Config
Log for #openttd on 1st June 2021:
Times are UTC Toggle Colours
01:12:49  *** Wormnest has quit IRC
03:12:24  *** glx has quit IRC
03:56:14  *** Kitrana1 has joined #openttd
04:03:54  *** Kitrana has quit IRC
04:27:28  *** tokai|noir has joined #openttd
04:27:28  *** ChanServ sets mode: +v tokai|noir
04:34:24  *** tokai has quit IRC
04:39:59  *** Flygon has joined #openttd
05:21:46  *** Eddi|zuHause has quit IRC
06:16:13  *** snail_UES_ has quit IRC
06:25:26  *** snail_UES_ has joined #openttd
06:33:28  *** snail_UES_ has quit IRC
06:33:57  <DorpsGek> [OpenTTD/OpenTTD] danuker commented on issue #8206: Vehicle lists "Manage list" options inconsistent https://git.io/JfM79
06:39:30  *** Eddi|zuHause has joined #openttd
07:20:44  *** snail_UES_ has joined #openttd
07:28:29  *** HerzogDeXtEr has joined #openttd
07:30:42  *** gelignite has joined #openttd
07:35:13  *** snail_UES_ has quit IRC
08:00:47  *** sla_ro|master has joined #openttd
08:15:52  *** snail_UES_ has joined #openttd
08:31:25  *** Xaroth has quit IRC
08:32:28  *** snail_UES_ has quit IRC
08:32:40  *** Xaroth has joined #openttd
08:36:13  *** andythenorth has joined #openttd
08:40:24  *** tokai has joined #openttd
08:40:25  *** ChanServ sets mode: +v tokai
08:47:14  *** tokai|noir has quit IRC
08:52:05  *** WormnestAndroid has quit IRC
08:52:18  *** WormnestAndroid has joined #openttd
08:52:52  *** WormnestAndroid has quit IRC
08:53:09  *** WormnestAndroid has joined #openttd
08:59:20  *** sla_ro|master has quit IRC
09:12:29  *** snail_UES_ has joined #openttd
09:29:44  *** gelignite has quit IRC
09:31:13  *** snail_UES_ has quit IRC
09:37:25  *** tokai|noir has joined #openttd
09:37:25  *** ChanServ sets mode: +v tokai|noir
09:44:09  *** tokai has quit IRC
09:48:19  *** tokai has joined #openttd
09:48:19  *** ChanServ sets mode: +v tokai
09:55:13  *** tokai|noir has quit IRC
10:15:21  *** snail_UES_ has joined #openttd
10:21:46  <DorpsGek> [OpenTTD/OpenTTD] MaXenaeL opened issue #9326: Bug Report https://git.io/JGE4x
10:22:03  <DorpsGek> [OpenTTD/nml] PeterN commented on pull request #212: Fix #209: Sign-extend value before cargo_profit callback result calculation https://git.io/JGE4p
10:23:32  <LordAro> might be zbase issue?
10:24:09  *** andythenorth has quit IRC
10:25:48  *** snail_UES_ has quit IRC
10:27:36  <DorpsGek> [OpenTTD/OpenTTD] James103 commented on issue #9326: Bug Report https://git.io/JGE4x
10:29:43  <LordAro> thanks james
10:32:56  <peter1138> No "might" about it.
10:34:20  <LordAro> of course
10:36:49  <DorpsGek> [OpenTTD/OpenTTD] MaXenaeL commented on issue #9326: Bug Report https://git.io/JGE4x
10:39:21  *** tokai|noir has joined #openttd
10:39:21  *** ChanServ sets mode: +v tokai|noir
10:42:44  <TrueBrain> What is long time version? :D
10:46:03  *** tokai has quit IRC
11:03:10  *** snail_UES_ has joined #openttd
11:22:13  *** snail_UES_ has quit IRC
11:58:12  *** snail_UES_ has joined #openttd
12:07:54  <DorpsGek> [OpenTTD/OpenTTD] PeterN closed issue #9326: Bug Report https://git.io/JGE4x
12:07:57  <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on issue #9326: Bug Report https://git.io/JGE4x
12:18:03  *** snail_UES_ has quit IRC
12:29:05  *** Samu has joined #openttd
12:34:41  *** glx has joined #openttd
12:34:41  *** ChanServ sets mode: +v glx
12:38:43  *** snail_UES_ has joined #openttd
12:43:08  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/JGESZ
12:53:02  <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/JGE9Q
13:13:41  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/JGE7h
13:17:23  *** sla_ro|master has joined #openttd
13:21:14  <TrueBrain> a variable called "l" .. really glx :P
13:22:46  <TrueBrain> no clue if it is better to have ChunkHandler or ChunkHandler * in the std::vector
13:22:52  <TrueBrain> the latter is strictly seen using less memory
13:23:10  <TrueBrain> but it requires an extra dereference .. and it could potentially have nullptr .. so from a coding point of view, the first seems nicer
13:27:13  <LordAro> it's already heap allocated, so you're not gaining anything from the extra pointer apart from if it could be null or ownership/copying reasons
13:27:35  <glx> initializer_list for a constructor is the typical usage
13:29:31  <TrueBrain> glx: sorry, what do you mean?
