IRC logs for #openttd on OFTC at 2021-01-05
⏴ go to previous day
00:08:41 *** DasPoseidon has left #openttd
03:08:27 *** Wormnest__ has quit IRC (Quit: Leaving)
03:21:58 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:18:04 *** tokai|noir has joined #openttd
04:18:04 *** ChanServ sets mode: +v tokai|noir
04:24:59 *** tokai has quit IRC (Ping timeout: 480 seconds)
04:46:59 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
05:20:31 *** roadt has quit IRC (Ping timeout: 480 seconds)
06:57:29 *** andythenorth has joined #openttd
07:01:25 *** andythenorth has quit IRC ()
07:02:30 *** sla_ro|master has joined #openttd
07:13:31 *** andythenorth has joined #openttd
07:16:28 *** snail_UES_ has quit IRC (Quit: snail_UES_)
08:00:42 *** WormnestAndroid has quit IRC (Remote host closed the connection)
08:00:55 *** WormnestAndroid has joined #openttd
09:19:38 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLbws
09:26:28 <andythenorth> also I should not be allowed to rebase anything ever
09:26:33 <andythenorth> I think I might alias rebase to ''
09:27:19 *** DasPoseidon has joined #openttd
09:29:45 <andythenorth> I should stop following the instructions
09:29:56 <andythenorth> " (use "git pull" to merge the remote branch into yours)" gets me every time
09:30:03 <andythenorth> it's force-push at that point
09:30:44 <TrueBrain> "git pull" is a horrible command to use
09:30:49 <TrueBrain> I literally never use it
09:30:57 <TrueBrain> if I would be to alias anything to "", it would be "git pull"
09:31:02 <TrueBrain> shitty command being shitty
09:39:07 <TrueBrain> yup .. too much magic
09:39:18 <LordAro> i think it does have a mode for automatically rebasing now
09:39:22 <LordAro> which probably makes it worse
09:52:56 <TrueBrain> yeah, as people will not understand the errors :P
09:53:01 <TrueBrain> but it is better than the magic merge
09:56:48 *** Compu has quit IRC (Remote host closed the connection)
10:11:23 *** Westie has quit IRC (Ping timeout: 480 seconds)
10:33:37 <andythenorth> seems vanilla to me
10:37:44 <TrueBrain> it is already based on population; guess we should get the other PR in for bridges to be the same :)
10:40:19 <TrueBrain> guess we can just pick them up both, clean the minor boo-boos, sync them to be the same, and push them .. would be nice :)
10:41:49 *** jottyfan has joined #openttd
10:42:29 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8485: Codechange: improve performance for complex vehicle chains by resolving sprites less often https://git.io/JL5pN
10:42:56 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8458: Fixes for issues with rotated (non-rectangular) airports https://git.io/JLy1U
10:42:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #8437: Airports: Improper aircraft movement when the northernmost airport tile is not part of the layout https://git.io/JLMPY
10:43:34 <TrueBrain> yeah, no clue why you didn't merge them yet LordAro :P
10:44:46 <LordAro> mostly it's being unsure of things
10:44:54 <LordAro> and wanting someone else to also approve it
10:45:11 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8457: Change: Ability to convert between road types for town roads and bridges now indi… https://git.io/JLyEp
10:45:39 <TrueBrain> shrug .. pretty sure we can have lengthy debates about these kind of PRs .. but nobody found a regression, so to the public! :P
10:49:07 <TrueBrain> guess it is fine with it being a setting .. it is not so much about gameplay, but more about how to use the interface, I guess
10:49:12 <LordAro> idk, i think i like it
10:49:27 <TrueBrain> it is one of these PRs I am fine with it being a setting :D
10:49:47 <TrueBrain> it would be horrible for new players
10:49:51 <TrueBrain> that is mostly my argument
10:50:02 <LordAro> mm, having it default off is fine, imo
10:50:12 <LordAro> needs a rebase though :D
10:50:16 <TrueBrain> yeah, and code cleanup
10:50:32 <TrueBrain> _settings_game is sync'd between client and server, is it not?
10:50:40 <TrueBrain> I wonder if sending the bit via DoCommand is useful
10:50:57 <LordAro> i was wondering about that
10:51:07 <LordAro> i certainly would expect it to be synced
10:51:08 <TrueBrain> I think it can just pick up the setting where it is needed .. no need for all this routing in of information, I think
10:51:17 <TrueBrain> okay, something on my todo-list :)
10:51:31 <LordAro> though you could argue that passing it in the command uses less global state :)
10:51:45 <TrueBrain> if the state wasn't already there, you are right :P
10:53:25 <TrueBrain> fun, DC_NO_RAIL_OVERLAP is not used anywhere :P
10:53:29 <TrueBrain> seems a remnant of the old AI
10:54:57 <andythenorth> 8274 signals, the existing behaviour does annoy me in game
10:55:18 <andythenorth> I have a twitch about the PR, not sure why, maybe just it's more combinatorial behaviour
10:56:06 <LordAro> TrueBrain: freeing up command numbers would please me
10:56:33 <TrueBrain> should I renumber the DC numbers?
10:56:37 <TrueBrain> or just leave a hole
10:57:06 <LordAro> renumbering would probably cause all sorts of issues with external things, i would imagine
10:57:15 <LordAro> andythenorth: possibly
10:58:03 <LordAro> though... that comment from UnsuspiciousGooball implies that it's already a thing?
10:58:09 <LordAro> not sure i understand
10:58:44 <andythenorth> that's a different issue, afaict
10:58:47 <andythenorth> I reproed that behaviour
10:59:28 <andythenorth> the train can't find a path after converting to the powered railtype, but it's fixed by reversing the train
11:00:05 <Eddi|zuHause> that sounds like a path reservation thing
11:00:07 <andythenorth> I have seen the behaviour Gwyd talks about, but only after reloading railtype grfs where the railtype definition changed
11:00:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbi0
11:00:27 <andythenorth> the solution is then to crash a train into the stuck train
11:00:34 <TrueBrain> I do not understand what the HasSignals block does, but I think I did that PR correctly :)
11:00:54 <TrueBrain> signals are fine if you make straight lines or something? I dunno
11:01:21 <andythenorth> anyway no repro, no grf list, no save, no screenshot etc for 8314, so not the best of reports :)
11:02:44 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbiV
11:03:07 <TrueBrain> LordAro: would that be helpful to anyone?
11:03:15 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbir
11:03:31 <TrueBrain> andythenorth: I FOUND A MEERKAT
11:04:02 <Eddi|zuHause> TrueBrain: i imagine someone in 10 years wants to add a flag and check whether he can reuse that?
11:04:12 <TrueBrain> it is unused; he can use it
11:04:16 <TrueBrain> this is not stored in savegames or something
11:04:24 <TrueBrain> it is just runtime information
11:04:36 <Eddi|zuHause> yeah, but he will probably embark on a discovery first what this was before
11:04:45 <TrueBrain> why would anyone do that? I heavily doubt that
11:05:01 <TrueBrain> neither are you going to do that when you want to use 0x1000
11:05:19 <TrueBrain> I left a comment as otherwise you might miss there is a hole there, that is all
11:05:43 <TrueBrain> remember it hasn't been used for 11 years now :P
11:05:58 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbi1
11:06:45 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbiy
11:06:58 <TrueBrain> LordAro: what is your reason for triggering on that line? :)
11:07:07 <TrueBrain> why you think history is important here?
11:07:37 <TrueBrain> (again, no sarcasm; IRC can be tough on that :D)
11:09:19 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbi7
11:09:42 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbid
11:10:01 <michi_cc> TrueBrain: Just renumber, then nobody will complain, even if you have a lot more changed lines :p
11:10:21 <TrueBrain> well, I honestly wonder what is in LordAro 's mind :) It might be insightful ;)
11:10:32 <TrueBrain> something triggered it :D I might be missing something :)
11:10:53 <LordAro> basically the same as Eddi
11:11:01 <TrueBrain> okay, I can understand that argument
11:11:04 <TrueBrain> it just doesn't apply here :)
11:11:08 <TrueBrain> this is not a spec :)
11:11:10 <TrueBrain> it is an internal value
11:11:17 <TrueBrain> but I went through the history
11:11:19 <michi_cc> It makes it a tiny bit to look like some of the bitfields that are saved somewhere.
