IRC logs for #openttd on OFTC at 2021-03-03
⏴ go to previous day
00:09:41 *** Progman has quit IRC (Remote host closed the connection)
00:15:24 *** gelignite has quit IRC (Quit: Stay safe!)
00:20:08 <DorpsGek> [OpenTTD/OpenTTD] pelya opened issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
00:26:10 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
00:28:53 <DorpsGek> [OpenTTD/OpenTTD] pelya commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
00:43:16 <DorpsGek> [OpenTTD/OpenTTD] Rau117 commented on issue #8797: Wrong distance between signals with «Drag signal distance» https://git.io/JthDs
01:00:20 *** azubieta6 has joined #openttd
01:01:42 *** orudge` has joined #openttd
01:02:09 *** m1cr0m4n has joined #openttd
01:02:13 *** ^Spike^ has quit IRC (Ping timeout: 480 seconds)
01:02:23 *** tneo has quit IRC (Ping timeout: 480 seconds)
01:02:23 *** Yexo has quit IRC (Ping timeout: 480 seconds)
01:03:33 *** Terkhen has quit IRC (Ping timeout: 480 seconds)
01:03:38 *** orudge has quit IRC (Ping timeout: 480 seconds)
01:03:38 *** azubieta has quit IRC (Ping timeout: 480 seconds)
01:03:48 *** m1cr0man has quit IRC (Ping timeout: 480 seconds)
01:03:48 *** Mek_ has quit IRC (Ping timeout: 480 seconds)
01:06:48 *** ^Spike^ has joined #openttd
01:07:29 *** Terkhen has joined #openttd
01:07:29 *** ChanServ sets mode: +o Terkhen
01:08:53 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:06:01 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
02:08:06 *** orudge` has quit IRC (Ping timeout: 480 seconds)
02:37:23 *** TinoDidriksen has quit IRC (Ping timeout: 480 seconds)
02:44:12 *** Wuzzy has quit IRC (Remote host closed the connection)
02:44:28 *** TinoDidriksen has joined #openttd
02:45:01 *** TinoDidriksen is now known as Guest472
02:45:24 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8773: Add: Support Zstandard(zstd) savegame compression https://git.io/JqeIv
03:05:43 *** Guest472 has quit IRC (Ping timeout: 480 seconds)
03:09:38 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:18:14 *** mirrorb2rd has joined #openttd
03:20:24 *** TinoDid|znc has joined #openttd
03:23:58 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
03:32:55 *** Wormnest has quit IRC (Quit: Leaving)
03:45:55 *** TinoDid|znc is now known as TinoDidriksen
03:46:30 *** TinoDidriksen is now known as Guest479
04:00:53 *** heffer has quit IRC (Quit: heffer)
04:07:18 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
04:12:32 *** Speedy` has joined #openttd
04:18:00 *** didac has quit IRC (Remote host closed the connection)
04:25:56 *** snail_UES_ has quit IRC (Quit: snail_UES_)
05:59:51 *** mirrorb2rd has quit IRC (Quit: Leaving)
06:12:33 *** Guest479 is now known as TinoDidriksen
07:29:02 *** didac has quit IRC (Remote host closed the connection)
08:02:58 *** andythenorth has joined #openttd
08:06:11 *** sla_ro|master has joined #openttd
08:10:59 *** andythenorth has quit IRC (Quit: andythenorth)
08:17:05 *** andythenorth has joined #openttd
08:35:31 *** gelignite has joined #openttd
08:45:46 *** supermop_Home has quit IRC (Ping timeout: 480 seconds)
08:52:37 *** HerzogDeXtEr has joined #openttd
09:02:59 <TrueBrain> one area we did not touch, multiplayer GUIs :P
09:07:16 <rptr_> how do you mean multiplayer GUI?
09:08:03 <TrueBrain> the GUIs for multiplayer :)
09:08:35 <TrueBrain> you know, all those windows guiding you through setting up and joining another multiplayer game
09:11:58 <_dp_> guiding? more like discouraging :p
09:12:26 <TrueBrain> yes yes, everything is broken, OpenTTD sucks, blabla
09:13:02 <andythenorth> we can rebuild him!
09:13:07 <andythenorth> we have the technology!
09:13:57 <andythenorth> TrueBrain tell me your observations!
09:14:09 <TrueBrain> a sun on a blue sky
09:14:13 <TrueBrain> some houses in the distance
09:14:19 <TrueBrain> a car parked in front of my house
09:14:30 <TrueBrain> a garden that hasn't been taken care for in a while
09:14:30 <andythenorth> there is an egg here
09:14:47 <TrueBrain> that is not how that game works andythenorth :P You cannot, as controller, say what is where :P
09:15:04 <andythenorth> IMAGINE IF YOU COULD
09:15:09 <andythenorth> think we just invented a new genre
09:15:10 <TrueBrain> would make a boring game :)
09:15:49 <andythenorth> and now we arrive immediately at why god invented people
09:16:00 <andythenorth> * according to some
09:16:21 <Wolf01> Exactly, people with free will are making my life a hell
09:16:44 <andythenorth> ludic theory of omnipotent deity
09:16:50 <andythenorth> omnipotence gets boring
09:17:15 <andythenorth> anyway TrueBrain I think the button should be pink
09:17:19 <andythenorth> that's my UI contribution
09:17:41 <TrueBrain> ty, your opinion has been filed with the classification /dev/null
09:18:31 <_dp_> TrueBrain, OpenTTD is old :p
09:19:15 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
09:20:38 * andythenorth might have a shave
09:28:16 <Wolf01> 70% probability... sure he tested at least 10 times
09:30:07 <Wolf01> Updated mirc, nice, 64 colours now!
09:41:05 <Timberwolf> Blimey, that worked on irssi.
09:41:22 <Timberwolf> Well. The 'o', 'a' and 'y' are near-invisible because black terminal background, but still.
09:41:33 <peter1138> There are only 7 colours. It is the law.
09:41:34 <LordAro> it worked writing it on irssi too :p
09:41:42 <rptr_> how do i make all my text light grey
09:41:54 <Tulitomaatti> unexpected irssi colour.
09:44:35 <Timberwolf> Sudden ZZT flashbacks...
10:04:27 <glx> about #8800, maybe we should add this check to one of the linux CI
10:12:05 <DorpsGek> [OpenTTD/OpenTTD] orudge commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
10:17:25 <glx> _dp_: yeah forcing minimum version for the lib seems ok
10:18:09 <glx> knowing theres API changes between each version IIRC
10:18:20 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
10:21:52 <DorpsGek> [OpenTTD/OpenTTD] ldpl updated pull request #8773: Add: Support Zstandard(zstd) savegame compression https://git.io/Jtx6j
10:25:59 <peter1138> My manager has a severe case of hanging-on-itis when it comes to ending a Teams call. Speak you later, see you, bye, bye. Always feels like I've hung up on him.
10:26:29 <TrueBrain> I know that type of manager
10:26:40 <TrueBrain> I like to toy with those manager a bit by not hanging up, and keep the bye going
10:26:50 <TrueBrain> it becomes so awkward at a certain point
10:26:53 <TrueBrain> fucking hilarious :D
10:32:47 <peter1138> Hmm, when I pause a game the Simulation rate: frames/s shows some seemingly random value.
10:32:57 <TrueBrain> the last value; fixed in master
10:34:03 <TrueBrain> LordAro: I am trying to get clang to compile against libc++ here, but .. CMake is refusing in so many ways :P
10:34:04 <peter1138> Not, I'm on g95462493ef which is after that change.
10:34:21 <peter1138> Okay, non-FFWD it shows 0.01 frames/s
10:34:33 <peter1138> Or 0.00. Logical, it's paused.
