:19:07  <LordAro> i feel like it should be...
08:19:10  <TrueBrain> well, keeping it lowered should be possible too I guess
08:19:16  <TrueBrain> to the C++ code!
08:19:43  <Rubidium> maybe abort the refresh-presser's connection with a special "error" code, so sentry can distinguish between "impatient" and other issues
08:19:50  <TrueBrain> I can also just keep it disabled whatever for 10 seconds or something
08:20:00  <TrueBrain> this has nothing to do with sentry
08:20:03  <LordAro> might be easier
08:20:08  <TrueBrain> sentry showed me a bug which I fixed, nothing that is important btw
08:20:22  <TrueBrain> this is about these kind of people being really loud to the server
08:20:56  <TrueBrain> LordAro: which of the two comments "might be easier"? :D IRC sucks to see what you refer to :P
08:20:57  <LordAro> SetLowered seems to do what you want
08:21:04  <LordAro> ("keep it disabled")
08:21:32  <TrueBrain> do we want to keep it lowered for 10s?
08:22:03  <TrueBrain> kinda want two things: an indicator a refresh is still pending
08:22:08  <TrueBrain> and a preventing for people smashing that button
08:22:13  <TrueBrain> so lets do the first first
08:22:14  <TrueBrain> :D
08:23:30  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #69: Fix: only delete the client token if we are the owner https://git.io/J27fw
08:26:55  <TrueBrain> can't believe how trigger-happy people actually are :P
08:28:38  <TrueBrain> Rubidium: the thing I try to address is that currently if I press refresh 10x rapidly, the server also does 10 STUN requests in a short amount of time, etc :)
08:29:13  <TrueBrain> I picked up this information because of the stats the GC keeps, which shows a high amount of "aborted"; but till now I didn't really understand where it came from
08:29:26  <TrueBrain> a bug in Sentry accidentally showed me what was actually going on :D
08:35:18  <TrueBrain> LordAro: lowered state is an excellent idea
08:35:24  <TrueBrain> it feels really sensible
08:39:08  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9600: Add: [Network] Keep the refresh button in lowered state while refreshing https://git.io/J27wF
08:39:11  <LordAro> :)
08:41:03  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain created new tag: 1.4.1 https://git.io/J27rR
08:43:43  *** HerzogDeXtEr has joined #openttd
08:45:01  *** Samu has joined #openttd
08:46:00  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #70: Fix: only close connections to servers the instance owns https://git.io/J27Kk
08:46:07  <TrueBrain> new bug with every new release :P
08:46:44  <LordAro> lol
08:47:34  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #70: Fix: only close connections on shutdown which the instance owns https://git.io/J27Kk
08:48:28  <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9600: Add: [Network] Keep the refresh button in lowered state while refreshing https://git.io/J27KS
08:49:07  <DorpsGek> [OpenTTD/OpenTTD] LordAro closed issue #9597: [Bug]: Badly localized string for the language name "Română" https://git.io/J2gmu
08:49:10  <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9598: Fix: Incorrect Romanian own name https://git.io/J227m
08:49:35  <TrueBrain> so far only 1 entry for the title competition
08:49:44  <TrueBrain> guess we can release tomorrow after all? :P
08:50:21  <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J276u
08:53:15  *** Progman has joined #openttd
08:54:31  <TrueBrain> oof, many outdated "backport requested" labels
08:55:25  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick dismissed a review for pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J276u
08:55:28  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2RyN
08:55:46  <LordAro> TrueBrain: oh no
08:56:01  <TrueBrain> I am going to try to script this :D
09:00:49  <LordAro> huh, i am surprised there are that many
09:00:55  <LordAro> i thought we were keeping on top of them
09:01:07  <LordAro> (or rather, i was, given it's generally me doing the backporting)
09:01:12  <TrueBrain> those were planned for 1.11.3 (if that would ever happen)
09:01:25  <TrueBrain> so we should have wiped those labels when creating 12.0 branch :)
09:01:32  <TrueBrain> was not in the release procedure :P
09:01:55  <TrueBrain> not difficult to solve btw, just need to pay attention to the merge date
09:02:32  <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #9601: [Crash]: Assertion failed - An AI's `info.nut` fills up the maximum memory allocation almost exactly. https://git.io/J271B
09:02:35  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9600: Add: [Network] Keep the refresh button in lowered state while refreshing https://git.io/J27wF
09:06:51  <LordAro> james stahp
09:08:10  <TrueBrain> our in-house tester :) Finding bugs glx didn't :P :P :P
09:16:44  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8461: Feature: Permanent rivers https://git.io/JLS3A
09:20:18  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8609: Fix: Equalise the number of frames needed for road vehicles to traverse different radius curves https://git.io/JtnEF