11:11:22 <Eddi|zuHause> i wasn't going to comment on it, but then LordAro did and i had to chime in
11:11:26 <TrueBrain> we removed DC flags more often without any mentioning it ever existing :)
11:11:27 <LordAro> even for things like admin port?
11:11:47 <TrueBrain> only if you have a fake client, this matters
11:11:55 <TrueBrain> (as DoCommands are sent over TCP)
11:12:00 <LordAro> i thought we cared about thirdparty versions now? :p
11:12:02 <TrueBrain> but ... yeah ... :P
11:12:02 <michi_cc> Renumbering will trick the mind (and kick all fake clients :)
11:12:27 <TrueBrain> pretty sure you cannot fire a DoCommand via the admin port :)
11:12:35 <TrueBrain> (not on DC_ flags level)
11:13:25 <LordAro> guess i thought they were a bit more exposed than they are
11:13:37 <TrueBrain> LordAro: if we care on this level about thirdparty, we should not remove it and leave it as a DONEVERUSEAGAIN :)
11:13:45 <TrueBrain> not as far as I know :)
11:14:22 <LordAro> that enum should probably have an explicit size on it too
11:14:43 <TrueBrain> I was triggered on that too because of another PR
11:18:21 <TrueBrain> owh, it is not even send over the network
11:18:26 <TrueBrain> CommandFlagsToDCFlags
11:18:34 <TrueBrain> only 3 flags are converted from network to DC_
11:18:39 <TrueBrain> DC_ purely is internal :)
11:19:20 <TrueBrain> DoCommandP is network-safe
11:19:29 <TrueBrain> DoCommand is the internal processing
11:19:35 <TrueBrain> the first does not have flags; they are embedded in Cmd
11:21:23 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbPu
11:21:58 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbPz
11:22:15 <TrueBrain> it is not about happy or not :) It is about: am I missing something and are we about to break everything, or not :D
11:22:30 <TrueBrain> it also answers: does the size of DoCommandFlag matter; answer appears to be: no
11:22:40 <TrueBrain> 4 bytes, 8 bytes, give it a size, it is fine, from what I can tell :)
11:24:22 <TrueBrain> DoCommand vs DoCommandP is just very confusing :D
11:26:46 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain dismissed a review for pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbPu
11:26:49 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbi0
11:26:58 <TrueBrain> (forgot a dot in the commit message, and changed 0x0010 to 0x010, as to be more like the other values)
11:28:07 <LordAro> well i'm not approving it again :p
11:28:47 <Eddi|zuHause> i'm fairly sure me approving anything doesn't actually help anyone :p
11:29:12 <Samu> oh, I'm gonna use that new entry for my DC_NO_FLOOD thingy
11:29:51 <TrueBrain> LordAro: but honestly, it was a good question you asked. As clearly I did not have an answer ready .. which is never a good thing :P
11:30:08 <TrueBrain> assumption can be evil :D
11:30:31 <TrueBrain> ah, I see why the auto-signal-remove needs a bit ... the setting is client-side only
11:31:08 <LordAro> all questions answered!
11:32:13 <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbXG
11:32:35 <TrueBrain> bit 0..2 ... that fits 7 values .. right
11:32:37 <TrueBrain> bits are annoying :P
11:32:53 <michi_cc> It's fitting even 8 values
11:32:56 <TrueBrain> doWeUseThis or do_we_use_this as variable-names?
11:33:01 <TrueBrain> michi_cc: haha, yes :D
11:33:40 <TrueBrain> well, strictly seen, zero is not a number :P
11:33:50 <TrueBrain> but that fact doesn't help anyone, ever
11:35:06 <TrueBrain> a | b << 3 === a | (b << 3) , right? Do we prefer one of the other?
11:35:37 <TrueBrain> of = over, typing hard
11:36:29 <michi_cc> As zero is part of N₀, I would very much consider it a number. But I'm sure Eddi has a dedicated opinion about that :D
11:37:01 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8496: Codechange: DC_NO_RAIL_OVERLAP is a remnant of the OldAI https://git.io/JLbi0
11:37:16 <Eddi|zuHause> you have to add more specifiers to "Number" to make 0 not one of them
11:37:46 <TrueBrain> you guys are looking at this too much from a math perspective :P
11:37:57 <TrueBrain> zero wasn't "invented" till years after the other numbers :D
11:38:32 <Eddi|zuHause> you don't "invent" maths. you "discover" maths, as it has been there before you found it
11:39:03 <TrueBrain> guess airquotes carry no meaning anymore :'(
11:39:06 <TrueBrain> okay, my_variable, fine
11:39:24 <Eddi|zuHause> and after you found it, you can usually find clues that it has always been there because you find people using workarounds for the thing that had been missing all along
11:39:27 <TrueBrain> coding style doesn't mention it ...
11:39:39 <TrueBrain> LordAro / michi_cc : any preferences on a | b << 3 vs a | (b << 3) ?
11:40:09 <LordAro> a friend of mine has "0 ∈ ℕ" on a tshirt
11:40:56 <Eddi|zuHause> TrueBrain: i was told to order it big endian, so (b << 3) | a
11:41:25 <TrueBrain> I .. do not even know what to say to that :P
11:41:47 <Eddi|zuHause> it's in #7000 somewhere, i think it was frosch123
11:42:56 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain requested changes for pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLbXb
11:43:47 <TrueBrain> okay, that are very minor things to fix in that PR
11:44:12 <TrueBrain> I could have done it myself, but user was active recently enough; so hopefully he is still interested :D
11:48:13 <LordAro> i looked at it earlier, then got distracted by removing our own implementations of std::min/max
11:48:28 <LordAro> then i got an incomprehensible compile error, and stopped
11:48:42 <Eddi|zuHause> those are the "best" :p
11:49:30 <TrueBrain> oeh, coding style issue :P
11:50:28 <TrueBrain> that PR cannot possibly be correct
11:51:10 <TrueBrain> IsValidID cannot be run before the "join_as--" is done
11:52:04 <LordAro> all those commands are very inconsistent about what they do with the company id
11:52:08 <LordAro> i looked into it before
11:52:18 <LordAro> beyond that, no idea what the state is :p
11:53:56 <TrueBrain> I ask because you did the last review ;) But I see ... this is one of N problems related to this :) And the solution is incorrect ... guess that needs more attention :)
11:55:06 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLb1E
11:56:28 <TrueBrain> I think I would prefer it as a client-setting .. but a company-setting would mean less code .. interesting balance :)
11:56:53 <TrueBrain> LordAro: challenge accepted!
11:57:25 <michi_cc> I wouldn't like signal auto-remove as a company settings. IMHO it is a pure GUI thingy.
11:58:18 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLb1y
11:59:03 <TrueBrain> LordAro: okay, that is just funny .. wth? :D
12:00:04 <LordAro> my only thought was that it's something to do with std::max not being (explicitly) inline
12:00:43 <TrueBrain> you at least get a relocate warning
12:00:47 <TrueBrain> I don't even get that :D
12:01:09 <TrueBrain> anyway, the code is really fishy
12:01:14 <TrueBrain> who defines static consts in a struct?!
12:01:25 <TrueBrain> moving them out fixes the issue
12:01:46 <TrueBrain> owh, this struct is really just a class hiding as a struct
12:02:36 <TrueBrain> ah, it is the Window system
12:04:03 <TrueBrain> I am not sure what a 'static const' in a struct is suppose to do, honestly
12:05:15 <TrueBrain> you want a variable to be part of a struct
12:05:18 <TrueBrain> but you also make it static
12:05:35 <TrueBrain> I cannot put my finger on it, but it hurts
12:05:38 <LordAro> that's one of those weird edgecases in C++, isn't it?