10:35:13 <TrueBrain> and now it refuses to find algorithm, lol
10:35:16 <peter1138> FFWD I see 5787.61 frames/s one time, 12054.44 another time.
10:35:19 <LordAro> in theory, CXXFLAGS=-stdlib=libc++ should be enough
10:35:33 <peter1138> Unpaused in FFWD it's a fairly steady 830 frames/s
10:35:33 <TrueBrain> yeah, CFLAGS didn't work :P
10:35:40 <TrueBrain> with CXXFLAGS I get include errors .. hmm
10:35:49 <LordAro> well it's definitely doing something then!
10:35:58 <peter1138> 33921.93, that'd be nice.
10:36:00 <LordAro> do you have libc++ installed? :p
10:36:05 <TrueBrain> made sure of that :D
10:36:12 <TrueBrain> took me a while to figure out how :P
10:37:10 <peter1138> Also weird: when paused, the window is 1 pixel wider. Unpausing sets it back -1 pixel again.
10:37:35 <TrueBrain> fatal error: 'algorithm' file not found .. hmm ...
10:37:35 <peter1138> I guess that's my font size.
10:38:04 <peter1138> " Road vehicle ticks:" is removed, I guess "Graphics rednering:" is 1 pixels shorter
10:38:30 <peter1138> Wrong way around above, paused is 1 pixel narrower, because the entries in the list change.
10:39:16 <TrueBrain> anyway, LordAro , maybe edit your reply to read `CXXFLAGS="-stdlib=libc++"` .. that at least picks the right stdlib :) Now to figure out how to make it find includes ..
10:44:32 <TrueBrain> lol, had to add the include folder to /usr/lib/llvm-11/include/c++/v1/ myself ... what-ever clang :P
10:47:22 <TrueBrain> at least that crashes too now :)
10:48:55 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
10:51:08 <TrueBrain> the pointer given by algorithm is wrong, funny
10:52:12 <TrueBrain> there is only 1 thread active, so that is not an issue either
10:52:29 <LordAro> are the iterators valid?
10:53:25 <TrueBrain> it uses ::begin() and ::end()
10:53:31 <TrueBrain> so in that sense, I would surely hope so :D
10:53:39 <TrueBrain> std::sort(std::vector<T>::begin(), std::vector<T>::end(), ...)
10:54:26 <TrueBrain> but one of the calls to the lambda seems to contain an invalid pointer nevertheless
10:54:43 <TrueBrain> I guess the change we did in beta2, was to fix the sort-order of servers again
10:55:16 <TrueBrain> but even there, I do not see why it crashes now and not before
10:56:44 <TrueBrain> ::end() is the first invalid entry, right?
10:57:08 <TrueBrain> if my C++ knowledge remembers correctly, that was the odd thing about these things, not? :D
10:57:29 <_dp_> can the vector change while being sorted?
10:57:38 <TrueBrain> as the pointer to ::end() is given to the sorter lambda :P
10:57:52 <LordAro> [begin, end) is common
10:58:30 <LordAro> MSVC errors if the iterators are invalid, so that would've caught it normally
10:59:11 <TrueBrain> "first, last - the range of elements to sort "
10:59:19 <TrueBrain> that gives me no indication if they are inclusive or exclusive
10:59:29 <TrueBrain> Sorts the elements in the range [first, last) in non-descending order.
10:59:32 <TrueBrain> at least that is clear
10:59:51 <TrueBrain> so our call is okay, just libc++ seems to get confused
11:00:21 <TrueBrain> they correct their "last" pretty quick to be [begin, end]
11:00:36 <TrueBrain> but for some reason, it still calls the lambda with the end() value :P
11:03:42 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
11:06:51 <TrueBrain> and solves the issue
11:08:16 <LordAro> horrendous crash behaviour though
11:11:25 *** gelignite has quit IRC (Quit: Stay safe!)
11:13:32 <TrueBrain> LordAro: and you even fixed this issue before! How could you not spot this before JGR did :P :P :P
11:14:18 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8801: Fix: NGameAllowedSorter() is not imposing strict weak ordering relation https://git.io/Jqe9S
11:15:28 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
11:16:04 <LordAro> oh, well, i copied JGR's solution :p
11:16:19 <TrueBrain> it was also a lifetime ago, so it is fine :)
11:17:47 <TrueBrain> so Discord denied the Verification of the server, with no reason given, except that we do not meet the requirements. They list them again, and we meet them one by one
11:17:57 <TrueBrain> so .. yeah ... I emailed them, like: what does one have to do more to meet them? :D
11:18:29 <TrueBrain> well, I requested it via DorpsGek .. that might be a bit of the problem, I guess
11:19:30 <TrueBrain> the process is also a bit weird ... you fill in a very simple form with: state your name, state the server, and you have to link (via authenticatin) to twitter / GitHub / ..
11:19:40 <TrueBrain> no field to explain or point to other places
11:24:28 <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #8801: Fix: NGameAllowedSorter() is not imposing strict weak ordering relation https://git.io/JqeHc
11:26:14 <_dp_> that's not a very good translation...
11:27:16 <glx> we don't have eints for basesets, so we have to trust PRs
11:32:49 <LordAro> not like we check the quality of eints translations either :p
11:38:28 *** WormnestAndroid has quit IRC (Remote host closed the connection)
11:38:41 *** WormnestAndroid has joined #openttd
11:41:16 *** roadt__ has joined #openttd
11:47:53 *** roadt_ has quit IRC (Ping timeout: 480 seconds)
11:50:43 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
11:59:29 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
11:59:41 *** WormnestAndroid has joined #openttd
11:59:47 *** gelignite has joined #openttd
12:12:33 <DorpsGek> [OpenTTD/OpenTTD] LordAro closed issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
12:12:36 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #8801: Fix: NGameAllowedSorter() is not imposing strict weak ordering relation https://git.io/Jqe9S
12:15:08 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/Jqe5l
12:23:29 <TrueBrain> right, time for a monthly round of: have you updated this?
12:27:47 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8706: Feature: rail station class name filtering https://git.io/Jqedq
12:29:35 <TrueBrain> owh, bah, aiohttp backported the fix for chardet wrong .. so I still have to manually fix every repo .. bah :P
12:33:38 <TrueBrain> I couldn't resist trying :D
12:41:29 <DorpsGek> [OpenTTD/bananas-api] TrueBrain updated pull request #82: Scheduled monthly dependency update for March https://git.io/Jtp5Y
12:42:05 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #18: Scheduled monthly dependency update for March https://git.io/Jtp5u
12:43:17 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain updated pull request #54: Scheduled monthly dependency update for March https://git.io/Jtp51
12:45:07 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #18: Scheduled monthly dependency update for March https://git.io/Jtp5u
12:47:13 <DorpsGek> [OpenTTD/bananas-server] TrueBrain updated pull request #42: Scheduled monthly dependency update for March https://git.io/Jtp5h
12:48:33 <DorpsGek> [OpenTTD/master-server] TrueBrain updated pull request #26: Scheduled monthly dependency update for March https://git.io/Jtpdx
12:49:10 <TrueBrain> we need something better than pyup really ..