13:31:32  <TrueBrain> LordAro: well, heap allocated or not, an extra copy takes more memory :P
13:31:39  <glx> I mean using it as an argument type is more usual than using it as a variable type
13:32:18  <TrueBrain> (to be clear, "more memory", we talk about .. 4 pointers per entry .. there are like 80 entries .. so ...
13:32:21  <TrueBrain> @calc 80 * 4 * 8
13:32:21  <DorpsGek> TrueBrain: 2560
13:32:25  <TrueBrain> 2560 bytes :P)
13:33:00  <glx> there are 33 tables, most with only one element ;)
13:33:07  <TrueBrain> total of 79 chunks
13:33:11  <TrueBrain> I know, as I have been converting them :P
13:33:21  <TrueBrain> well, no, not 79 chunks
13:33:24  <TrueBrain> 79 "objects"
13:33:27  <TrueBrain> so far less indeed
13:33:29  <TrueBrain> you are right :)
13:34:08  <TrueBrain> (most chunks has ~1 object, but some have (many) more)
13:34:20  <glx> yeah biggest is map
13:34:28  <glx> I think
13:34:29  <TrueBrain> so personally, I would do ChunkHandler, not ChunkHandler *, in the vector
13:34:47  <TrueBrain> just as it makes a lot of the other code a bit easier to read / less error prune
13:35:03  <TrueBrain> prone
13:35:05  <TrueBrain> ugh, typing hard
13:35:35  <TrueBrain> one thing I really like of C++ over C .. the ability to say: this is never a nullptr :)
13:36:12  <TrueBrain> btw, glx, the reason I think you used an initializer_list, is that you first need to store the ChunkHandlerTable[] somewhere in a variable before you can create a span<> out of it
13:36:48  <TrueBrain> so I just repeated the trick you did for the chunks itself :D
13:39:03  <TrueBrain> lol @ mail to @info .. "Hi OpenTTD (32-Bit), .."
13:39:07  <TrueBrain> didn't know that was our name now :)
13:41:50  *** nielsm has joined #openttd
13:42:33  <Rubidium> an alternative *could* be to just have one ChunkHandlerTable instead of it being spread over many _sl.cpp files. Means you need to declare the list of save/load functions, but that might be clearer in the long run?!?
13:43:51  <glx> I used initializer_list to don't need to create an array, and have a constructor usage similar to std::vector<int> v { 0, 1, 2, 3 };
13:44:38  <TrueBrain> it is creating the array anyway :)
13:45:54  <Rubidium> glx: I'm not bothered about the type of collection per se, but rather instead suggest having one collection, instead of making one collection from many collections
13:45:55  <TrueBrain> Rubidium: instead of having 1 extern per saveload file, there would be several externs for all the Save/Load procs .. not sure that is clearer in the long run, honestly
13:47:13  <Rubidium> generate them from an ini ;)
13:47:32  <TrueBrain> that is just moving the issue around
13:47:45  <glx> one of the previous version did auto register, but it also modified chunk orders
13:52:28  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/JGExp
13:54:38  <TrueBrain> sorry for all the bikeshedding glx :D
13:54:51  <glx> no it's good :)
13:55:47  <TrueBrain> trying to get the hang of these list storages myself .. not easy :P
13:56:17  <glx> yeah basically we just want an easy way to merge different lists :)
13:57:48  <TrueBrain> Range v3 ;)
14:26:11  <Rubidium> multilevel ranges!