09:31:18  *** Flygon has joined #openttd
09:33:13  <andythenorth> upstream to AWS oof
09:33:21  <andythenorth> maybe I could get a bitbucket pipeline to do it
09:33:27  <andythenorth> it's just pushing the compile of a tag
09:34:41  <TrueBrain> I told you before, but really: Cloudflare Pages :)
09:34:54  <TrueBrain> also reduces your AWS bill to 0
09:35:26  <andythenorth> stuff and things
09:35:38  <andythenorth> 5.2 KB/s upload though, GG
09:35:56  <andythenorth> it's like have my first 1 Mb/s cable connection
09:36:22  <andythenorth> 1999 called and wants its speed back
09:39:23  * andythenorth reading how cloudflare pages can do custom frameworks
09:41:01  <andythenorth> think I have to write a file telling it what deps to install
09:41:08  <andythenorth> seems like all the others
09:41:20  <andythenorth> then it's just 'make' and it serves pages
09:41:38  <andythenorth> no cloudfront, no route 53, no need to manually upload and invalidate CF
09:42:06  <TrueBrain> there is I reason I mention it ;)
09:43:09  <andythenorth> is the openttd main site now on Pages?
09:43:15  <TrueBrain> not yet
09:43:29  <andythenorth> well
09:44:59  *** Progman has quit IRC
09:45:08  <andythenorth> this looks like the best 'click here for site' so far
09:45:15  <andythenorth> Github Pages is frankly....weird
09:47:01  <andythenorth> the cloudflare flavour of lambdas looks interesting also https://workers.cloudflare.com/
09:48:32  <andythenorth> hmm
10:02:47  <TrueBrain> okay, got backporting mostly scripted
10:02:49  <TrueBrain> now for languages
10:02:56  <TrueBrain> where is that script we made for that .. hmm
10:04:49  <TrueBrain> LordAro: do you remember where that script is by any chance?
10:22:44  <TrueBrain> ha, found one of my own old scripts back for that
10:30:31  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2d3N
10:31:32  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9602: Backport master into release/12 https://git.io/J2dGb
10:32:29  <TrueBrain> that seems to work fine, sweet
10:35:09  <TrueBrain> https://gist.github.com/TrueBrain/bd73f25d687008d1b9497df422a979fc as a WIP storage of the scripts used :)
10:43:48  <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2d15
10:46:33  <TrueBrain> Lol .. so we suck in naming functions, nice
10:50:40  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2dxx
10:51:29  <TrueBrain> I still dont understand 9591
10:53:14  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2Ffi
10:54:22  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9594: Fix #9591: Update station docking tiles upon placing a water object on a docking tile https://git.io/J2RyN
10:54:25  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #9591: [Bug]: Building a water object on a plain water docking tile does not correctly remove the docking tile https://git.io/J2cXx
10:56:18  *** gelignite has joined #openttd
11:00:37  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9602: Backport master into release/12 https://git.io/J2Fc5
11:08:52  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #9590: [Bug]: strgen should be integrated with the translation system https://git.io/J2ZMG
11:12:11  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #9590: [Feature Request]: allow for easier language preview from Translator to in-game https://git.io/J2ZMG
11:12:36  <TrueBrain> shitty tittle still, but at least it doesn't it describes what the user wants, instead of how he thinks we should do it :)
11:14:35  <TrueBrain> doesn't = does
11:14:36  <TrueBrain> lol
11:18:30  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #9593: [Bug]: Assertion fails when loading the save while missing a newgrf https://git.io/J28qA
11:18:38  <TrueBrain> ^^ is not a bug, I think. Would love a second opinion
11:37:41  <TrueBrain> very curious what our users will think about the moving title game .. especially the pixel-based panning can be a bit annoying to see, I noticed :)
11:37:56  <TrueBrain> and zoomed-in shots look weird, as in: not used to it :)
11:39:09  <nielsm> it probably needs some practice to make really good ones
11:39:35  <nielsm> while working on mine I noticed it would probably be good to have a built-in way to ease over corners in a pan-sequence
11:39:55  <nielsm> some kind of curve interpolation, maybe not quite bezier but similar
11:41:13  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9603: Doc: Prepare for 12.0 release & title game https://git.io/J2bDo
11:41:32  <TrueBrain> it also reminds me I really want subpixel movements :P
11:41:34  <TrueBrain> but one thing at the time
11:43:10  <TrueBrain> lets see if creating a preview out of that PR is possible :)
11:43:16  <TrueBrain> I really do not know :P
11:44:38  *** kaomoneus has joined #openttd
11:45:21  <TrueBrain> right, guess I might as well write a news announcement, as I am rather bored today :P
11:47:22  <kaomoneus> Hi there! I want to intruce new string. What is default string localization? Is it english?