12:50:23 <DorpsGek> [OpenTTD/master-server-web] TrueBrain updated pull request #20: Scheduled monthly dependency update for March https://git.io/JtpFU
12:51:31 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain updated pull request #45: Scheduled monthly dependency update for March https://git.io/JtpdO
12:51:55 <TrueBrain> especially as it is refused to make a PR for eints
12:51:57 <TrueBrain> which is really annoying
12:53:26 <TrueBrain> owh, update is set to false
12:54:34 <DorpsGek> [OpenTTD/bananas-api] TrueBrain updated pull request #82: Scheduled monthly dependency update for March https://git.io/Jtp5Y
12:55:03 <DorpsGek> [OpenTTD/bananas-api] TrueBrain updated pull request #82: Scheduled monthly dependency update for March https://git.io/Jtp5Y
12:56:30 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #8700: Change: Recolour graph windows to brown https://git.io/JqebY
12:56:49 <DorpsGek> [OpenTTD/eints] TrueBrain opened pull request #34: Fix: [pyup] automatic creating of PRs was disabled https://git.io/JqebG
12:57:50 <TrueBrain> owh, eints is just broken
12:57:53 <TrueBrain> well, of course it is :P
12:58:06 <LordAro> the branches aren't going to help there
12:58:19 <TrueBrain> yeah .. we should make 2 projects out of it
12:58:23 <TrueBrain> and push eints to pyup, honestly
12:58:41 <TrueBrain> or just say: fuck everyone else, and merge openttd-github in master
13:00:45 <DorpsGek> [OpenTTD/eints] TrueBrain updated pull request #34: Fix: [pyup] automatic creating of PRs was disabled https://git.io/JqebG
13:01:46 <TrueBrain> okay .. at least that works now :)
13:02:00 <TrueBrain> after merging that, pyup can finally make a PR
13:02:10 <TrueBrain> guess I might as well manually bump the other versions well there
13:04:05 <_dp_> or revert window resizing
13:04:19 <_dp_> because right now 1.11 puts the most important info in the worst possible place
13:04:53 <DorpsGek> [OpenTTD/eints] TrueBrain updated pull request #34: Fix: [pyup] automatic creating of PRs was disabled https://git.io/JqebG
13:07:13 <TrueBrain> okay, all these update PRs are now ready for review
13:07:30 <TrueBrain> and next, upgrading of AWS stuff \o/
13:08:48 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8709: Change: Don't show global goals in company goal windows https://git.io/JqeNt
13:09:49 <DorpsGek> [OpenTTD/eints] LordAro approved pull request #34: Fix: [pyup] automatic creating of PRs was disabled https://git.io/JqeNG
13:10:33 <LordAro> TrueBrain: why is idna pinned?
13:11:10 <TrueBrain> LordAro: everything I pinned is because of dependency issues .. chardet for aiohttp, idna for requests, and ssmap (or what was it called) for ... something else
13:11:18 <TrueBrain> I forgot to write it in the PRs this time, didn't I? :D
13:12:26 <TrueBrain> I wish I had a cleaner way to make it clear why we are holding back some packages ..
13:14:04 *** supermop_Home has joined #openttd
13:15:46 <LordAro> TrueBrain: section at the bottom of requirements.base, perhaps?
13:16:11 <TrueBrain> LordAro: by pyup doesn't use that :(
13:16:36 <LordAro> does it not respect it at all?
13:17:07 <TrueBrain> it only looks at requirements.txt and bumps it
13:17:12 <TrueBrain> maybe it uses setup.py too
13:17:20 <TrueBrain> but requirements.base is something I "invented"
13:17:25 <TrueBrain> because we lack setup.py in all those repos
13:18:23 <TrueBrain> right, time to scale up our AWS instances to break down the old ones after :)
13:21:31 <DorpsGek> [OpenTTD/eints] TrueBrain merged pull request #34: Fix: [pyup] automatic creating of PRs was disabled https://git.io/JqebG
13:22:25 *** andythenorth has quit IRC (Quit: andythenorth)
13:35:56 *** DorpsGek` has joined #openttd
13:35:56 *** ChanServ sets mode: +o DorpsGek`
13:35:56 *** DorpsGek is now known as Guest517
13:35:56 *** DorpsGek` is now known as DorpsGek
13:35:56 *** DorpsGek` has quit IRC (Killed (synthon.oftc.net (Nick change collision)))
13:35:58 *** DorpsGek` has joined #openttd
13:35:58 *** ChanServ sets mode: +o DorpsGek`
13:35:58 *** DorpsGek is now known as Guest518
13:35:58 *** DorpsGek` is now known as DorpsGek
13:36:06 <TrueBrain> I forgot something didn't I? :D
13:36:11 *** DorpsGek is now known as Guest519
13:36:11 *** Guest519 is now known as DorpsGek
13:36:06 <TrueBrain> I forgot something didn't I? :D
13:36:11 *** DorpsGek is now known as Guest519
13:36:11 *** Guest518 is now known as DorpsGek
13:36:11 *** Guest519 has quit IRC (Killed (synthon.oftc.net (Nick change collision)))
13:36:13 *** DorpsGek has joined #openttd
13:36:13 *** ChanServ sets mode: +o DorpsGek
13:36:42 *** DorpsGek` has joined #openttd
13:36:42 *** DorpsGek is now known as Guest520
13:36:42 *** ChanServ sets mode: +o DorpsGek`
13:36:42 *** Guest520 is now known as DorpsGek
13:36:42 *** Guest520 has quit IRC (Killed (synthon.oftc.net (Nick change collision)))
13:36:52 *** DorpsGek has joined #openttd
13:36:53 *** ChanServ sets mode: +o DorpsGek
13:37:14 *** DorpsGek` has joined #openttd
13:37:14 *** DorpsGek is now known as Guest521
13:37:14 *** ChanServ sets mode: +o DorpsGek`
13:37:14 *** Guest521 is now known as DorpsGek
13:37:14 *** DorpsGek` has quit IRC (Killed (charon.oftc.net (Nick change collision)))
13:37:24 *** DorpsGek has joined #openttd
13:37:24 *** ChanServ sets mode: +o DorpsGek
13:37:26 <TrueBrain> I keep forgetting :(
13:37:54 *** DorpsGek has joined #openttd
13:37:54 *** ChanServ sets mode: +o DorpsGek
13:38:25 <TrueBrain> ready for review ^^
13:41:56 <TrueBrain> okay, AWS upgraded .. was ... without incident, it seems
13:47:28 *** erle- has quit IRC (Quit: Leaving)
13:58:19 <TrueBrain> okay, I should have solved the DorpsGek issue for real now :P
14:18:52 <DorpsGek> [OpenTTD/website] TrueBrain opened pull request #196: Update both Python and Gem dependencies to latest https://git.io/JqehE
14:23:35 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
14:23:42 *** snail_UES_ has joined #openttd
14:23:50 *** WormnestAndroid has joined #openttd
15:00:12 <DorpsGek> [OpenTTD/website] LordAro approved pull request #196: Update both Python and Gem dependencies to latest https://git.io/JqvfY
15:01:28 <TrueBrain> Why that smiley, LordAro? :)
15:02:08 <LordAro> testing is what we have staging for, right?