14:26:54  <glx> I'm merging stuff inside the original branch of the PR, confilcts everywhere
14:27:15  <Rubidium> glx: I'm truely sorry
14:27:40  <glx> it's my fault, I should have removed my commit first
14:27:54  <glx> because I'm basically replacing it
14:32:45  <glx> crazy idea, use the same variable name for all static const ChunkHandler[] (I won't do that)
14:36:17  * Rubidium ups the crazy... maybe something like how currently drivers are registered. They already have a priority. After that just sort on the name of the chunk. Should be stable, although not necessarily the same as it is now
14:40:52  <TrueBrain> I haven't followed why auto-register wasn't working, but I would guess we only need 1 chunk at the start
14:40:57  <TrueBrain> and the rest is irrelevant in order
14:41:07  <TrueBrain> (gamelog we want at the start for various reasons, if I remember correctly)
14:41:27  <glx> for individual chunk handler list, the order is important
14:41:39  <glx> but each list can be anywhere
14:41:59  <TrueBrain> although, I can understand that the game is now so used to a certain order
14:42:03  <TrueBrain> it might break in unexpected ways
14:42:05  <glx> like all MAP chunks must be in a precise order
14:42:08  <TrueBrain> like, expecting settings to be loaded or not
14:42:44  <TrueBrain> but honestly, it sounds more and more we are trying to overcomplicate this by a lot :D
14:43:03  <glx> settings don't matter, I think it's handled in AfterLoad
14:43:12  <TrueBrain> I can imagine someone didn't get that memo
14:43:17  <TrueBrain> and some line in some chunk depends on it :P
14:43:28  <TrueBrain> some are really messy .. not everyone understood the saveload system that well :P
14:44:06  <glx> but really order inside chunk groups is very important
14:45:08  <Rubidium> yay... dependency tree resolution ;)
14:55:14  <glx> extracted the auto register version (and rebased to current master) https://github.com/OpenTTD/OpenTTD/compare/master...glx22:saveload_3d3f995 it works but chunk groups are in "random" order, while order inside groups is maintained
14:55:38  <DorpsGek> [OpenTTD/OpenTTD] vituscze opened pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGuLs
14:56:06  <glx> advantage of chunk based savegame is that order should not matter
15:00:31  <glx> but it still partially does
15:05:38  *** WormnestAndroid has quit IRC
15:07:04  <TrueBrain> personally, not a fan, but that has to do with: if you forget to include the file, it doesn't work, and you spend a long time figuring out why :P
15:07:09  <TrueBrain> I like the explicit dragging shit in :D
15:07:17  <TrueBrain> but that is more angst of the past :P
15:08:00  *** WormnestAndroid has joined #openttd
15:08:06  <glx> if you forget to include it doesn't compile ;)
15:10:55  <TrueBrain> and yeah, I kinda agree with JGR .. deterministic savegames sound like a nice thing to have :)
15:11:10  <glx> I fully understand this point yes
15:11:14  <TrueBrain> (sorry, I was reading up on the auto-register )
15:11:18  <TrueBrain> so I changed my mind :P
15:11:53  <DorpsGek> [OpenTTD/OpenTTD] rubidium42 approved pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGumO
15:12:09  <glx> auto register is nice, but deterministic is better
15:12:10  <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGuLs
15:12:28  <TrueBrain> well said :)
15:12:47  <TrueBrain> just reminds me of the whole Python dict shit :P
15:12:58  <glx> ideally we want both, but really not easy
15:12:58  <TrueBrain> "what did we learn boys and girls?" :D
15:13:04  <TrueBrain> sort by name
15:13:06  <TrueBrain> sort by chunk-name
15:13:10  <TrueBrain> etc etc :)
15:13:20  <Rubidium> that's why I mentioned priority and then sorting by chunk name. As for the map chunk the *only* important thing to happen first is MAPS (it sets the sizes), the rest seems not to care about the order
15:13:21  <TrueBrain> just the code gets uglier and uglier
15:13:21  <DorpsGek> [OpenTTD/OpenTTD] vituscze commented on pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGumo
15:17:54  <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGuYK
15:19:00  <DorpsGek> [OpenTTD/OpenTTD] vituscze commented on pull request #9327: Fix b791ffc6: use the correct parameter in CmdSignalTrackHelper https://git.io/JGuY7
15:30:14  <DorpsGek> [OpenTTD/OpenTTD] glx22 updated pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/J3B9h
15:31:35  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #9180: Codechange: Remove FOR_ALL_CHUNK_HANDLERS https://git.io/JGuGz
15:38:33  *** Wormnest has joined #openttd
15:40:33  *** WormnestAndroid has quit IRC
15:41:11  *** WormnestAndroid has joined #openttd
15:53:31  *** luaduck has joined #openttd
15:54:33  <DorpsGek> [OpenTTD/OpenTTD] vituscze opened pull request #9328: Fix b791ffc6: use the correct name in CmdPlantTree https://git.io/JGuWW
16:18:22  <Rubidium> glx, TrueBrain: I was thinking about something along the lines of https://github.com/OpenTTD/OpenTTD/compare/master...rubidium42:saveload-chunk-attempt It seems to work (tm) for a simple savegame with some NewGRFs. I haven't rigorously tested it though
16:19:03  <Rubidium> works as in, I can save and then reload it again
16:26:33  <DorpsGek> [OpenTTD/OpenTTD] rubidium42 approved pull request #9328: Fix b791ffc6: use the correct name in CmdPlantTree https://git.io/JGuEe
16:26:39  <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9328: Fix b791ffc6: use the correct name in CmdPlantTree https://git.io/JGuWW
16:26:55  <glx> splitting the arrays is probably not needed
16:28:38  *** Progman has joined #openttd
16:37:46  *** andythenorth has joined #openttd
16:38:13  <andythenorth> nothing wrong with python dicts TrueBrain
16:38:30  <andythenorth> it's just more FUN to use a lists of 2-tuples and unpack them
16:44:20  *** gelignite has joined #openttd
16:46:57  <Rubidium> glx: if I would've known that, it would've saved me some time ;)
17:04:30  *** Progman has quit IRC
17:07:23  *** Progman has joined #openttd
17:14:18  *** Beer has joined #openttd
17:38:03  *** jottyfan has joined #openttd
17:55:36  *** frosch123 has joined #openttd
18:00:42  <andythenorth> quak
18:06:14  <frosch123> hoi mammals
18:15:06  *** Flygon has quit IRC
18:22:21  <TrueBrain> owh boy, incoming spam
18:22:42  <DorpsGek> [OpenTTD/bananas-api] pyup-bot opened pull request #93: Scheduled monthly dependency update for June https://git.io/JGuHK
18:23:20  <DorpsGek> [OpenTTD/bananas-frontend-cli] pyup-bot opened pull request #26: Scheduled monthly dependency update for June https://git.io/JGuHx
18:24:20  <TrueBrain> https://github.blog/changelog/2021-06-01-react-to-any-github-release-to-show-your-support/ <- w00p!
18:24:27  <DorpsGek> [OpenTTD/bananas-frontend-web] pyup-bot opened pull request #58: Scheduled monthly dependency update for June https://git.io/JGuQl
18:25:56  <DorpsGek> [OpenTTD/bananas-server] pyup-bot opened pull request #50: Scheduled monthly dependency update for June https://git.io/JGuQH
18:26:50  <DorpsGek> [OpenTTD/DorpsGek] pyup-bot opened pull request #55: Scheduled monthly dependency update for June https://git.io/JGu7q
18:30:01  <TrueBrain> grr @ requests .. fucking update your deps .. it is getting annoying :D
18:30:17  <DorpsGek> [OpenTTD/eints] pyup-bot opened pull request #45: Scheduled monthly dependency update for June https://git.io/JGudU
18:30:35  <LordAro> few big releases in there too
18:30:42  <LordAro> might actually warrant checking
18:31:16  <TrueBrain> not as big as it looks
18:31:20  <TrueBrain> lot of projects dropped Python 2 and 3.4
18:31:26  <DorpsGek> [OpenTTD/master-server] pyup-bot opened pull request #31: Scheduled monthly dependency update for June https://git.io/JGudu
18:31:28  <TrueBrain> and even 3.5
18:31:40  <TrueBrain> some reason that was a thing this month
18:32:52  <DorpsGek> [OpenTTD/master-server-web] pyup-bot opened pull request #25: Scheduled monthly dependency update for June https://git.io/JGudA
18:33:12  <LordAro> fair :)
18:33:22  <TrueBrain> I switched TrueWiki to dependabot
18:33:24  <andythenorth> what a spammer
18:33:26  <TrueBrain> so I already knew what to expect :P
18:34:47  <andythenorth> what did I want to do with newgrf towns again?
18:34:50  <andythenorth> can anyone remember?
18:35:30  <LordAro> i thought it was split cargo ships?