11:48:11  <kaomoneus> set(MASTER_LANG_FILE         ${CMAKE_CURRENT_SOURCE_DIR}/english.txt )
11:48:27  <kaomoneus> Joined IRQ to answer question myself, haha!
11:48:45  <nielsm> yes, and remember it's british english
11:49:21  <nielsm> also, you should only add the new string to english.txt, the other translations will be handled by the web translator tool
11:49:34  <nielsm> even if you know one of the other languages, do not add the string to it
11:50:12  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9603: Doc: Prepare for 12.0 release & title game https://git.io/J2bDo
12:01:20  <Samu> i just resolved conflicts a few hours ago, and now there's another
12:01:22  *** frosch123 has joined #openttd
12:01:25  <Samu> :(
12:01:55  <DorpsGek> [OpenTTD/website] TrueBrain opened pull request #226: Add: 12.0 announcement https://git.io/J2NEm
12:02:25  *** sla_ro|master has quit IRC
12:02:36  <TrueBrain> hmm, guess I should remove the GameScript reference from the intro game too, lol
12:05:09  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8609: Fix: Equalise the number of frames needed for road vehicles to traverse different radius curves https://git.io/JtnEF
12:05:49  <frosch123> i always read "rotating views" as "rotate map" :p
12:06:01  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9603: Doc: Prepare for 12.0 release & title game https://git.io/J2bDo
12:06:23  <TrueBrain> better wording is welcome ;)
12:07:03  <LordAro> mm, "changing views" would be better
12:07:41  <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8461: Feature: Permanent rivers https://git.io/JLS3A
12:08:29  <kaomoneus> frosch123: sorry for being meddled, are you working on rotating game view?
12:08:41  <nielsm> no
12:08:45  <nielsm> I can tell you that
12:09:13  <frosch123> noone is working on rotating game view, except maybe that opengl guy
12:09:14  <nielsm> TrueBrain: better wording may be "moving camera on title screen background"
12:09:36  <nielsm> or "multiple camera angles"
12:09:41  <DorpsGek> [OpenTTD/website] James103 commented on pull request #226: Add: 12.0 announcement https://git.io/J2NDr
12:10:11  <FLHerne> nielsm: I think 'multiple camera angles' also gives the wrong idea
12:10:23  <nielsm> hmm yeah
12:10:38  <kaomoneus> frosch123: I think it is interesting, who is that "opengl guy"?
12:10:47  <FLHerne> kaomoneus: The problem with rotating the world view is that the sprites for buildings just don't exist :p
12:11:01  <nielsm> https://www.tt-forums.net/viewtopic.php?f=33&t=86412 this OpenGL
12:11:10  <FLHerne> OpenGFX could get some, but it would break all the existing grfs
12:11:20  <nielsm> yeah that's the big problem you have 20 years worth of add-ons that won't work with rotation
12:11:42  <kaomoneus> FLHerne: yep, I agree. I still have couple of hopes though
12:11:57  <FLHerne> Maybe in another 20 years we can have an AI that predicts what the back-side of building sprites look like :D
12:12:01  <kaomoneus> 1. Perhaps the latest sprites are based on 3d models.
12:12:07  <FLHerne> I've seen something like that
12:12:32  <TrueBrain> lol @ james's suggestion .. does anyone mind if I do NOT use those idiotic quotes? :P :P
12:12:38  <TrueBrain> I never understood why they even exist ..
12:12:39  <nielsm> yes some sprite artists do use 3D models as base, but many also don't and prefer hand-drawing
12:13:07  <kaomoneus> 2. There is a nice progress with DNN algrorithms, which might create remaining sides of all building. It might be with some noice or even weird and inadequate, and yet might be a good starting point.