15:02:46 *** supermop_Home has quit IRC (Ping timeout: 480 seconds)
15:20:24 *** sla_ro|master has quit IRC ()
15:27:47 <DorpsGek> [OpenTTD/website] TrueBrain merged pull request #196: Update both Python and Gem dependencies to latest https://git.io/JqehE
15:27:56 <TrueBrain> 2 down, 7 more to go :P We need to find a better way to do this tbh
15:29:09 <TrueBrain> owh, I now see where your idna question came from .. had to do when both the app and regression_runner were installed
15:29:14 <TrueBrain> one downgraded it, before testing
15:29:18 <TrueBrain> so that was .. not ideal
15:38:22 <DorpsGek> [OpenTTD/bananas-api] TrueBrain approved pull request #82: Scheduled monthly dependency update for March https://git.io/JqvU6
15:38:45 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain approved pull request #54: Scheduled monthly dependency update for March https://git.io/JqvUi
15:38:51 <DorpsGek> [OpenTTD/bananas-server] TrueBrain approved pull request #42: Scheduled monthly dependency update for March https://git.io/JqvUP
15:38:55 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain approved pull request #45: Scheduled monthly dependency update for March https://git.io/JqvU1
15:39:02 <DorpsGek> [OpenTTD/master-server] TrueBrain approved pull request #26: Scheduled monthly dependency update for March https://git.io/JqvUD
15:39:05 <DorpsGek> [OpenTTD/master-server-web] TrueBrain approved pull request #20: Scheduled monthly dependency update for March https://git.io/JqvUy
15:39:18 <DorpsGek> [OpenTTD/bananas-api] TrueBrain merged pull request #82: Scheduled monthly dependency update for March https://git.io/Jtp5Y
15:39:28 <DorpsGek> [OpenTTD/master-server-web] TrueBrain merged pull request #20: Scheduled monthly dependency update for March https://git.io/JtpFU
15:39:45 <DorpsGek> [OpenTTD/master-server] TrueBrain merged pull request #26: Scheduled monthly dependency update for March https://git.io/Jtpdx
15:40:00 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain merged pull request #45: Scheduled monthly dependency update for March https://git.io/JtpdO
15:40:13 <DorpsGek> [OpenTTD/bananas-server] TrueBrain merged pull request #42: Scheduled monthly dependency update for March https://git.io/Jtp5h
15:40:37 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain merged pull request #54: Scheduled monthly dependency update for March https://git.io/Jtp51
15:40:51 *** supermop_Home has joined #openttd
15:46:35 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
15:49:19 <TrueBrain> lets see if I fixed DorpsGek ...
15:52:06 *** DorpsGek` has joined #openttd
15:52:07 *** ChanServ sets mode: +o DorpsGek`
15:52:07 *** DorpsGek is now known as Guest526
15:52:07 *** DorpsGek` is now known as DorpsGek
15:52:10 *** Guest526 has quit IRC (Remote host closed the connection)
15:52:17 <TrueBrain> much better, not? :D
15:52:51 <DorpsGek> LordAro: I have not seen Yexo.
15:55:17 <TrueBrain> more than 5000 lines are lost in the Seen database
15:55:25 <TrueBrain> so yeah ... that file got heavily corrupted at some point
15:55:34 <TrueBrain> if I am really bored one day, I will correct the issue :P
15:55:45 <TrueBrain> requires replaying logs .. so it is a bit of a work :D
15:56:37 <TrueBrain> okay, that should be all
15:56:57 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain approved pull request #18: Scheduled monthly dependency update for March https://git.io/JqvkG
15:57:04 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain merged pull request #18: Scheduled monthly dependency update for March https://git.io/Jtp5u
16:00:01 <TrueBrain> all names after H are gone, basically
16:01:53 <TrueBrain> means I just have to download the logs since my last backup, and replay it to recover it
16:03:07 <TrueBrain> might do that soon, just ... because I can
16:03:41 <TrueBrain> the hickups you are seeing are now also reported on "sdl" (the non-OpenGL)
16:03:47 <TrueBrain> someone really should debug that :D
16:03:54 <TrueBrain> I can help, but I cannot reproduce it, so ... meh
16:08:10 <peter1138> Does anyone still care about 80 character columns?
16:11:38 <TrueBrain> the Python projects OpenTTD has put it on 120
16:29:50 <peter1138> 80 is about 1/3rd of my developing window :-)
16:36:20 <_dp_> perfect for split screen ;)
16:37:00 *** Wormnest has joined #openttd
16:38:29 <peter1138> I somehow miss text-only computing.
16:53:47 *** _dp_ has quit IRC (Ping timeout: 480 seconds)
16:55:00 <Eddi|zuHause> you've been eaten by a grue
17:11:31 <_dp_> master server doesn't seem to be feeling well
17:26:04 <TrueBrain> pproxy seems to fail .. product has no changelogs ..
17:26:14 <TrueBrain> best guess, 2.3 server and 2.7 client cannot talk to each other
17:26:22 <TrueBrain> which would be really annoying if true ... but lets a 2.7 server
17:27:20 *** Progman has joined #openttd
17:37:24 <TrueBrain> its funny, it is fully operational .. just traffic is not flowing back out of the network :D
17:41:48 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
17:47:16 *** andythenorth has joined #openttd
17:50:57 <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8706: Feature: rail station class name filtering https://git.io/Jt9ev
17:51:02 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/Jqvs4
17:51:34 *** sla_ro|master has joined #openttd
17:54:19 <TrueBrain> ugh, it really is the new pproxy
17:56:45 <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/JthCX
17:59:41 <TrueBrain> we can bitch about our commit messages, and how difficult it can be
17:59:45 <TrueBrain> but the other side of the spectrum is this:
17:59:49 <TrueBrain> check the commit message
17:59:52 <TrueBrain> check the change ..
17:59:57 <TrueBrain> breaking APIs like there is no tomorrow
18:01:08 <DorpsGek> [OpenTTD/master-server] TrueBrain opened pull request #27: Fix: pproxy changed their API without telling anyone https://git.io/JqvG8
18:01:30 <TrueBrain> I am not annoyed at all or anything
18:02:01 <LordAro> i suppose it is an internal function...
18:02:10 <LordAro> it's internal to us :)
18:02:19 <LordAro> is that the only one?
18:02:34 <TrueBrain> we really only use it to set the destination address and to get the SOCKS header
18:02:45 <LordAro> worth adding "try this function or this other function" block, in case they change it back again? :p
18:02:57 <TrueBrain> that is rather unlikely
18:02:59 <DorpsGek> [OpenTTD/master-server] LordAro approved pull request #27: Fix: pproxy changed their API without telling anyone https://git.io/JqvGV
18:03:07 <TrueBrain> looking for another library, that is better advise :P
18:03:22 <DorpsGek> [OpenTTD/master-server] TrueBrain merged pull request #27: Fix: pproxy changed their API without telling anyone https://git.io/JqvG8
18:03:52 <TrueBrain> well, at least the rest of AWS did get a nice health-check
18:04:15 *** sugarman has joined #openttd
18:04:47 <TrueBrain> there is this "edit" button on the wiki
18:04:53 <TrueBrain> really odd, but our wiki allows editing :D
18:04:54 <sugarman> it not work in new version when put "="
18:05:18 <sugarman> TrueBrain: me no want make account for 1 ez thing
18:05:28 <TrueBrain> and so ..... you want us to do the work instead?
18:05:37 <LordAro> you're probably gonna have a bad time if you spend time pointing out every tiny incorrect detail with the wiki
18:05:48 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
18:06:15 <sugarman> that is technical problem LordAro
18:06:25 <TrueBrain> okay, curious if it will now hit the limit of DynamoDB .. pam pam pammmm
18:06:39 <sugarman> will hurt player trying play older scenario :(
18:06:46 <TrueBrain> so please fix it ! :)
18:06:48 <LordAro> also, not sure you're supposed to edit "Archive" pages
18:06:52 <FLHerne> sugarman: Anyway, that setting has been `on` by default for 15 years or so
18:06:58 <TrueBrain> LordAro: I do not see why not :)
18:07:02 <glx> IIRC = works, but without spaces
18:07:07 <sugarman> FLHerne: it is, but not on old scenarios 4some reason
18:07:33 <sugarman> config have dynamic_engines = true but still do this
18:07:51 <sugarman> glx: either way, is wrong on wiki if u try to paste
18:07:58 <LordAro> glx: sounds trivial to fix
18:08:25 <sugarman> do scenarios have own config settings??