12:06:06 <LordAro> the thing that you have to declare inside, but define outside the struct
12:06:20 <LordAro> though i thought that changed with C++11 or something
12:06:28 <TrueBrain> well, for non-const, yes
12:06:32 <TrueBrain> but this is also a const
12:06:42 <TrueBrain> a static const inside a struct ... I dunno what the intentions are, honestly
12:06:52 <TrueBrain> as a const a compiler can already resolve compile-time
12:07:06 <TrueBrain> a "static int" inside a struct can also be resolved compile-time
12:07:30 <TrueBrain> error: ISO C++ forbids in-class initialization of non-const static member ‘FramerateWindow::VSPACING’
12:07:44 <TrueBrain> seems it is not fixed yet in C++17 LordAro :P
12:08:17 <LordAro> i presume scoping is the intention
12:08:46 <TrueBrain> so my second suggestion does the same, does it not?
12:09:05 <TrueBrain> I am applying C logic here on a C++ world, so don't believe anything I write :P :D
12:09:16 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLbMC
12:10:32 <TrueBrain> Eddi|zuHause: valid point; a setting wouldn't hurt if the NewGRF controls when it triggers, I guess :) Nice reply!
12:11:11 <TrueBrain> michi_cc: but what does it do, a "static const" vs "const" in a struct?
12:11:47 <michi_cc> Struct::member versus instance.member
12:11:58 <TrueBrain> owh, like that, fair
12:12:05 <TrueBrain> well, we use neither
12:12:07 <TrueBrain> we just use "member"
12:12:13 <TrueBrain> so I guess that is an easy solution in this case :)
12:12:33 <michi_cc> a non-static const is (formally) initalized at instatiation, not startup.
12:12:49 <TrueBrain> but compilers optimize it away anyway, do they not?
12:13:16 <TrueBrain> but it might still use memory on initialization, I guess
12:14:13 <TrueBrain> but okay, given that page, changing `const` into `constexpr` solves it too
12:14:57 <michi_cc> static const they should, const alone should in principal affect the memory layout of the struct unless the optimizer can prove that nobody takes the address of it anywhere.
12:15:46 <TrueBrain> I wonder if this is just a GCC bug LordAro found .. what does clang do :)
12:15:57 <TrueBrain> don't have it installed, bah
12:18:19 <TrueBrain> seems constexpr was also the intention here
12:22:06 <TrueBrain> did this happen in more cases, or just for this window?
12:22:23 <TrueBrain> and constexpr is pretty cool :D Hints even harder to compilers to use it as a constant :P
12:22:39 <TrueBrain> learning new things :D
12:22:45 <TrueBrain> owh, I forgot lunch ... that needs fixing!
12:24:25 <LordAro> constexpr is a *huge* thing in new C++ now
12:25:22 <LordAro> /usr/bin/ld: CMakeFiles/openttd.dir/src/framerate_gui.cpp.o: in function `FramerateWindow':
12:25:25 <LordAro> /home/lordaro/dev/openttd/src/framerate_gui.cpp:438: undefined reference to `FramerateWindow::MIN_ELEMENTS'
12:25:28 <LordAro> clang-11: error: linker command failed with exit code 1 (use -v to see invocation)
12:25:43 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8492: Fix: Don't allow towns to terraform floodable tiles https://git.io/JLdA0
12:26:27 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLbDG
12:27:39 <Samu> there is some code repetition in my code, kinda sucks to make a helper function for another helper function
12:37:20 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLbDS
12:54:04 <TrueBrain> LordAro: I can understand constexpr being huge .. it seems it can replace a lot of macros and shit :)
13:03:19 <michi_cc> consteval (C++20) even more. constexpr means "could be evaulated at compile time", consteval is "will be evaluated at compile time". Can be used for complete functions and stuff.
13:07:22 <TrueBrain> might reduce a lot of template magic, I can imagine
13:25:31 <TrueBrain> lol ... "connect" console command is indeed very broken :D
13:30:58 <Eddi|zuHause> for some reason my brain connected that to "this is fine" and now imagines you sitting in a burning room :p
13:31:58 <TrueBrain> okay, that whole company-joining-stuff is just broken, but also a bit by design
13:32:16 <TrueBrain> I am very tempted to just collect all the network issues, plan them out, and make a v3 protocol
13:32:26 <TrueBrain> not sure that patching up individual issues is going to help
13:32:44 <TrueBrain> so many bugs on bugs over the years ..
13:33:31 <andythenorth> 'but now we have experience'
13:34:04 <TrueBrain> well, that for sure is a true statement ... there are "several" years of working on network protocols between then and now :P
13:34:23 <TrueBrain> but no, it is more that there are a few things that are hard to resolve in the way the network currently works
13:34:42 <TrueBrain> for example, the bug I just reproduced: when you join a server which requires a company password, your client hasn't left the last server, but also hasn't joined the new
13:34:47 <TrueBrain> so you are in a very weird partial state
13:35:09 <TrueBrain> and in fact, the server itself hasn't forgotten about you :P
13:36:26 <LordAro> v3 will be json-based, right?
13:36:42 <TrueBrain> ..... I join you in leaving andythenorth :)
13:36:55 <TrueBrain> LordAro: I should be more specific: I didn't mean that I want to revise the protocol itself
13:36:59 <TrueBrain> just the packets (and order) therein
13:37:22 <andythenorth> TrueBrain new channel, everyone but LordAro ? :)
13:37:27 <andythenorth> oops bad behaviour
13:37:32 <TrueBrain> I kinda like LordAro , so nah, I am good here :)
13:37:53 * LordAro leaves before he is kicked
13:37:56 <TrueBrain> OpenTTD over WebSocket would be interesting :)
13:38:20 <SpComb> and then port the pathfinder to use GraphQL
13:39:08 <TrueBrain> right, lets put this bug in the freezer for now; there is no easy solution without breaking 10 other things, it seems :)
13:39:34 <Eddi|zuHause> so we are finally going web 2.0 now?
13:52:04 <TrueBrain> it looks more than it really is; some PRs just need some minor work
13:52:07 <TrueBrain> some issues are trivial to fix
13:52:28 <TrueBrain> would love to hear your opinion on the matter :)
13:53:55 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8492: Fix: Don't allow towns to terraform certain floodable tiles https://git.io/JLdA0
13:56:20 *** snail_UES_ has joined #openttd
13:59:41 <TrueBrain> well, there was a reason I suggested to not do 1.11 yet :) If you feel some don't belong there, do let me know :)
14:00:40 <LordAro> it's just a matter of whether we can get them all done
14:00:45 <LordAro> it's very optimisitic
14:01:18 <TrueBrain> yeah, I don't expect us to process them all
14:01:26 <TrueBrain> but it might give some priority in what to pick up :)
14:02:55 <TrueBrain> ah, it was my settings
14:03:59 <TrueBrain> if the aircraft is cleaned up before the news pops up, it is never shown
14:04:38 <andythenorth> such forest, such sunlight
14:10:09 <Eddi|zuHause> i guess this is a disguised steam train? :p
14:11:48 <TrueBrain> okay, crashed vehicle news is REALLY evil :)
14:11:51 <TrueBrain> no wonder you rarely see it
14:12:30 <Eddi|zuHause> is github slow or is it my internet? because that page is not loading
14:12:47 <TrueBrain> (duh; otherwise I could not have posted it :P)
14:13:16 <TrueBrain> but I have had issues with GitHub for weeks now .. it sometimes just loads VERY slow
14:13:54 <Eddi|zuHause> i thought we fixed that by separating "real" news from ticker messages
14:14:07 <TrueBrain> totally unrelated to the problem :)
14:14:47 <TrueBrain> that made it less bad
14:15:01 <TrueBrain> before that, 0 chance you seeing a crashed vehicle .. now 1% :P
14:16:04 <TrueBrain> it is worse with fast-forward, as news is tied to "realtime" clock
14:16:59 <TrueBrain> andythenorth: sadly, that has been a known problem
14:17:03 <TrueBrain> something with sanctions against a country
14:18:14 <TrueBrain> one solution for the news I can see, is to make crash-news-reports not tight to the vehicle, but to a viewport location
14:18:18 <Eddi|zuHause> new category: *BREAKING NEWS*
14:18:19 <TrueBrain> in that case the news remains
14:18:26 <TrueBrain> just might show an empty plot
14:19:24 <Eddi|zuHause> that sounds reasonable. make sure the news keeps the map location, and doesn't rely on the vehicle to access that location
14:20:09 <supermop_Home> ugh did TTO invent the idea of an endless scrolling ticker of news at the bottom of screen for cable news?