18:36:13  <TrueBrain> okay, so idna 3 dropped Python2
18:36:17  <TrueBrain> many projects depends on idna
18:36:28  <TrueBrain> new pip punishes dependency mismatches
18:36:34  <TrueBrain> so projects are forced to upgrade to idna3
18:36:40  <TrueBrain> needing to drop Python2 for it
18:36:44  <TrueBrain> that is a nice cycle :D
18:36:47  <TrueBrain> can't say I mind
18:36:58  <TrueBrain> we have given people enough time to upgrade to Python3 :P
18:38:42  <frosch123> we have given people enough time to move away from winxp :p
18:38:59  <TrueBrain> exactly why I really wouldn't mind if we break building for WinXP :)
18:39:12  <TrueBrain> https://github.com/psf/requests/pull/5711 <- they are not quick in any sense with accepting PRs ..
18:40:53  <andythenorth> split cargo ship towns?
18:40:54  <andythenorth> hmm
18:41:12  <andythenorth> I think it was control town growth via grf
18:41:31  <frosch123> TrueBrain: the funny part is how many closed/duplicated PRs are linked from there
18:42:03  <frosch123> hmm, no, actually those are from other people
18:42:15  <frosch123> silly link spam :p
18:50:22  <TrueBrain> spam, yes :)
18:50:39  <TrueBrain> GitHub is just terrible in this interlinking
18:50:53  <TrueBrain> I wish it wouldn't do it by default, and that you have to ask it to do it
18:50:55  <TrueBrain> but okay ..
18:57:53  *** Wolf01 has joined #openttd
19:06:15  *** OsteHovel_ has joined #openttd
19:09:06  *** OsteHovel has quit IRC
19:40:34  *** OsteHovel_ is now known as OsteHovel
19:40:55  *** OsteHovel has joined #openttd
19:41:27  *** OsteHovel has joined #openttd
19:43:01  *** OsteHovel has quit IRC
19:43:13  *** OsteHovel has joined #openttd
19:44:04  *** OsteHovel has joined #openttd
19:44:31  *** OsteHovel has quit IRC
19:44:50  *** OsteHovel has joined #openttd
19:45:58  *** OsteHovel has joined #openttd
19:58:42  *** tokai has joined #openttd
19:58:42  *** ChanServ sets mode: +v tokai
20:02:36  *** jottyfan has quit IRC
20:05:34  *** tokai|noir has quit IRC
20:33:51  *** Wormnest_ has joined #openttd
20:40:23  *** Wormnest has quit IRC
20:50:24  *** nielsm has quit IRC
20:51:27  *** andythenorth has quit IRC
20:58:46  <DorpsGek> [OpenTTD/OpenTTD] DorpsGek pushed 1 commits to master https://git.io/JGzGE
20:58:47  <DorpsGek>   - Update: Translations from eints (by translators)
20:59:17  <frosch123> haha, in 21 days it will be on time
20:59:47  <frosch123> github moved to some planet with a 25h day
21:01:09  <Xaroth> Maybe it's a person doing the job starting
21:01:18  <Xaroth> it's said that the human body has a natural 25h cycle
21:07:06  *** EER has quit IRC
21:13:10  <peter1138> Hmm, never knew that Chris Sawyer (along with David Braben) was involved with Conqueror on the Archimedes.
21:14:13  *** Samu has quit IRC
21:15:07  <frosch123> he ported games to different machines before making own games
21:15:34  <peter1138> Yeah, he did the DOS port.
21:16:23  <peter1138> 16 colour VGA :D
21:17:56  *** gelignite has quit IRC
21:19:16  <frosch123> i always found old graphics cards weird, when you had to choose a subset of the available colors. cga: pick 4 of 8 colors. ega: pick 16 of 64 colors
21:27:56  <peter1138> CGA only had 2 official palettes (although low and high intensity variants too)
21:29:27  <peter1138> and of course MCGA/VGA, pick 256 out of 262144
21:30:32  <peter1138> And true VGA was planar which was just a pain in the arse.
21:44:35  *** frosch123 has quit IRC
21:46:28  *** HerzogDeXtEr has quit IRC
21:49:37  *** sla_ro|master has quit IRC
22:06:12  *** Wolf01 has quit IRC
22:49:10  *** Progman has quit IRC
23:41:54  *** Tirili has joined #openttd
23:50:08  *** tokai|noir has joined #openttd
23:50:08  *** ChanServ sets mode: +v tokai|noir
23:55:51  *** Wormnest_ has quit IRC
23:56:59  *** tokai has quit IRC

Powered by YARRSTE version: svn-trunk