12:13:59  <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #226: Add: 12.0 announcement https://git.io/J2NEm
12:14:14  <FLHerne> TrueBrain: yeah, just remove them IMO
12:14:40  <FLHerne> kaomoneus: current things seem to really struggle with pixel art
12:14:46  <TrueBrain> FLHerne: that works too
12:15:00  <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #226: Add: 12.0 announcement https://git.io/J2NEm
12:15:20  <FLHerne> I suppose a network specifically designed/trained for it might help, but then you have a dataset problem
12:15:46  <DorpsGek> [OpenTTD/website] TrueBrain commented on pull request #226: Add: 12.0 announcement https://git.io/J2Npt
12:16:12  <FLHerne> there are very few (by NN training standards) existing isometric building sprites in an OTTD-like scale with all angles drawn
12:17:03  <TrueBrain> hmm, emscripten doesn't seem to load the title game .. oh-oh
12:17:22  <TrueBrain> "Internal error: Loader for 'lzma' is not available."
12:17:23  <TrueBrain> hmmm
12:17:35  <TrueBrain> who broke LZMA ..
12:18:03  <frosch123> FLHerne: if you account for lightning, there are zero :p
12:18:07  <frosch123> not even terrain sprites
12:18:24  <frosch123> s/lightning/shading/ ?
12:20:31  <TrueBrain> ah, no, the github workflow doesn't apply the lzma patch .. right, explains a thing or two :)
12:22:40  *** HerzogDeXtEr has quit IRC
12:31:32  <FLHerne> frosch123: well, there are some in similar games like SimCity 2000 and some Simutrans paks
12:31:54  <FLHerne> I don't think SC2k rotates the lighting properly either
12:32:07  <nielsm> simcity 2000 mostly just mirrors and reuses sprites
12:32:18  <nielsm> and has every building be square
12:33:06  <frosch123> FLHerne: yep, it doesn't. and that's the reason why i never rotated the map in s2k... whenever i rotate i get completely lost
12:33:22  <frosch123> i do no longer recognize things, it's like an entirely different map :p
12:42:35  <kaomoneus> Guys, is it intended to use tabs instead of spaces?
12:43:24  <nielsm> depends on where
12:43:39  <nielsm> line indentation in the C++ code is with tabs
12:46:54  *** roadt__ has quit IRC
12:47:14  *** roadt__ has joined #openttd
13:05:27  *** glx has joined #openttd
13:05:27  *** ChanServ sets mode: +v glx
13:09:17  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9604: Change: upgrade Emscripten to 2.0.31 and enable LZMA by default https://git.io/J2xbP
13:09:22  <TrueBrain> lets see if this actually works .. difficult to test :D
13:09:50  <glx> ok so I have 2 reports to analyse
13:10:04  <TrueBrain> James has been busy finding bugs in your code, yes :P
13:10:17  <glx> but they may already be fixed
13:10:57  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9604: Change: upgrade Emscripten to 2.0.31 and enable LZMA by default https://git.io/J2xbP
13:11:23  <TrueBrain> right, emscripten works, sweet :)
13:11:23  <glx> especially #9599, looks very similar to #9548
13:11:42  <glx> and the crashlog being from jgrpp
13:12:52  *** kaomoneus has quit IRC
13:15:43  <LordAro> why does jgrpp have the crashlog and we don't?
13:17:08  *** gelignite has quit IRC
13:17:56  <TrueBrain> asking the real questions!
13:20:26  <TrueBrain> while talking about crashes, is https://github.com/OpenTTD/OpenTTD/issues/9593 a bug or expected behaviour?