18:08:26 <TrueBrain> so by now you spend more time talking about it sugarman , than it would have taken you to fix it yourself ;)
18:08:56 <glx> ah no, I checked the source, no "="
18:09:09 <LordAro> sugarman: yes, there are many settings that are stored in the savegame/scenario itself
18:09:30 <sugarman> even ones also in openttd.config LordAro ?
18:09:55 <TrueBrain> owh, funny, servers are signing in one by one of course, so the load is spread out .. sweet :D By accident, that is implemented in a good way :D
18:10:36 <LordAro> sugarman: yes, the ones in openttd.cfg (that are also stored in the savegame) are the defaults for new savegames
18:10:55 <sugarman> if tokyo metro give fatal newgrf issue becuz dynamic engine is off, if i turn on by console will it just work or do i have to relaunch scenario? LordAro
18:11:35 <sugarman> becuz if i have to relaunch scenario, then is impossible 2 get it to work cuz the scenario would override it anyway
18:12:21 <LordAro> (i have no idea, try it and find out!)
18:12:57 <sugarman> i cant tell if its working or broke bcuz i cant see what it looks like when it works f
18:14:58 <TrueBrain> we should rename this channel honestly ... #everything-sucks would be a better name ;)
18:16:55 <TrueBrain> okay, master-server bounced back nicely now
18:17:02 <TrueBrain> weird that sentry was not picking up on the exception
18:17:09 <TrueBrain> happened in a weird thread in Python
18:17:14 <TrueBrain> and it got eaten somewhere
18:17:33 <TrueBrain> glx: only after you fix the problem, but yes :P
18:17:56 <TrueBrain> well, also please make it an extra entry in the matrix
18:18:05 <TrueBrain> instead of using "libsdl2", as now libsdl2 won't be tested I think :P
18:18:21 <TrueBrain> owh, it is in addition to
18:18:37 <sugarman> this is the best scenario in openttd
18:18:40 <TrueBrain> that might be a bit weird? Dunno .. I would add to the matrix an entry like: extra-cmake-parameters glx :)
18:18:43 <glx> it's an extra step to one of the matrix builds yes
18:18:52 <TrueBrain> as now you "cd" into the wrong build folder :P
18:19:10 <TrueBrain> deduplication there might be nice
18:19:13 <TrueBrain> simpler == better :)
18:19:46 <glx> so extra-cmake-parameters and another addition to the matrix
18:19:53 <TrueBrain> I think that is better
18:19:56 <TrueBrain> easier to see what breaks
18:20:23 <TrueBrain> and we might want to test other parameters too :D
18:21:03 *** frosch123 has joined #openttd
18:23:40 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8784: Alt-Enter to make OpenTTD fullscreen desyncs actual fullscreen state from toggle in Game Options https://git.io/JtpHd
18:23:41 <frosch123> hoi, alt+tab issue was fixed after beta2?
18:24:05 <LordAro> alt+tab has been fixed
18:24:15 *** sugarman has quit IRC (Remote host closed the connection)
18:24:56 <frosch123> ok, that was after beta2 :)
18:25:20 <frosch123> i can't await your website PR :)
18:26:11 <_dp_> I know there are some fans of breakdowns here...
18:26:13 <frosch123> well, this mail was before the merge :)
18:26:29 <TrueBrain> owh, you are talking about an email
18:26:32 <_dp_> How about you implement some way to know where the damn train actually got lost? ;)
18:26:34 <TrueBrain> sorry, context is everything frosch123 :)
18:26:45 <_dp_> should make hunting for wrong depots much easier
18:26:49 <frosch123> the timeline is important here :) beta2 < alt-tab fix < bug report via mail < website contact update :)
18:26:56 <TrueBrain> frosch123: and the new contact page is live btw :)
18:27:20 <frosch123> yes, i just learnt that :)
18:27:53 <TrueBrain> hopefully that helps
18:28:03 <TrueBrain> I have also been pressing Unsubscribe lately on stupid emails from things like Docker
18:29:04 <frosch123> do we still use dockerhub? or was it completely replaces by gh actions?
18:29:15 <TrueBrain> as far as I know, we do not make use of it anymore, no
18:29:49 <TrueBrain> well, replaced by GHCR (GitHub Container Repository)
18:35:20 <glx> oh I see the issue, we detect everything independant of OPTION_DEDICATED, then we link some depending on OPTION_DEDICATED
18:36:00 <glx> and of course some stuff is found (like Fontconfig) then ignored, but still known to CMake
18:40:27 <TrueBrain> WITH_ variables exist for that reason :)
18:41:25 <glx> but add_files happen before link_package
18:43:22 <TrueBrain> ah, for adding files, yes :D
18:44:02 <glx> I'll just add extra dedicated checks around lib detection I think
18:48:57 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8773: Add: Support Zstandard(zstd) savegame compression https://git.io/JqvCr
18:52:56 <TrueBrain> LordAro: just as a FYI, ^^ would effectively remove Ubuntu 18.04 and Debian Buster from our release targets. I don't really have feeling about that one way or the other, but you are our biggest advocate for these targets, so just wanted to keep you in the loop there :)
18:53:35 <LordAro> rules it out of 1.11, at the very least
18:53:56 <TrueBrain> I have no clue if it is difficult to make it zstd1.3 compatible too, or if they are even compatible
18:54:28 <LordAro> that would be the ideal
18:55:28 <glx> it builds, but fails testing
18:55:46 <TrueBrain> something is not skipping timestamps, lol
18:56:23 <glx> I guess I need to add a regex
19:00:08 <peter1138> Uhoh, my Windows tablet is updating... takes a while with 1GB RAM on an Atom...
19:01:08 <DorpsGek> [OpenTTD/OpenTTD] ldpl commented on pull request #8773: Add: Support Zstandard(zstd) savegame compression https://git.io/JqvW0
19:02:47 *** qwebirc75610 has joined #openttd
19:03:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8773: Add: Support Zstandard(zstd) savegame compression https://git.io/JqvWw
19:06:31 <_dp_> I guess it's not that big of a deal to add it just for network
19:06:51 <_dp_> worst case server owners will have to learn how to change compression :p
19:07:06 <TrueBrain> well, basically that, yes :)
19:07:27 <TrueBrain> and if it breaks terribly .. we just do a patch-release, I mean ..
19:07:43 <TrueBrain> they are non-persisting saves, so boo-fucking-hoo :P
19:07:56 <_dp_> it's just that openttd does not differentiate between saving reasons
19:08:08 <TrueBrain> but it feels like that is an easy thing to add :)
19:08:09 <_dp_> so now have to think how to do it properly :/
19:08:21 <TrueBrain> but in my book, it would be for the better :)
19:08:33 <_dp_> well, jgr added it quite simply
19:09:02 <_dp_> but if you think about it server and client can both have different set of libraries compiled
19:09:20 <LordAro> yeah, there's no reason why it couldn't just be an optional dependency like all the others, no need to change any defaults
19:09:21 <_dp_> so ideally client should somehow let server know what it has and server pick what it agrees on
19:09:25 <TrueBrain> we do not handshake compression algorithms :D
19:09:31 <LordAro> no need to drop platform support because of it
19:09:56 <TrueBrain> not sure I follow you here LordAro
19:10:09 <LordAro> server supports zstd and uses it for its save
19:10:13 <TrueBrain> _dp_: doable for sure, just .. more work :D
19:10:17 <LordAro> client does not, and just gets an error on connection
19:10:18 <_dp_> TrueBrain, jgr just made all network saves and autosaves use zstd
19:10:41 <TrueBrain> LordAro: you consider that correct behaviour?