14:20:34 <supermop_Home> Chris Sawyer is responsible for Fox News D:
14:21:24 <Eddi|zuHause> "news ticker" pretty much predates TV by like a century :p
14:22:21 <supermop_Home> yes but more as a series of stock prices by telegraph that you make a concerted decision to look at
14:22:26 <TrueBrain> hmm ... the news ticket does appear this way, but you no longer know WHICH vehicle crashed
14:22:29 <TrueBrain> not sure that is better
14:23:22 <Eddi|zuHause> TrueBrain: long haul suggestion: keep the vehicle around in vehicle lists, but clear it from maps. and add a "rebuild in nearest depot" button or something
14:23:46 <TrueBrain> yeah, I was considering the same, but that would also keep vehicles in the vehicle pool
14:23:51 <TrueBrain> it needs cleaning some day, I guess
14:24:36 <supermop_Home> not as a endless distraction splitting your attention
14:24:50 <LordAro> can't the news message just be "precalculated" when it's generated? then no need for worrying about keeping references to vehicles around at all
14:25:05 <TrueBrain> LordAro: yes, that is the "based on tile, not vehicle" solution
14:25:12 <TrueBrain> but the problem is, the vehicle is cleaned up by then
14:25:17 <TrueBrain> so .. what vehicle crashed?
14:25:24 <LordAro> well, just save the location instead of the vehicle
14:25:27 <TrueBrain> news messages are "precalculated" in that sense :)
14:25:37 <TrueBrain> but what vehicle was it?
14:25:45 <Eddi|zuHause> supermop_Home: to my understanding, a "news ticker" was typically put up in places like newspaper offices, where they ran in the background and someone went through them whenever they ran out of interesting things to write about
14:26:01 <TrueBrain> how does the user know what vehicle crashed
14:26:03 <LordAro> but you've saved the name of the vehicle when generating the news message
14:26:16 <TrueBrain> you were thinking from the game :D That is solved, that is easy ;)
14:26:21 <TrueBrain> not from the player perspective :D
14:26:31 <LordAro> i'm not sure what the difference is
14:26:38 <TrueBrain> a player gets: aircraft crashed
14:26:38 <Samu> does the news about crashed vehicle also go missing for AIs?
14:26:41 <Eddi|zuHause> LordAro: how would the user find out what the order list was, after the vehicle is cleared?
14:26:45 <TrueBrain> so ... what aircraft was it? Theone from A to B, or from A to C?
14:27:11 <TrueBrain> sorry, I was not clear in explaining :)
14:27:20 <Samu> if the AI is busy doing something else before checking events?
14:27:48 <supermop_Home> Eddi|zuHause exactly - it wasn't bombarding a regular audience trying to look at something else
14:29:03 <Eddi|zuHause> supermop_Home: i'm trying to think back when i first was noticing something like that. must have been around 2000-ish
14:29:25 <LordAro> idk, location should give a pretty good clue to the player, along with the vehicle name
14:29:35 <TrueBrain> it wouldn't for my games :D
14:29:40 <Eddi|zuHause> supermop_Home: which was around the same time as 24h-news-channels started popping up
14:29:41 <TrueBrain> but okay, this was already a problem
14:30:10 <TrueBrain> disasters have the same issue atm :) News is removed when the disaster is removed
14:30:12 <Eddi|zuHause> supermop_Home: presumably in america this was a bit earlier than in europe
14:30:19 <TrueBrain> so an UFO can crash your rail, and by the time you would get the message
14:30:22 <TrueBrain> you see and hear nothing :D
14:31:15 <Eddi|zuHause> TrueBrain: i agree, "keep the news around" and "keep the vehicle info around" are two different issues.
14:31:44 <TrueBrain> one bug at the time :D
14:32:03 <Eddi|zuHause> technically those are features, not bugs :)
14:33:14 <Eddi|zuHause> because, as far as i can tell, the current behaviour is not broken, just undesirable
14:34:05 <TrueBrain> what is also funny: town celebrates, new aircraft at airport!
14:34:16 <TrueBrain> all the news abut the celebration is also removed
14:34:23 <TrueBrain> "THIS NEVER HAPPENED PEOPLE"
14:34:44 <supermop_Home> Eddi|zuHause here it was 95 or 96 i think - suspiciously just 1 year after Transport Tycoon!
14:35:56 <Eddi|zuHause> TrueBrain: imagine some people never got the news "First Zeppelin lands at Lakehurst, New Jersey" on account of it going up in flames
14:36:20 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLb55
14:36:30 <TrueBrain> anyway, easy fix :)
14:36:30 <supermop_Home> oooh precedent: first record of a news ticker being used as part of a regular broadcast was on NBC's Today, used when the program first premiered on January 14, 1952. Without the benefit of computer-generated headlines and digital on-screen graphics, the ticker was very different from the ones in use today. The Today ticker was an actual piece of paper with typewritten headlines superimposed on the lower third of the screen. T
14:36:30 <supermop_Home> on the smaller screens of the time, and was dropped not long afterward.
14:37:03 <supermop_Home> can we add breakdowns for the ticker?
14:37:16 <supermop_Home> complete with smoke sprite?
14:41:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLb55
14:41:09 <Eddi|zuHause> i think part of your message was cut
14:41:34 <TrueBrain> so happy it is not only me who has that issue :)
14:42:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLb55
14:42:25 <Eddi|zuHause> TrueBrain: i'm fairly sure it's a clientside issue of the sender, where it's not pre-cutting the message properly to fit into an IRC message body
14:42:40 <TrueBrain> it is a problem between client and server for sure
14:42:50 <TrueBrain> not sure what client supermop_Home uses
14:43:03 <TrueBrain> but I can imagine it could also be that the server sends a max-message-size that is incorrect
14:43:09 <Eddi|zuHause> so the server cuts down the message further, and doesn't bother sending the rest as a second message
14:43:10 <TrueBrain> so it might be somewhere in between the two :)
14:43:36 <TrueBrain> might be as easy as a disagreement of specs :)
14:43:49 <TrueBrain> (total-message-size, including NOTICE etc, or "message size" as parameter)
14:44:09 <supermop_Home> well the gist of the idea was that NBC tried a news ticker in 1952, superimposing a feed of an actual paper ticker tape at the bottom of the screen
14:44:39 <supermop_Home> but it was basically impossible to read on the like 6" screens back then, and the ticker kept breaking down
14:45:29 <supermop_Home> on my client side it shows it sent as two messages?
14:45:49 <Eddi|zuHause> at least, in europe the inflation of "Breaking News" wasn't ever adopted
14:46:23 <TrueBrain> on the smaller screens of the time, and was dropped not long afterward.
14:46:25 <TrueBrain> is what we received
14:46:26 <Eddi|zuHause> supermop_Home: yes, it was two messages, but part of the first message was silently dropped by the server
14:47:01 <TrueBrain> I have the same issue with my client :) (Matrix IRC bridge)
14:48:00 <supermop_Home> Eddi|zuHause here the permanent tickers didn't start until 9/11, but then never went away
14:49:03 <supermop_Home> so in the absence of ongoing crises whatever rehashes of regular news stories became the crisis
14:49:31 <Eddi|zuHause> supermop_Home: i mean stuff like live-helicopter-car-chases
14:49:33 <supermop_Home> before then they were used occasionally for sever weather / snow day school closures
14:50:20 <milek7> this is likely server bug, IRC allows 512 bytes per message
14:50:36 <milek7> so client calculates proper length to fit into 512b messages
14:50:41 <Eddi|zuHause> "Breaking News: In areas with more than 200 new infections per 100.000 inhabitants, people aren't allowed to travel further than 15km from their homes"
14:50:53 <TrueBrain> hmm .. can you use money-cheat in multiplayer as server ...
14:51:16 <milek7> but irc server stucks its own hostname or something into message, and now it overflows
14:51:24 <Eddi|zuHause> TrueBrain: i think you can't
14:52:00 <TrueBrain> starts singleplayer, gives lot of money, start server from savegame
14:52:22 <milek7> hmm, or maybe client should also substract hostname length, but I don't know if client knows it fully
14:54:43 <milek7> so it seems clients needs to substract its own hostname to calculate proper length
14:54:58 <milek7> but RFC don't say anything that it has to do it, so...