13:21:03  <glx> I'd say expected, we even have a special crash message
13:21:53  <glx> but could also be a weird thing in NRT specs
13:22:23  <TrueBrain> at least it is not a regression :)
13:32:32  <glx> ok #9601 doesn't exactly do the same in a a2cf81e722 debug build as in the report, but it seems to be a duplicate of #9548
13:42:25  <glx> same for #9599
13:42:35  <glx> let's test both in current master
13:52:36  <TrueBrain> hmm, seems Cloudflare Pages always deploy on pushes into main
13:52:41  <TrueBrain> to production
13:52:43  <TrueBrain> and not on tag
13:52:46  <TrueBrain> that is slightly annoying
13:52:49  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on issue #9601: [Crash]: Assertion failed - An AI's `info.nut` fills up the maximum memory allocation almost exactly. https://git.io/J271B
13:52:52  <DorpsGek> [OpenTTD/OpenTTD] glx22 closed issue #9601: [Crash]: Assertion failed - An AI's `info.nut` fills up the maximum memory allocation almost exactly. https://git.io/J271B
13:53:01  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on issue #9599: [Crash]: An AI fills up the maximum memory allocation almost exactly (just a few bytes left) and crashes the game. https://git.io/J27UX
13:53:04  <DorpsGek> [OpenTTD/OpenTTD] glx22 closed issue #9599: [Crash]: An AI fills up the maximum memory allocation almost exactly (just a few bytes left) and crashes the game. https://git.io/J27UX
13:54:18  *** gelignite has joined #openttd
13:54:35  <TrueBrain> nice glx, that was easy :)
13:54:52  <glx> I recognised the pattern :)
13:55:47  <glx> and debug build don't even reach the assert, crt debug free fails before
13:56:52  <glx> but both test ai were reaching meomory limit in realloc
14:00:51  <TrueBrain> okay, so Cloudflare Pages basically is Netlify .. at least, I get errors pointing to their docs :P
14:01:20  <TrueBrain> I have to say, it is only because our website needs some special love, but Cloudflare Pages does make deploying your website rather trivial
14:02:36  <TrueBrain> just too bad it doesn't know "staging" from "production"
14:09:30  <TrueBrain> I wonder if I can trick this .. hmm .. lets try
14:19:32  <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #9604: Change: upgrade Emscripten to 2.0.31 and enable LZMA by default https://git.io/J2j8T
14:20:10  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9604: Change: upgrade Emscripten to 2.0.31 and enable LZMA by default https://git.io/J2xbP
14:21:58  <glx> hmm maybe we should check https://github.com/actions/virtual-environments/issues/4060 doesn't affect us :)
14:22:16  <TrueBrain> read: someone should test a nightly :)
14:22:47  <TrueBrain> ah, no, those aren't upgraded yet
14:23:33  <glx> I could upgrade the workflow in my fork (at least to check it still builds)
14:25:22  <TrueBrain> let's just hope we can do 12.0 before they upgrade
14:25:30  <TrueBrain> at least we know it works as it is now :)
14:26:33  <glx> https://github.com/glx22/OpenTTD/actions/runs/1300681195 <-- queued (because I pushed the rebase right before adding my test commit)
14:27:32  <glx> ok no vcpkg cache found (expected)
14:32:15  <glx> looking at the issue, they removed python 2.7, I can see how some workflows may be affected (not ours of course)
14:33:43  *** jottyfan has joined #openttd
14:36:11  <glx> at least CI builds and regression pass
14:36:44  <glx> we "should" be safe
14:36:56  <TrueBrain> lets not risk it if we can help it :)
14:40:25  <glx> I guess I could make a test PR (not to be merge, keeping macOS-latest is ok in any case), but allows to publish test build so we have something to test for apple users
14:42:24  <TrueBrain> go for it
14:44:14  <glx> oh releases are currently forced to 10.15 image (not a bad idea ;))
14:47:11  *** kaomoneus has joined #openttd
14:48:31  <kaomoneus> https://github.com/OpenTTD/OpenTTD/blob/3046a6ce39d755f50207a2c4a1235069a0eb23fb/src/economy.cpp#L1156
14:48:39  <kaomoneus> is min used to prevent overflow?
14:50:21  <nielsm> kind of
14:51:05  <nielsm> it prevents the industry from having more than 0xFFFF units waiting, but it's not used to prevent integer overflow in the regular sense
14:51:41  <nielsm> compare it to saturating arithmetic rather, the calculation is performed as 32 bits but the result is capped to 16 bits
14:52:07  <kaomoneus> OK.
14:52:34  <DorpsGek> [OpenTTD/OpenTTD] glx22 opened pull request #9605: Change: [Actions] Use macOS-11 image for releases https://git.io/Jaea0
14:57:35  *** Progman has joined #openttd
15:14:06  <glx> hmm manual trigger with PR ref won't work I think
15:14:55  <glx> same issue you had with emscripten PR
15:16:25  <TrueBrain> what do you mean?