19:10:56 <LordAro> no different from someone compiling OTTD without lzma support today
19:10:58 <_dp_> well, that's the behaviour we currently have, zstd or not :p
19:11:07 <TrueBrain> LordAro: self-compiled, yes .. but binaries we publish?
19:11:17 <TrueBrain> it is the reason our binaries have LZO :P
19:11:17 <_dp_> you can compile without lzma and get screwed xD
19:11:28 <TrueBrain> we really cannot publish binaries where you cannot join servers with :D
19:11:32 <TrueBrain> like .. you get "just" an error
19:11:38 <TrueBrain> huh? I downloaded it from the official website?
19:11:41 <TrueBrain> that .. is not good :D
19:11:49 <_dp_> but yeah, I agree, official ones should have it
19:11:51 <LordAro> like i said, the better solution would be being compatible with zstd 1.3
19:12:02 <TrueBrain> or, what _dp_ suggests, handshake compression algorithms available
19:12:11 <TrueBrain> and make the server fallback to a matching one
19:12:28 <LordAro> can the server change compression algorithms as easy as that?
19:12:35 <TrueBrain> it really is that easy
19:12:45 <peter1138> With something better than just a bool like in JGR's patch :)
19:12:58 <_dp_> handshake seems much easier to implement
19:13:13 <LordAro> depends on the incompatibilities in zstd1.3 vs 1.4
19:13:14 <_dp_> than trying to understand anything in that crap of documentation zstd has :/
19:14:14 <LordAro> TrueBrain: i was specifically referring to network save compression :p
19:14:23 <TrueBrain> all saves go through there :)
19:14:33 <TrueBrain> so there is need for some flag to indicate origin, for sure
19:14:49 <TrueBrain> SAVEMODE_AUTOSAVE, SAVEMODE_NETWORK, SAVEMODE_LOCAL, I dunno :)
19:15:15 <peter1138> Ooh, it's up to 27%
19:15:23 <TrueBrain> there is already a threaded flag
19:15:28 <TrueBrain> possibly that can be collapsed into one
19:15:34 <TrueBrain> SAVEMODE_THREADED | SAVEMODE_AUTOSAVE :P
19:15:43 <LordAro> sounds overly generic to do something like that
19:15:48 <TrueBrain> someone already poked a hole in everything for "threaded" :)
19:16:04 <TrueBrain> what part sounds overly generic?
19:16:14 <LordAro> should be 3 functions that all call this one with a specific compression format
19:16:35 <TrueBrain> if you want to do it proper, you maintain the layered approach we have no tbh
19:16:54 <TrueBrain> but again, holes are already poked for this with "threaded"
19:17:29 <LordAro> why is threaded saves even a thing that can be disabled?
19:17:33 <_dp_> not only it's 27% but also a very fast compression which is the most important thing for server itself
19:17:40 <TrueBrain> LordAro: emergency saves, I guess
19:18:11 <TrueBrain> LordAro: yup, emergency saves
19:19:09 <TrueBrain> owh, and autosave vs normal load
19:19:20 <TrueBrain> uses the same code :P
19:19:34 <TrueBrain> never look too close into these things, in general you regret the choice :D
19:19:41 *** Smedles has joined #openttd
19:20:17 <TrueBrain> funny, threaded is true by default, but more places set it to false :D
19:20:58 <frosch123> there is already a dependency in the saveload code for saving for joining
19:21:19 <frosch123> it's probably ugly, but depot-order-backups involve that magic
19:22:17 <frosch123> if (!_networking || !_network_server) return; <- Save_BKOR(): i expected something more fancy :p
19:22:50 <TrueBrain> that is the nice alternative way of doing that, yes :p
19:23:07 <TrueBrain> I am a bit surprised "threaded" is added as parameter, and not as global
19:23:14 <TrueBrain> not sure it is bad/good, just surprised
19:23:33 <TrueBrain> we also have this of course:
19:23:34 <TrueBrain> if (_network_server || !_settings_client.gui.threaded_saves) threaded = false;
19:24:22 <TrueBrain> LordAro: now if you ask me why it is (still) a setting to disable threading for saves, that I really do not know :P
19:24:40 <TrueBrain> back when 2 cores were the exception, I am sure there was a reason for that :D
19:27:52 <TrueBrain> okay, I should stop reading this code .. just funny to see years of evolution :)
19:28:06 <_dp_> wait, network server doesn't have threaded saves?
19:28:25 <TrueBrain> _dp_: I was surprised too, honestly, but it seems they are forcefully disabled, yes
19:28:33 <TrueBrain> must be a reason there :D
19:28:35 <TrueBrain> not documented of course
19:28:43 <TrueBrain> because .. "it is obvious, not" :D
19:28:56 <TrueBrain> past-us hunting us down :P
19:30:15 <TrueBrain> it was never possible, it seems
19:31:02 <TrueBrain> "The server is disabled from threaded-saving, but might be enabled in the future."
19:32:30 <TrueBrain> _dp_: it is threaded these days for joins
19:32:47 <TrueBrain> so "save" in console is blocking
19:34:05 <TrueBrain> the worst part, I am not even surprised :D
19:35:17 <TrueBrain> so basically, for joins, the servers starts making a savegame in a thread
19:35:22 <TrueBrain> and when data becomes available, it is send to the client
19:35:29 <TrueBrain> so during compression
19:35:33 <TrueBrain> it doesn't wait till compression is done
19:36:18 <TrueBrain> _dp_: I would guess you did not take that into account with your matrix overview? (no judgement btw; as I wouldn't)
19:37:26 <frosch123> maybe it's just to give a synchronous success/failure response?
19:37:28 <_dp_> well, it changes nothing for that matrix
19:37:34 <_dp_> as it doesn't even measure C++ :p
19:37:53 <TrueBrain> _dp_: no, but the time difference between lzma and zstd would be less, I would guess?
19:37:57 <TrueBrain> as things happen at the same time
19:38:01 <TrueBrain> where you assumed they do not?
19:38:10 <_dp_> well, if you add common blocking part yes
19:38:23 <_dp_> but it's relatively short afaict
19:38:47 <TrueBrain> btw, did you really assume a connection speed of 100 mbit/s? :D
19:39:12 <TrueBrain> that might not be realistic ;)
19:39:58 <_dp_> technology doesn't stay still
19:40:03 <_dp_> we need to look into the future :p
19:40:23 <TrueBrain> other random question: does your data mean that the average map-size, after compression, is 2MB atm, over all public servers?
19:41:01 <TrueBrain> that is really small
19:41:28 <frosch123> dp probably disabled trees
19:41:40 <DorpsGek> _dp_: 2118.4166666666665
19:41:44 <TrueBrain> well, he says all public servers, so I assume he connected to all :P
19:42:11 <_dp_> yeah, all 180 server I could connect to
19:42:12 <TrueBrain> what is the biggest you have?
19:42:53 <TrueBrain> either way, 2 MB means you need a 20 mbit/s connection to download it in 1 seconds .. so download-time is a non-factor, basically :P
19:43:01 <TrueBrain> 31MB, okay, more what I expected :D
19:43:21 <_dp_> second biggest < 10mb xD
19:43:55 <TrueBrain> yeah, but so the actual download speed is really not important .. so zstd would really make a difference
19:44:00 <TrueBrain> even with making the savegames bigger
19:44:46 <TrueBrain> anyway, enough joking around in saveload code .. _dp_ , if I were you I would look into zstd1.3 compatibility, as that sounds easier, but I personally would not be against handshakes and a server picking the best one :)
19:45:35 <_dp_> yeah, I looked, 1.3 seems doable but there were some performance improvements in 1.4
19:45:54 <TrueBrain> it is client-only, so they are less relevant, not?