14:56:30 <Borg> irv private message looks like this: PRIVMSG <target> :<string>
14:56:41 <Borg> so.. usualy.. you need to substract around 20 bytes from 510...
14:56:57 <TrueBrain> hmm, found another fun bug when looking at another .. you can give other people 0 money in network games :)
14:56:59 <TrueBrain> that makes no sense :P
14:57:09 <Borg> but of course.. if you target multiple targets.. your <string> shortens
14:57:10 <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #8498: Add some way to sort or filter the vehicle list such that crashed vehicles are shown first. https://git.io/JLbdS
14:57:41 <Borg> also.. that 510 bytes per message limitation is only or client -> server... server <-> server can talk full blown
14:58:53 <Borg> just supermop_Home IRC client is b0rked.. thats all
14:59:14 <FLHerne> milek7: The RFC explicitly says that the 512-byte limit includes the command and any parameters
15:00:55 <FLHerne> > IRC messages are always lines of characters terminated with a CR-LF (Carriage Return - Line Feed) pair, and these messages shall not exceed 512 characters in length, counting all characters including the trailing CR-LF. Thus, there are 510 characters maximum allowed for the command and its parameters.
15:01:20 <milek7> but I don't think command is the issue there
15:01:23 <FLHerne> TrueBrain: Well, why not?
15:01:54 <DorpsGek> [OpenTTD/OpenTTD] Eddi-z commented on issue #8498: Add some way to sort or filter the vehicle list such that crashed vehicles are shown first. https://git.io/JLbdS
15:02:20 <FLHerne> supermop_Home: What was the missing section of your message?
15:02:20 <milek7> FLHerne: but the issue that server adds client hostname before broadcasting to other servers
15:02:33 <Borg> milek7: no! its not the issue
15:02:45 <Borg> servers <-> server and server -> client does NOT impose that limit
15:02:59 <milek7> his message was 430 char long
15:03:02 <Borg> its only client -> server to protect servers from flooding and excessive large messages
15:03:02 <FLHerne> supermop_Home: (or, how long is it)
15:03:20 <milek7> and PRIVMSG #openttd is around 20 or so
15:03:41 <milek7> so where ~60 characters went?
15:03:46 <Borg> milek7: well, then OFTC ircd is broken.. but.. I will do RTFS now..
15:04:05 <Borg> milek7: servers append this to every command originating from client: <nick>!<user>@host COMMAND ...
15:04:16 <Borg> lets see if they are as well length checked.
15:07:08 <Borg> yep.. my good old IRCd doesnt fiddle with messages.. but OFTC ircd perphaps do... its new age ircd after all.. added a lot of things..
15:07:50 <Eddi|zuHause> can we teleport crashed vehicles to a random depot, instead of deleting them?
15:08:15 <Eddi|zuHause> (imagine some crash investigation authority gathering the parts)
15:08:50 <Borg> milek7: yep.. and I found it.. its truncating the message to 510 bytes...
15:09:37 <Timberwolf> Eddi|zuHause: Go full classic Microprose: when there is a crash, switch to an X-COM tactical mission.
15:09:53 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8499: Fix: allow input of numbers greater than INT32_MAX for GiveMoney https://git.io/JLbF1
15:10:20 <Eddi|zuHause> Timberwolf: i haven't actually ever played X-COM
15:12:25 <Borg> milek7: so I wonder if this is standard or not.... if it is on majority of ircds.. to truncate messages to 510 bytes.. then clients should actully use 410 limit on PRIVMSG/NOTICES to be on safe side
15:13:04 <FLHerne> Eddi|zuHause: but why
15:13:39 <Borg> Timberwolf: haha :D good one
15:13:48 <Eddi|zuHause> FLHerne: so a player can figure out what he needs to replace
15:17:51 <Borg> twpol: nice... lets see irssi out of curiosity
15:20:16 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLbb0
15:21:44 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLbbV
15:22:37 <Samu> Is this good english? * Get the number of raised corners of two given slopes combined.
15:23:04 <Borg> twpol: ugh.. irssi code is fuzzy... primary function that talks to ircd.. checks for 510 bytes.. but trying out to figure how PRIVMSG are sent
15:23:09 <Samu> i smash a slope into another with a &, and then i count corners
15:29:46 <FLHerne> Samu: That sounds like a "why would you do that?!" problem more than an English one
15:30:23 <Samu> it's so that I don't repeat code
15:30:36 <Samu> sec, let me update PR in a min
15:31:51 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8492: Fix: Don't allow towns to terraform certain floodable tiles https://git.io/JLdA0
15:34:52 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
15:35:24 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8492: Fix: Don't allow towns to terraform certain floodable tiles https://git.io/JLdA0
15:36:55 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8492: Fix: Don't allow towns to terraform certain floodable tiles https://git.io/JLdA0
15:41:15 <supermop_Home> Eddi|zuHause i do wish a way to clone crashed vehicles lasted longer than the wreck on the road
15:41:48 <supermop_Home> this was more an issue for me playing tto in the 90s when i constantly was crashing busses and planes
15:42:19 <supermop_Home> and you couldn't clone so id have to build a new one and hope i could get the orders replicated
15:44:34 <TrueBrain> I think we all agree there :)
15:44:45 <TrueBrain> it seems I found someone willing to give that a spin to implement :)
15:45:10 <Eddi|zuHause> let's see about that :p
15:46:00 *** D-HUND is now known as debdog
15:46:50 <TrueBrain> well, you can't say I am not trying :D
15:47:00 <debdog> a list of formerly crashed vehicles where the user can replace each listed vehicle or remove it manually from that list. sounds nice
15:48:21 <TrueBrain> I am trying to backport a patch from JGRPP to upstream, but boy, that is not easy :D Git doesn't really make it simple to do so ... :D
15:50:36 <Eddi|zuHause> that might be tricky once the branches diverged a bunch
15:52:07 <FLHerne> TrueBrain: `git cherry-pick`?
15:52:20 <FLHerne> If there are a ton of conflicts, that's not really git's fault
15:52:56 <FLHerne> (and you can play with different merge strategies to see which gets least confused)
15:53:22 <TrueBrain> I know how cherry-pick works, tnx ;)
15:53:36 <TrueBrain> the problem is changes on top of the change, that you want to know about
15:57:29 <Borg> TrueBrain: did you tried to merge master to that sha1 on separate branch?
16:00:12 <TrueBrain> owh boy, and now I have to learn the Window system ... not sure if I want to :D
16:07:56 *** DasPoseidon1 has joined #openttd
16:07:59 *** DasPoseidon has quit IRC (Read error: Connection reset by peer)
16:07:59 *** DasPoseidon1 is now known as DasPoseidon
16:16:02 *** DasPoseidon has quit IRC (Ping timeout: 480 seconds)
16:17:58 *** Gustavo6046 has joined #openttd
16:27:03 <Samu> towns can build tunnels, hmm
16:27:45 <Samu> i say hmm, because tunnels can terraform
16:29:27 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8498: Add some way to sort or filter the vehicle list such that crashed vehicles are shown first. https://git.io/JLbdS
16:30:46 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8498: Add some way to find and recover crashed vehicles https://git.io/JLbdS
16:36:44 <supermop_Home> now i'm listening to the wave race 64 soundtrack
16:47:22 *** Wormnest has joined #openttd
16:48:04 *** Borg has quit IRC (Quit: leaving)
16:49:42 <TrueBrain> bah .. so I finish backporting a nice patch, to find out there are several bugs with it :)
16:49:58 <TrueBrain> I wonder if JGRPP also has those bugs, or if they are fixed there
16:50:56 <TrueBrain> see, another patch I missed .. lol
16:55:13 <glx> looking at how to do #8468 nicely, I just noticed GetCargoTransportedSortValue() is not aware of 16 out cargo :)
16:55:27 *** gelignite has joined #openttd
16:56:13 <glx> IndustryProductionSorter() is not either
16:56:53 <andythenorth> I recall some kind of shenanigan about how we'd fit all the cargos into the window :)
16:57:00 <glx> but transported only look at first 2 cargos
16:57:01 <andythenorth> I _thought_ it was solved :)
16:57:07 <andythenorth> oof transported :P
16:57:27 *** frosch123 has joined #openttd
17:03:47 <TrueBrain> OpenTTD chat is a bit weird, I am finding out :D
17:03:58 <TrueBrain> you can send from yourself to a company
17:04:08 <TrueBrain> so you can sneak money away without anyone noticing
17:09:07 *** DasPoseidon has joined #openttd
17:12:28 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
17:12:35 <TrueBrain> not perfect yet (see limitations), but looking for opinions ^^
17:12:56 <TrueBrain> LordAro is going to love me, as I removed 2 globals :)
17:18:32 *** jottyfan has quit IRC (Quit: jottyfan)
17:18:38 *** jottyfan has joined #openttd
17:19:06 <Samu> you're working around the clock
17:19:41 <frosch123> sounds like tax evasion
17:22:16 <andythenorth> oof tax return time looms
17:22:33 <andythenorth> January, the bleakest month :D
17:22:48 <milek7> borg: twpol: tbh it looks like protocol flaw
17:23:12 <Samu> i'd rather have the "give money" not exposed so easily
17:23:27 <milek7> imo only sensible thing would be to server splitting message, if it doesn't fit after processing
17:30:22 <andythenorth> so many ships I could draw
17:33:19 *** Progman has joined #openttd
17:33:32 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8499: Fix: allow input of numbers greater than INT32_MAX for GiveMoney https://git.io/JLNTi
17:37:09 <frosch123> i still don't know how integer promotion works :)
17:37:56 <Samu> can somebody fix that unary minus something already :p
17:38:14 <Samu> or do you want me to do it?