15:16:50  <glx> use "workflow from" will be master
15:17:37  <TrueBrain> hehe, yeah, you can only test that on your fork :)_
15:17:38  <glx> and I need to use workflow from the PR itself
15:18:12  <glx> but on my fork I can't publish binaries
15:18:35  <TrueBrain> they are in the artifacts I think
15:18:42  <TrueBrain> but yeah .. chicken-egg :D
15:18:49  <TrueBrain> lets wait till after 12.0
15:18:53  <TrueBrain> we can just merge it after that
15:18:56  <TrueBrain> and see what happens
15:18:58  <TrueBrain> its fine
15:19:12  <TrueBrain> https://www-test.openttd.org/ <- running from my fork, but I think I got it to work how I would like
15:19:13  <glx> yeah we can break future nightlies
15:19:17  <TrueBrain> the only issue is, the staging URL is a bit weird
15:20:12  <TrueBrain> (as that is https://main.openttd-website.pages.dev/ )
15:24:24  <TrueBrain> but I think this is a lot better to use, as Cloudflare Pages invalidates the cache
15:24:43  <TrueBrain> right, lets move it from my fork to the main repo, and just shadow-run it for a bit
15:33:44  *** sla_ro|master has joined #openttd
15:37:10  <TrueBrain> https://www-staging-test.openttd.org/ <- ha, got it to work :)
15:37:39  <kaomoneus> TrueBrain: does emscripten work well with OpenTTD servers now?
15:37:53  <TrueBrain> not more or less than before
15:38:19  <DorpsGek> [OpenTTD/website] TrueBrain opened pull request #227: Add: [Cloudflare Pages] automation to deploy website to Cloudflare Pages https://git.io/Jafqu
15:38:33  <TrueBrain> that should enable CF, next to AWS, so we can test it out :)
15:39:04  <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #227: Add: [Cloudflare Pages] automation to deploy website to Cloudflare Pages https://git.io/Jafqu
15:40:57  <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #227: Add: [Cloudflare Pages] automation to deploy website to Cloudflare Pages https://git.io/Jafqu
15:41:17  <TrueBrain> hmm .. for some reason, PRs aren't being build ..
15:41:25  <TrueBrain> guess they have to be in the upstream repo for that to work?
15:42:56  <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #226: Add: 12.0 announcement https://git.io/J2NEm
15:43:01  <TrueBrain> yeah, that is it
15:43:02  *** Wormnest has joined #openttd
15:48:46  <TrueBrain> https://github.com/OpenTTD/website/pull/226 <- click Show all checks, click Details on Cloudflare Pages, go to the URL :D
15:48:49  <TrueBrain> PR previews ;)
15:49:04  <TrueBrain> it can also post a comment to tell you that, but that sounds noisy ..
15:50:04  <glx> nice
15:50:25  <TrueBrain> but most of all: deployments will be visible instantly on Cloudflare, instead of a 2 hour cache
15:51:05  <TrueBrain> so after someone reviewed #227, we will have to wait 24 hours to see if the automation kicks in correctly .. and after that, we can flip AWS with Cloudflare Pages :D
15:54:37  <DorpsGek> [OpenTTD/website] glx22 approved pull request #227: Add: [Cloudflare Pages] automation to deploy website to Cloudflare Pages https://git.io/Jafyf
15:55:39  <DorpsGek> [OpenTTD/website] TrueBrain merged pull request #227: Add: [Cloudflare Pages] automation to deploy website to Cloudflare Pages https://git.io/Jafqu
15:56:33  <TrueBrain> Tnx :D
16:13:22  *** kaomoneus has quit IRC
16:29:40  <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on issue #9590: [Feature Request]: allow for easier language preview from Translator to in-game https://git.io/J2ZMG
16:30:05  *** Heiki has quit IRC
16:36:54  *** Heiki has joined #openttd
16:45:13  <DorpsGek> [OpenTTD/OpenTTD] michicc commented on issue #9590: [Feature Request]: allow for easier language preview from Translator to in-game https://git.io/J2ZMG
16:53:56  *** geli has joined #openttd
16:55:17  *** virtualrandomnumber has joined #openttd
16:57:49  *** jellyknight has joined #openttd
16:59:17  *** virtualrandomnumber has quit IRC
17:00:19  *** gelignite has quit IRC
17:04:19  *** geli has quit IRC
17:05:40  *** snail_UES_ has quit IRC
17:11:28  *** Flygon has quit IRC
17:38:33  *** Progman has quit IRC
17:39:58  *** tokai has joined #openttd
17:39:58  *** ChanServ sets mode: +v tokai
17:47:05  *** tokai|noir has quit IRC
17:57:09  <TrueBrain> any chance anyone can take a look at https://github.com/OpenTTD/OpenTTD/pull/9603 ? They are just backports :)
17:57:14  <TrueBrain> euh, wrong one
17:57:18  <TrueBrain> https://github.com/OpenTTD/OpenTTD/pull/9602
17:57:23  <TrueBrain> means I can finish my scripting tonight ;)
17:58:26  *** Progman has joined #openttd
17:59:15  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9605: Change: [Actions] Use macOS-11 image for releases https://git.io/JakHq
18:02:59  <andythenorth> hmm
18:03:05  <andythenorth> such grf things
18:03:10  * andythenorth drawing trains
18:04:46  <andythenorth> player choice for alternative train sprites when? :P