19:46:00 <TrueBrain> as long as the format is compatible :D
19:46:26 <TrueBrain> and additionally, making autosaves use zstd would really be cherry-on-top for me
19:46:36 <TrueBrain> as the sizes of those games is less important than getting it over with ASAP :P
19:46:48 <TrueBrain> honestly, personally, I could even live with making zstd:1 default
19:46:57 <TrueBrain> it is not like diskspace is expensive these days
19:47:56 <frosch123> is zstd that stable? the discussion read like it is still evolving
19:48:08 <frosch123> for long-term savegame storage you want a stable compression format
19:48:09 <TrueBrain> it is a Facebook product
19:48:13 <TrueBrain> and everything you now think of it, is true
19:48:21 <TrueBrain> frosch123: that is a very good point, okay, I take it back
19:48:25 <TrueBrain> no zstd:1 default :P
19:49:49 <frosch123> i only used one facebook product once (prestodb). it was a bit uff :p
19:50:10 <TrueBrain> it is really silly that most people have this idea how they make software, and it turns out to be true in many cases
19:50:13 <frosch123> luckily i could delegate dealing with it to an intern :p
19:50:19 <TrueBrain> what they did with the zstd API is just ... like .. wtf
19:50:27 <TrueBrain> their naming of functions was (and still somewhat is) a total shitshow
19:50:44 <TrueBrain> I read JGR also found this out recently :P
19:50:44 <frosch123> is it as bad as ruby?
19:50:53 <TrueBrain> I ignore that Ruby exists
19:50:55 <TrueBrain> so I wouldn't know :P
19:51:07 <frosch123> chop, chop! chunk, chunk!
19:51:07 <TrueBrain> we did not try gollum, and we did not use wikicloth
19:51:18 <frosch123> perfect names for a standard library
19:51:38 <_dp_> zstd seems stable but I can't find any definite statement on that so far
19:52:05 <_dp_> api is not that bad, they just have two names for same stuff atm and plan to deprecate one in future
19:52:08 <frosch123> if it is facebook, it's probably safe to assume it's dead in 5 years
19:52:25 <TrueBrain> it has been around for a few already, so who knows
19:52:31 <frosch123> s/dead/incompatible/ or similar
19:52:53 <_dp_> well, it's bsd license, we can always embed it :p
19:53:07 <TrueBrain> alternatives are popping up fast too
19:53:25 <TrueBrain> just take a few years for any to mature :)
19:53:48 <frosch123> joining is the only case where speed matters
19:53:57 <TrueBrain> autosaves too, I would say
19:54:06 <frosch123> autosave thread can take as long as it wants to?
19:54:40 <frosch123> why? because it has to finish within a month?
19:54:52 <TrueBrain> only know it annoyed me :P
19:55:08 <TrueBrain> we did discover that if you try 2 autosaves at once, the second is just ignored
19:55:12 <_dp_> well, it would be weird to have 2 autosaves going on at the same time...
19:55:12 <TrueBrain> so not really a reason in my opinion
19:55:22 <_dp_> and if there are 10 and they cycle...
19:55:30 <TrueBrain> but that doesn't happen
19:55:49 <frosch123> it's weird to tie autosaves to in-game time
19:56:00 <TrueBrain> shall we change it? :D
19:56:03 <TrueBrain> we have the meansi now
19:56:09 <frosch123> do autosaves never trigger when building in pause?
19:56:24 <TrueBrain> now there is a question ..
19:57:34 <TrueBrain> OnNewMonth is the only place that sets _do_autosave to true
19:57:45 <TrueBrain> so that would strongly suggest, during pause, no autosaves are ever made
19:58:18 <TrueBrain> really strange, if you think about it :D
19:58:30 <TrueBrain> well, the game used to have no other means of doing it, ofc
19:58:43 <TrueBrain> the same reason the whole network protocol is tight to ticks :)
19:58:52 <TrueBrain> lag-detection is not done on real time, but on game time
19:59:21 <frosch123> are you sure? i think building-in-pause also works in multiplayer
19:59:53 <TrueBrain> that would be new, let me try
20:00:05 <frosch123> i think network-frames also run while paused
20:00:16 <TrueBrain> ah, yeah, that would be the cheat, ofc
20:00:36 <TrueBrain> we can now change a few to real-time, would make it possible to change the game-speed server-side :D
20:00:55 *** didac has quit IRC (Remote host closed the connection)
20:01:34 <frosch123> did you already add the packet "upgrade connection to websocket"?
20:01:49 <TrueBrain> I have been staying away from the network for now, sorry :P
20:03:30 <TrueBrain> frosch123: and it seems I should now focus on stabilizing 1.11, making it even worse! :P
20:03:42 <supermop_Home> i was thinking, at least for the firs layout, maybe i should rotate the lobby building of the hotel
20:03:55 <TrueBrain> some weird reports about lag on both sdl and sdl-opengl
20:04:01 <TrueBrain> need to find a way to debug that shit
20:04:02 <frosch123> TrueBrain: by buying weird hardware with weird graphics cards?
20:04:31 <frosch123> TrueBrain: default for sdl changed from 8bpp to 32bpp
20:04:35 <supermop_Home> so both big atrium windows face out to the SE, instead of the current two staircase ends
20:04:39 <frosch123> just in case you forgot :)
20:04:43 <TrueBrain> frosch123: did it? It was 32bpp for me already
20:04:55 <TrueBrain> well, I dunno if that was already changed during 1.11 of course
20:04:55 <frosch123> 32bpp was for windows and osx
20:05:06 <TrueBrain> but my testing of OpenGL showed SDL was also doing 32bpp by default for me
20:05:20 <TrueBrain> which surprised me :D
20:05:48 <frosch123> maybe you were using zbase or abase :p
20:05:58 <TrueBrain> clean everything with OpenGFX
20:06:24 <TrueBrain> yeah, I even made note of it in the OpenGL thread
20:06:27 <TrueBrain> I wonder when that changed
20:07:59 *** qwebirc75610 has quit IRC (Remote host closed the connection)
20:08:49 <TrueBrain> dbg: [driver] Successfully loaded blitter '32bpp-sse2-anim'
20:09:20 <TrueBrain> clean checkout of the 1.10.3 tag
20:09:24 <TrueBrain> downloaded OpenGFX via bootstrap
20:13:09 <LordAro> zstd is stable enough that arch have switched to it for packaging
20:13:41 <TrueBrain> that is not nothing
20:15:47 <frosch123> TrueBrain: no, you are right... there is actually no os depencency... it's just a config-file-only setting "support8bpp" that defaults to "no"
20:16:05 <TrueBrain> otherwise I would have regretted my vote to make 32bpp default :P
20:17:23 <frosch123> even worse, it was me :p
20:17:37 <TrueBrain> 6 years ago, nice :D
20:18:19 <TrueBrain> directly after the release of 1.4 :P
20:26:30 *** arikover has joined #openttd
20:28:28 <frosch123> hmm, i am trying to write the changelog for NML...
20:28:52 <frosch123> apparently procedures were invented in the last 6 months :) i thought it was longer ago
20:32:26 <frosch123> switches with parameters and optimisations are new... FLHerne: glx: what do you think of going for 0.6.0 instead of 0.5.4 :p
20:33:31 <frosch123> oh damn, noone wrote docs for parameters...
20:33:48 *** arikover` has joined #openttd
20:37:06 <glx> procedures existed internally
20:37:38 *** arikover has quit IRC (Ping timeout: 480 seconds)
20:40:02 <andythenorth> frosch123 I think 0.6.0
20:42:23 <frosch123> andythenorth: do you know how nml procedures and parameters work?