17:44:16 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLNkB
17:44:30 <Samu> glx, I just noticed IndustryTransportedCargoSorter isn't used
17:44:52 <Samu> i commented it out, and apparently everything still runs fine
17:45:37 <glx> it's listed in sorter_func
17:47:12 <Samu> removed it from there too
17:49:07 <Samu> oh, maybe you're right, let me try sort by transported
17:51:44 *** Gustavo6046 has joined #openttd
17:52:30 <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #8468: Fix #8316: Make sort industries by production with a cargo filter possible https://git.io/JLNkF
17:56:18 <glx> it's still a "global" but it's scoped
17:58:23 <Samu> transported filter also needs a similar treatment
17:58:34 <Samu> sort by transported, with a cargo production
17:59:07 <glx> transported does support more than 2 cargos yet anyway
17:59:15 <Samu> like for example, livestock percentage transported, sort by transported
17:59:26 <Samu> it messes up with grain, doesn't look sorted
18:00:13 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8499: Fix: allow input of numbers greater than INT32_MAX for GiveMoney https://git.io/JLbF1
18:02:13 <Samu> i kinda forgot about Sort by Transported %
18:02:24 <Samu> was too focused on Sort by Production
18:04:19 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick commented on pull request #8468: Fix #8316: Make sort industries by production with a cargo filter possible https://git.io/JLNIG
18:04:45 <glx> and I forgot a "not" in my last message :)
18:09:29 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNI0
18:10:08 <frosch123> more comments to follow
18:11:53 <FLHerne> frosch123: It's whichever is largest out of both arguments and `int`
18:14:39 <Samu> wow, Sort by Transported % won't be easy to fix
18:15:21 <Samu> i just looked into it, seems that it only takes the 2 first producing cargoes from the industry
18:16:00 <Samu> need to investigate better
18:21:38 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNIh
18:22:24 <LordAro> almost certainly #8066
18:22:56 <TrueBrain> it is a good moment to test if our nightly really is fixing all those MacOS issues ;)
18:25:11 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNLf
18:26:20 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNLk
18:26:56 <TrueBrain> I am happy frosch123 has as many WTFs as I had :P
18:27:52 <frosch123> albert's gui system is actually quite nice. you should train a bit with it :)
18:28:06 <TrueBrain> I spend more than an hour with it, fiddling to get this icon correct
18:28:09 <DorpsGek> - Update: Translations from eints (by translators)
18:28:18 <TrueBrain> what I do not like, at all, is that there was a lack of hitting the freaking enter in all the definitions :P
18:28:25 <TrueBrain> they are just huge blobs of text :D
18:28:33 <TrueBrain> but most of it feels inuitive
18:34:11 <TrueBrain> frosch123: I only do not understand why the View HQ button is bigger than the Password button by default
18:34:15 <TrueBrain> I cannot discover what causes that
18:35:46 <TrueBrain> I guess the spacer with minimal size of 90 causes it
18:36:11 <frosch123> as long as you font size is small: "view hq" = "password icon" + "password button" = 90
18:36:29 <frosch123> but when your font becomes bigger, that all breaks down
18:36:32 <TrueBrain> but that 90 feels magic to me :D
18:36:44 <TrueBrain> (of the view HQ button)
18:37:03 <TrueBrain> as if widgets auto-align to the same size in their container?
18:37:12 <TrueBrain> I am missing a piece of the puzzle :)
18:37:13 <frosch123> yes, that's the point of containers :)
18:37:28 <TrueBrain> okay, so the selection with a spacer causes the whole container to act like that
18:37:29 <frosch123> or rahter, that's the point of SetFill
18:37:55 <TrueBrain> The Company GUI is build up a bit annoying ... I expected a left column and a right column
18:38:00 <TrueBrain> but instead it is build from top to bottom
18:38:01 <frosch123> "SetFill" means "use all the space in the container"
18:38:16 <TrueBrain> yeah, okay, and everything in a selection is precalculated
18:38:38 <TrueBrain> (as in, the size of the first item in a selection influence the second one
18:38:50 <TrueBrain> I wonder if I shouldn't just completely rebuild the Company GUI (with the same look&feel)
18:39:00 <TrueBrain> but .... it is so easy to get lost in that :P
18:39:22 <frosch123> with the infrastructure/vehicle counts, it's not that easy
18:39:33 <frosch123> you could just set the size of the buttons in UpdateWidgetSize
18:39:37 <frosch123> and remove all the minimumsize
18:39:47 <TrueBrain> it does require me to copy/paste the STR_ of them, it seems
18:39:55 <TrueBrain> at least, that is what they do in all other windows, from what I can tell
18:40:02 <TrueBrain> I don't like code-duplication :)
18:40:23 <frosch123> it's a lot shorter than looking up the text via GetWidget
18:42:25 <TrueBrain> okay, syncing the sizes looks a lot better
18:42:48 <frosch123> but yes, the company window is one of the worst :)
18:42:52 <frosch123> buttons all over the place
18:43:13 <frosch123> some buttons are right of the stuff they belong to
18:43:16 <frosch123> others are at the bottom
18:43:32 <frosch123> colour schema button could be next to the colour schema text
18:43:45 <frosch123> manager name button could be next to the manager name
18:44:14 <frosch123> i think that's the reason i did not touch that gui in the location/rename PR :p
18:44:15 <TrueBrain> I just wanted to fix this very ugly password buttno :P
18:44:31 <andythenorth> let's not talk about colour schemes window though :)
18:44:35 <TrueBrain> yeah, any idea why the A is misaligned btw frosch123 ?
18:45:19 <frosch123> TrueBrain: it's misaligned because there is nothing to align it to :p
18:45:30 <TrueBrain> but other buttons are aligned ..