18:05:06  <TrueBrain> we call that trains, yes
18:05:10  <TrueBrain> since 1994
18:05:55  *** WormnestAndroid has quit IRC
18:07:08  <andythenorth> very important that player can choose train skin
18:07:13  <andythenorth> missing feature
18:07:15  <andythenorth> pls add
18:07:26  * andythenorth back to drawing
18:07:34  <TrueBrain> only via lootboxes
18:07:57  *** WormnestAndroid has joined #openttd
18:13:23  *** Wuzzy has joined #openttd
18:20:57  <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9602: Backport master into release/12 https://git.io/JaIan
18:21:54  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9602: Backport master into release/12 https://git.io/JaIwC
18:24:05  <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #9602: Backport master into release/12 https://git.io/JaIiU
18:24:28  <TrueBrain> <3
18:25:17  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9602: Backport master into release/12 https://git.io/J2dGb
18:34:03  <nielsm> idea for new main menu layout: https://0x0.st/-g_G.png
18:35:25  <TrueBrain> not bad
18:35:37  <TrueBrain> although I did become a strong believer all settings should be in a single menu
18:35:44  <TrueBrain> especially options and settings should be merged :P
18:35:59  <TrueBrain> (it doesn't have to be in the same list, mind you, but in about the same place)
18:36:10  <frosch123> also, settings which affect only new games should be in the new-game window
18:36:13  <TrueBrain> the line between options and settings is incredibly thin, and often confusing :)
18:36:21  <nielsm> I think it's okay to separate settings that affect game rules from settings that only affect client behaviour
18:36:29  <frosch123> people are always confused that changing newgrf/whatever in main menu does not affect their game
18:36:41  <nielsm> but e.g. town names does affect game rules and shouldn't be where it is
18:37:05  <TrueBrain> I think we all say the same here :)
18:37:06  <nielsm> true, NewGRF and AI/GS setup should probably be put into the New Game window
18:37:12  <frosch123> TrueBrain: haha, the rule is exactly: options are those settings that do not work in the settings tree, because of implementation limitations
18:37:29  <TrueBrain> frosch123: I like how that is obvious to players indeed :P
18:40:08  <nielsm> also merge the Play Heightmap, and maybe Play Scenario, into New Game too, so it basically becomes a tabbed window where you choose how to generate the world
18:40:16  <frosch123> https://gist.githubusercontent.com/frosch123/889b6478a6d5d7b10f89d9b014845331/raw/b11d0d479c773153722a617b3d7b8c1d0041f57a/guiplan202012.png <- nielsm: the black text at the bottom were meant to be buttons
18:40:36  <frosch123> like, put only the importan settings into mapgen gui, and then link to a filtered tree via "more map settings"
18:43:06  <TrueBrain> and if we are sharing inspiration: https://user-images.githubusercontent.com/1663690/109415102-02547e80-79b7-11eb-8f2c-3a3a324d07a6.png
18:43:10  <TrueBrain> https://user-images.githubusercontent.com/1663690/109415702-77758300-79ba-11eb-8ebf-d86ba9a4fbd1.png
18:43:10  <TrueBrain> https://user-images.githubusercontent.com/1663690/109419810-93375400-79cf-11eb-97a2-fbcd6151a631.png
18:43:15  <TrueBrain> for what-ever it is worth :)
18:43:38  <TrueBrain> Feb 28 .. lol, that has been in the freeze for a while now :P
18:45:16  <nielsm> what kinds of settings do we have... things that only affect the client (like basesets, news message settings), things that are client-driven but affect defaults for what the player does in-game (new orders non-stop by default, train stopping position), company settings, game rules settings, world generation settings
18:46:19  <DorpsGek> [OpenTTD/OpenTTD] DorpsGek pushed 1 commits to master https://git.io/JaLRl
18:46:20  <DorpsGek>   - Update: Translations from eints (by translators)
18:47:07  <TrueBrain> https://gist.github.com/TrueBrain/976cab3b6c88f208636653191ba5c420 might help with that mindset
18:47:28  <TrueBrain> https://gist.github.com/TrueBrain/976cab3b6c88f208636653191ba5c420#what-does-a-preset-include I meant to link
18:48:33  <TrueBrain> w00p, my backport scripts work :D Now I don't have to think about all that shit anymore :P
18:50:47  <TrueBrain> nielsm: either way, the "continue game" button I am most looking forward to us implementing :D
18:50:53  <TrueBrain> I so love that in other games I play
18:51:32  *** Wuzzy2 has joined #openttd
18:51:58  <nielsm> hmm... random thought: compare the worldgen seed, if it's even stored in games, to highlight savegames from the same world when loading a game
18:52:21  <TrueBrain> that is a lot of games to load :)
18:52:34  <TrueBrain> but grouping of savegames is not a bad idea
18:52:55  <nielsm> I think we already do scan file headers on all savegames listed in the window?