20:42:45 <frosch123> can i volunteer you for the docs? :)
20:46:35 <andythenorth> we could write docs on livestream?
20:47:00 <FLHerne> frosch123: Yeah, besides that there's been a lot of code churn
20:54:51 <andythenorth> how to explain procedures in plain language?
20:54:56 <andythenorth> to non-programmers?
20:55:06 <andythenorth> I didn't understand them for years (in nfo)
20:55:45 <frosch123> "with nml 0.6.0 you can use switches inside expressions, and pass stuff to them, like this"
20:55:50 <andythenorth> "However, instead of branching, it is a subroutine call, with the value calculated by the called entry being used as variable value."
20:56:37 <frosch123> that's not for non-programmers, that's for fortran people
20:57:06 <andythenorth> I have lots of examples, mostly silly ones
20:57:11 <andythenorth> I won't include the loop unroll :P
20:57:40 <frosch123> oh, i forgot that. i agree, it was a bad idea to ask you :p
20:58:01 <FLHerne> frosch123: Should probably mention "Change: simplify the code and enforce usage of '()' for calls" as it's a breaking change
20:58:12 <andythenorth> I have some nice ones
20:58:31 <frosch123> FLHerne: to my understandting there were no calls before at all
20:58:39 <frosch123> so, nothing that could be broken
20:59:08 <andythenorth> can't remember why I do that instead of the nml built-in
20:59:48 <andythenorth> shouldn't I be using FORWARD_SELF(x) instead?
20:59:54 <andythenorth> with random switch?
21:01:22 <frosch123> no, that syntax should be labeled deprecated
21:01:31 <frosch123> it's in the top 5 of worst newgrf additions :)
21:01:51 <andythenorth> it's unfortunately presented
21:02:00 <andythenorth> because it seems like the 'correct' thing to use
21:02:16 <andythenorth> but I usually end up replacing it with random bits, because I don't understand random switch
21:02:29 <andythenorth> and discord say it all works fine
21:02:59 <frosch123> andythenorth: do you have an example without var[0x61,..] ? :)
21:03:41 <andythenorth> they all include templating so far
21:03:50 <andythenorth> big silly loops to build structures
21:05:02 <andythenorth> I found that they are rarely needed for simple cases
21:05:23 <andythenorth> they're worth using for complex situations
21:06:56 <andythenorth> there's a nice one in FIRS for terrain sprites, but it doesn't return anything useful, it relies on stuffing the sprite number into a register for reuse later
21:07:07 <andythenorth> cut a lot of repeated switches out though
21:09:52 <andythenorth> someone will tell me there's a built-in for that now :)
21:22:00 <peter1138> How do I explain regexs to normal people? :/
21:22:16 <TrueBrain> 1 problem + regex == 2 problems
21:22:17 <andythenorth> there are 10 kinds of people etc
21:22:34 * andythenorth has avoided regexp forever
21:22:40 <glx> and regex format depends on the language
21:22:59 <frosch123> peter1138: it's meta-magic done with runes
21:23:00 <andythenorth> glx procedures are really helpful, I've removed so much duplicated code
21:23:16 <andythenorth> and made my own pseudo-variables where the ottd ones are 'broken'
21:24:34 <frosch123> is that was grandparents tell their grandchildren? i wrote a working regex once?
21:25:34 <glx> usually it needs a lot of trial and error
21:27:10 <andythenorth> but I have used multi-threading to generate n problems
21:27:19 <andythenorth> I only fork-bombed one person I think
21:31:14 *** sla_ro|master has quit IRC ()
21:33:53 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
21:40:38 <glx> pff issue linking doesn't work with PR title
21:51:53 *** Samu has quit IRC (Quit: Leaving)
21:55:47 *** jottyfan has joined #openttd
22:29:08 *** didac has quit IRC (Remote host closed the connection)
22:33:43 *** andythenorth has quit IRC (Quit: andythenorth)
22:41:59 <TrueBrain> I still have Steam stuff to translate ... we need to find a way to extend eints for those things :D
22:43:37 <glx> I think eints can have multiple projects
22:45:32 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
22:46:35 <glx> but for baseset description translations (3 strings in total I think), creating 3 projects would be overkill
22:47:11 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:49:15 <TrueBrain> dunno, they are cheap to create as far as I know
22:49:19 <TrueBrain> but Steam stuff doesn't fit in there
22:49:31 <TrueBrain> I can always copy/paste them out of it
22:49:41 <TrueBrain> will have to ask frosch123 how he sees that, but he just run by the reading of it :P
22:49:56 <glx> ah no there are 4 strings in opengfx (2 matters), 1 in opensfx and 1 in openmsx
22:50:19 <TrueBrain> best would be if they are separate projects but show as one, I think
22:50:29 <TrueBrain> we should do something with it anyway :)
22:52:09 <glx> they seem to be typical newgrf lang format, so eints should handle them (I think eints was used for them when they were in ottdcoop repo)
22:53:06 <TrueBrain> yeah, the Open**X shouldn't be a problem
22:53:25 <TrueBrain> it is just that we need to undo some hacks if we want that :)
22:53:28 *** jottyfan has quit IRC (Quit: jottyfan)
22:53:33 <TrueBrain> eints is not build for 1 project .. it normally shows a list, if I remember correctly
22:53:39 <TrueBrain> but Steam is more of an odd duck
22:53:47 <TrueBrain> but I can put it in our lang format
22:53:52 <TrueBrain> and just once in a while copy it out of there
22:53:53 <glx> for steam you first need a way to automate stuff on the steam side
22:53:55 <TrueBrain> or maybe Steam has an API
22:54:03 <TrueBrain> it doesn't -need- to be automated, honestly
22:54:27 <glx> well if you need to copy/paste by hand it's suboptimal ;)
22:54:38 <TrueBrain> but better than no translations!
22:57:01 <glx> anyway, time to sleep for me :)
23:00:50 <LordAro> TrueBrain: we could just copy all the sfx/gfx/msx/dteam strings into the bottom of english.txt, then copy them whereever before doing a release
23:01:15 <LordAro> not exactly the most automated thing in the world, but would save all the extra infrastructure
23:08:58 <FLHerne> Supposedly it launched about three hours ago, so we'll see :p
23:15:57 <LordAro> it's doing rather more up this time
23:21:24 *** jottyfan has joined #openttd
23:25:09 <DorpsGek> [OpenTTD/OpenGFX] Andrew350 updated pull request #46: Change: Use a white cursor for better visibility https://git.io/JLyY4
23:26:39 <TrueBrain> LordAro: my question was about Steam strings, but you and glx keep hijacking it for Open**X :p
23:27:01 <TrueBrain> Don't think they should be in OpenTTD repo :p
23:27:25 <michi_cc> No boom this time :P
23:27:31 <TrueBrain> (See the About the Game on Steam)
23:32:21 <michi_cc> I retract my earlier statements.
23:33:06 <TrueBrain> Your fault now michi_cc :p
23:33:10 <DorpsGek> [OpenTTD/OpenTTD] pelya commented on issue #8799: Memory corruption in 1.11.0-beta2 on Android NDK in multiplayer servers list https://git.io/JtjjW
23:34:09 *** grossing has quit IRC (Remote host closed the connection)
23:34:12 *** grossing has joined #openttd
23:35:37 <arikover`> I was wondering: is it possible to change the base sounds with a NewGRF? Other than a new Base set?
23:48:07 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:51:12 <Eddi|zuHause> i think that's possible, but not sure. newgrfs can include sounds, but i don't know if they can override the existing ones
23:58:28 *** Extrems has joined #openttd
continue to next day ⏵