18:45:38 <frosch123> aligning icons and texts does not work, because icons and text are scaled differently
18:45:49 <frosch123> aligning icons does not work, because the X is not aligned in the first place
18:45:59 <TrueBrain> X is off by a pixel or so, yes :D
18:46:24 <frosch123> technially i have to align the icon with the X of the original graphics
18:46:31 <frosch123> ogfx can align the A differently
18:46:35 <frosch123> but it will never align with the text
18:46:55 <frosch123> so, a pencil is probably better than a letter
18:47:03 <TrueBrain> that solves the issue too, yes :D
18:47:08 <frosch123> but a pencil in 10x10 in monochrome is kind of hard
18:47:25 <TrueBrain> the URL I linked had some other suggestions, I guess
18:47:41 <TrueBrain> grrrrr, now my password icon is again in the wrong place :P
18:47:52 <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLNLb
18:48:22 <TrueBrain> I have to add a spacer in the vertical container, where I expect the horizontal container
18:49:06 <TrueBrain> oops .. SetFill takes x / y
18:49:12 <frosch123> TrueBrain: i guess "DrawSprite(SPR_LOCK, PAL_NONE, r.left, r.top);" should be changed to "DrawSprite(SPR_LOCK, PAL_NONE, rtl ? r.left : r.right - r.width + 1, r.top);"
18:49:46 <frosch123> so the icon is right-aligned for LTR languages, and left-aligned for RTL languages
18:51:06 <TrueBrain> owh, you can right-align stuff .. I was just padding it on the left, or at least, trying to :D
18:51:18 <TrueBrain> r.width does not exist :(
18:51:41 <andythenorth> I will probably draw some icons
18:51:42 <frosch123> well, i means the sprite width
18:51:53 <frosch123> r.width was wrong anyway :)
18:51:58 <andythenorth> 10x10 mono is quite a jump from drawing ships :P
18:52:34 <frosch123> TrueBrain: ah, you can also insert another spacer
18:52:44 <frosch123> make one space do the fill, and make the icon one fixed size
18:52:55 <TrueBrain> well, I expected I had to do that inside the horizontal container
18:52:58 <TrueBrain> but I have to do it in the vertical
18:53:07 <TrueBrain> which I do not understand
18:53:10 <TrueBrain> but it does work :P
18:53:17 <frosch123> andythenorth: i used some grey in the eye icon, so not pure mono
18:53:23 <TrueBrain> owh, I think I get it
18:53:52 <andythenorth> ok pencil challenge :P
18:54:07 <andythenorth> are our icons now just mono then?
18:54:11 <andythenorth> I guess they're just sprites eh
18:54:29 <glx> when touching GUI IIRC it's nice to try hebrew to check alignment ;)
18:55:30 <frosch123> but you have to remember the location of the change language combobox in advance, and be able to mirror it in your head :)
18:55:52 <Xaroth> or learn hebrew very very quickly.
18:56:06 <Wolf01> Or change the openttd.cfg manually
18:56:11 <TrueBrain> is there a way to debug windows, like where it draws rectangles around the widgets or something?
18:56:22 <TrueBrain> or start the game with -x guys ... use the parameters available! :P
18:56:57 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
18:57:11 <frosch123> i don't think we have that
18:57:11 *** WormnestAndroid has joined #openttd
18:58:13 <TrueBrain> hmm .. vertical alignment of icon is a bit weird
18:58:34 <TrueBrain> frosch123: you don't think we have the -x I have been using all day? :D
19:00:24 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNtY
19:00:31 <TrueBrain> (sorry, out of sync comments are fun :P)
19:00:47 <TrueBrain> it is just hard to debug now from time to time to see where a container is .. guess I could color them somehow
19:00:54 <TrueBrain> well, got it to work, so happy days \o/
19:01:24 <andythenorth> I broke the image how?
19:02:29 <andythenorth> maybe it hates indexed pngs
19:03:51 <LordAro> i was just about to say
19:04:08 <TrueBrain> interesting, what you did there andythenorth :)
19:04:21 <TrueBrain> andythenorth: you clicked save too quickly
19:04:38 <andythenorth> anyway a version
19:07:58 * supermop_Home had mulled wine at lunch
19:11:59 <andythenorth> eye icon is harder
19:12:05 <andythenorth> ellipse shape in 10 px
19:12:11 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
19:12:12 <TrueBrain> not done yet, still need to fix chat ^^
19:12:18 <andythenorth> with round pupil and highlight :P
19:12:18 <TrueBrain> but the window thingies should be fine now :D
19:12:29 <andythenorth> we want 'eye' yes, not 'cross-hairs' like maps use?
19:12:29 <TrueBrain> tnx a lot frosch123 for the pointers :)
19:13:23 <supermop_Home> andythenorth i feel like a google map pin would be intuitive
19:13:49 <andythenorth> pin pins the window though
19:14:30 <supermop_Home> i like the crosshairs, but to a newcomer idk how they could possibly be understood to mean move the view to this location
19:15:30 <supermop_Home> the eye is a little surreal though
19:16:00 <frosch123> andythenorth: it only has to fit the original baseset, make it SF style :p
19:16:50 <TrueBrain> okay, a real n00b question: DoCommandP is called with CMD_MSG(STR_ERROR_INSUFFICIENT_FUNDS)
19:16:54 <TrueBrain> but this is not always the correct error ...
19:17:21 <TrueBrain> can you have more than one error from a DoCommandP?
19:17:46 <frosch123> the command can return a message
19:18:00 <frosch123> the CMD_MSG is only used, if the command returns fail without message
19:18:56 <TrueBrain> so much fun! A "grep" on the wrong thing can crash VSCode
19:19:00 <frosch123> andythenorth: i'll make some in-game screenshots with different window colors
19:20:54 <andythenorth> I'm a big fan of 8455 :P
19:21:09 <andythenorth> creates space in the vehicle window...
19:21:18 <TrueBrain> it is awesome, that is for sure :)
19:22:10 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNt7
19:23:06 <andythenorth> "Over 90 percent of pre-2000 gaming source code may already be gone."
19:24:05 <TrueBrain> sending chat in singleplayer crashes the game, shocker :)
19:24:22 <DorpsGek> [OpenTTD/OpenTTD] kabal2020 commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLNtp
19:25:48 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLNqU
19:32:30 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
19:32:30 <TrueBrain> okay ... now for the stupid chat stuff ..... hmmm
19:35:56 <supermop_Home> i think sim city 3000 was underrated
19:36:04 <TrueBrain> ah, chat is done in a callback as you want to know self vs other, I guess
19:37:20 <supermop_Home> sc4 was more fun tho infuriatingly buggy
19:37:39 <supermop_Home> 3000 felt a little lost in the cracks
19:37:54 <supermop_Home> but had a lot of nice touches
19:38:07 <supermop_Home> cute sprite design, good music
19:38:16 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNqV
19:38:40 <DorpsGek> [OpenTTD/OpenTTD] kabal2020 commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLNqr
19:38:56 <frosch123> i think we need another icon iteration
19:39:39 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNq6
19:40:16 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JLNqP
19:40:40 <TrueBrain> pencil is a lot more clear than the A at least :D
19:41:00 <frosch123> andythenorth: i changed the pencil to 9x9, that's the size of the X
19:41:26 <TrueBrain> chat doesn't work in single-player .. what to use to tell about the money transfer there .. hmm
19:42:24 <frosch123> well, the X is 8x8, but it looks weird when the tip of the pencil extends over the bottom of the X
19:42:36 <frosch123> on the top the pencil can be longer
19:43:21 <frosch123> TrueBrain: does it need telling? only one player could issue it
19:50:28 <supermop_Home> write an AI that likes to send money?
19:51:08 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on issue #8498: Add some way to find and recover crashed vehicles https://git.io/JLbdS
19:51:29 <TrueBrain> frosch123: basically the callback was used as DoCommandP doesn't know the player who did somethng, only the company
19:51:39 <TrueBrain> but I guess it is perfectly fine to mention which company wired money to what company
19:51:42 <TrueBrain> instead of which player
19:53:08 <frosch123> haha, true, now the culprit can push blame onto another company member
19:53:18 <TrueBrain> that is the only drawback
19:53:27 <TrueBrain> I can leave the callback in, but you are right, clients can block the message
19:53:32 <TrueBrain> so it is all a bit .... so-so
19:54:00 <Wolf01> I'm not sure about the eye icon
19:54:03 <frosch123> make it company only :)
19:54:24 <frosch123> there is no point in attributing actions to clients
19:54:36 <frosch123> they have to trust each other in every build action
19:54:42 <frosch123> so money is nothing special
19:54:42 <TrueBrain> yeah, that is what I did now
19:54:47 <TrueBrain> you see money from company to company
19:54:49 <TrueBrain> not which player did that
19:55:02 <TrueBrain> (well, not pushed yet)
19:56:15 <TrueBrain> ugh, getting company name is HARD :P
19:57:19 <TrueBrain> yeah, you have to use GetString
19:57:46 <frosch123> oh, i guess the old message used the language of the sending client :)
19:58:21 <TrueBrain> -STR_NETWORK_MESSAGE_GIVE_MONEY :*** {RAW_STRING} gave your company {2:CURRENCY_LONG}
19:58:21 <TrueBrain> +STR_NETWORK_MESSAGE_GIVE_MONEY :*** {RAW_STRING} gave {2:CURRENCY_LONG} to {1:RAW_STRING}
19:58:24 <TrueBrain> how is eints going to handle this?