18:52:56  <frosch123> i think we already have a unique savegame id
18:53:14  <frosch123> isn't there some specific confirm overwrite in savegame window?
18:53:35  <frosch123> nielsm: only of the selected one
18:54:02  <nielsm> hmm okay... the original game did scan all the files to get the long names though
18:55:02  <TrueBrain> we wanted a cache for shit anyway, right? :D
18:55:08  <TrueBrain> scanning NewGRFs every time is also a bit .. much
18:55:11  *** Wuzzy has quit IRC
19:15:03  <DorpsGek> [OpenTTD/OpenTTD] glx22 merged pull request #9605: Change: [Actions] Use macOS-11 image for releases https://git.io/Jaea0
19:16:33  *** Wuzzy3 has joined #openttd
19:19:52  <glx> but how do you detect new files at start without a full rescan ?
19:20:11  *** Wuzzy2 has quit IRC
19:20:26  <glx> at least known newgrf may not need to be fully checked
19:22:49  <TrueBrain> modified time ;)
19:23:23  <glx> tar files, and their content list, are a good candidate for caching too
19:25:31  <glx> and someday I'll try to analyse script scans, to see why there are at least 5 full script rescan at launch
19:26:27  <TrueBrain> haha, yeah .. that would be an interesting deep dive :)
19:26:30  <glx> easy to spot with an info.nut reaching memoy limit
19:26:51  <glx> it's reported 2 times before newgrf scan, then 3 more times after
19:27:34  <glx> I think intro game is loaded 2 times too
19:27:50  <glx> so already 2 script scan I guess
19:28:20  <glx> then there is the initial empty game generation IIRC (another scan)
19:29:30  *** HerzogDeXtEr has joined #openttd
19:47:38  <TrueBrain> lol, the Cloudflare Pages removes .html from the URL
19:47:42  <TrueBrain> it does redirect properly
19:47:44  <TrueBrain> but that is unexpected
19:48:38  <TrueBrain> other than that, it seems to work as expected .. staging and production got updated as intended
19:58:58  *** WormnestAndroid has quit IRC
19:59:58  *** WormnestAndroid has joined #openttd
20:10:01  *** nielsm has quit IRC
20:14:46  <andythenorth> \o/
20:41:22  *** _aD has joined #openttd
20:45:08  *** jellyknight has quit IRC
21:02:39  *** sla_ro|master has quit IRC
21:06:19  *** Gustavo6046 has quit IRC
21:18:48  *** tokai|noir has joined #openttd
21:18:48  *** ChanServ sets mode: +v tokai|noir
21:20:06  *** Samu has quit IRC
21:25:50  *** tokai has quit IRC
21:35:41  *** Gustavo6046 has joined #openttd
21:38:28  *** Progman has quit IRC
21:42:30  *** frosch123 has quit IRC
21:44:07  *** Speeder_ has quit IRC
21:54:42  *** Gustavo6046 has quit IRC
21:54:45  *** Gustavo6046 has joined #openttd
22:05:00  *** jottyfan has quit IRC
22:06:33  *** Wuzzy2 has joined #openttd
22:07:11  *** Wuzzy3 has quit IRC
22:23:19  *** andythenorth has quit IRC
22:29:08  *** HerzogDeXtEr has quit IRC
22:49:22  *** tokai has joined #openttd
22:49:22  *** ChanServ sets mode: +v tokai
22:56:10  *** tokai|noir has quit IRC

Powered by YARRSTE version: svn-trunk