19:59:26 <frosch123> it will remove the invalid translations from git
19:59:38 <frosch123> but still remember the, and display them as "invalid" to the translators
20:00:04 <frosch123> unless you reset the eints storage
20:00:16 <frosch123> (don't do that :p )
20:01:31 <Samu> the all cargo filter looks kinda bad still
20:01:40 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
20:01:53 <TrueBrain> frosch123: good, that is what I would expect :)
20:02:10 <TrueBrain> well, this PR went a bit overboard
20:02:52 <frosch123> you found a rabbit hole :)
20:03:26 <TrueBrain> well, your fault :P
20:03:34 <TrueBrain> you confirmed what I wanted to do, but refrained from doing :D
20:03:48 <TrueBrain> owh, and I created some merge-conflicts with it for JGR with this PR .. it does things differently enough that it doesn't cleanly merge
20:04:23 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
20:05:18 <TrueBrain> but, 1 less Cc, 2 less globals, and one ugly hack gone :D That should count for something!
20:05:43 *** Yacht has quit IRC (Quit: Page closed)
20:06:52 <TrueBrain> hmm, I can send which player did it
20:07:00 <TrueBrain> DoCommandP have a "text" field ofc
20:07:33 <frosch123> there is a special CMD_XYZ in command.cpp to make the server put the clientid into p1 or p2
20:07:36 <frosch123> the clients cannot fake that
20:07:50 <TrueBrain> both p1 and p2 are already used, but that is a good point
20:07:54 <Samu> can AIs give money? doubtful
20:09:14 <frosch123> still see no reason why the player is important though
20:09:41 <TrueBrain> okay, I will not purse this :P
20:10:43 <TrueBrain> okay, this is weird ...
20:10:48 <TrueBrain> MAX_LENGTH_COMPANY_NAME_CHARS is 32
20:10:51 <andythenorth> frosch123 does it look good? :)
20:10:54 <TrueBrain> NETWORK_COMPANY_NAME_LENGTH is 128
20:11:10 <TrueBrain> char msg[64]; was used in GiveMoney
20:11:14 <TrueBrain> so ... what is it? :D
20:11:59 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNYr
20:12:19 <frosch123> TrueBrain: i cannot comment on commit messages, maybe mention that client-language thing in the last commit?
20:12:48 <TrueBrain> only the company names were used from the sending client
20:12:52 <TrueBrain> the rest was translated correctly
20:12:54 <TrueBrain> is that worth mentioning?
20:13:20 <frosch123> oh, i thought it was send like a regular chat message
20:13:29 <frosch123> but that explains why noone reported it
20:13:41 <TrueBrain> but how many bytes is a company name ....
20:14:19 <frosch123> ottd limits all names with 32 unicode chars, and then reserves 4 times the size in bytes
20:14:42 <frosch123> so 32 vs 128 is CHARS vs BYTES
20:14:46 <TrueBrain> well, times 2 in the case of company name, as it turns out :P
20:14:53 <TrueBrain> but for networking, it is times 4
20:14:57 <TrueBrain> which is confusing as fuck :D
20:15:29 <TrueBrain> I was hoping to find a constant for it ..
20:15:36 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
20:16:13 <TrueBrain> MAX_LENGTH_COMPANY_NAME_CHARS * 2 valid? Or just weird? :D
20:16:19 <frosch123> TrueBrain: you find both "MAX_LENGTH_COMPANY_NAME_CHARS" and "MAX_LENGTH_COMPANY_NAME_CHARS * MAX_CHAR_LENGTH" in the source
20:16:48 <frosch123> no idea where you get the "* 2" from
20:16:55 <frosch123> but that looks wrong :)
20:16:59 <TrueBrain> the code already read [64]
20:17:29 <frosch123> well, replace them with MAX_LENGTH_COMPANY_NAME_CHARS * MAX_CHAR_LENGTH :)
20:17:32 <andythenorth> the i icon suffers from being an ellipse with sharp tails, locked in a square
20:17:38 <andythenorth> maybe we should give it eyelashes
20:18:06 <frosch123> messed more with the icons, time for new screenshots
20:18:32 <andythenorth> but the shape fits the bounding box better
20:18:44 <TrueBrain> frosch123: except or the fact it is stored in a savegame in some places .. so I cannot replace them easily
20:18:48 <TrueBrain> right, the rabbit hole stops here :D
20:19:12 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
20:20:10 <TrueBrain> this wasn't even in the 1.11 milestone, ffs :P
20:20:40 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNO0
20:22:23 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNOK
20:22:33 <TrueBrain> I like the eye andy posted
20:22:50 <frosch123> andythenorth: your eye is aligned with the debug icon, it should be aligned with the other 3
20:22:56 <andythenorth> mine is quite Addams family
20:23:35 <andythenorth> I aligned it with the pin, but it depends if we think the line is with windowshade icon
20:24:10 <andythenorth> windowshade is off-grid in my opinion, but it also makes sense for the feature
20:27:09 *** Progman has quit IRC (Remote host closed the connection)
20:27:47 <LordAro> TrueBrain: got a bit of time, are you waiting on me for anything/need a review/etc?
20:28:23 <TrueBrain> LordAro: you expect me to say no to that? :D
20:28:31 <frosch123> i put your eye one pixel higer, it looks better when the pupil is aligned with the windowsize-dot
20:28:39 <TrueBrain> waiting on you is strong, but .. :)
20:30:57 <andythenorth> I have made a horrible alternative eye
20:31:00 * andythenorth isn't posting it
20:31:32 <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic updated pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJsQJ
20:35:14 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
20:38:03 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLN39
20:38:40 <LordAro> #8063 is probably a bit too much, it turned into something bigger than i was expecting :)
20:38:47 <TrueBrain> LordAro: I indeed add those when squashing if I don't forget .. I find it highly annoying that every force push, the cmmit is mentioned in an issue :P
20:39:12 <TrueBrain> LordAro: too bad; close it? keep it around? (8063)
20:39:30 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8497: Fix: keeps news about accidents around after vehicle is cleaned up https://git.io/JLb55
20:39:37 <LordAro> TrueBrain: i mean too much for right now :p
20:39:52 <LordAro> though there's not exactly consensus on how it should be handled
20:44:23 <TrueBrain> I haven't really read up on it, I just now I find it weird I cannot change towns when I don't have RVs :P
20:45:46 <Samu> glx, if (this->industries.SortType() != 3) {
20:45:54 <Samu> I didn't want to use a number here
20:46:13 <Samu> do the filters have a name, enum or so?
20:46:34 <Samu> 3 means Sort by Transported
20:46:40 <Samu> 2 means Sort by Produced
20:46:43 <frosch123> you have to look closely :)
20:46:49 <Samu> 1 is by Type, 0 is by Name
20:47:37 <TrueBrain> Y looks like a totem, and now I cannot unsee that :D
20:47:44 <TrueBrain> A vs B is a toss-up, really
20:47:56 <glx> Samu: I don't think there's an enum as it was not needed
20:49:08 <Samu> hmm, where would i place this enum?
20:51:06 <Timberwolf> B / X but I don't like any of the eye options.
20:51:19 <TrueBrain> Timberwolf: make us something better? :D
20:51:42 <Timberwolf> You've seen how bad my UI sprites are, right? :p
20:52:03 <andythenorth> the pencils are potato / potato
20:52:04 <Wolf01> I don't like the eye at all for that button
20:52:16 <andythenorth> I don't like the eye but if we shipped it I wouldn't complain
20:52:48 <DorpsGek> [OpenTTD/OpenTTD] frosch123 approved pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNsM
20:53:28 <Wolf01> What about a camera or a binocular?
20:54:02 <Timberwolf> Yup. What's the resolution here?
20:54:24 <frosch123> 10x10 or 10x9 depending on how well it aligns with the other icons
20:55:25 <andythenorth> binoculars look like handcuffs :P
20:55:28 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:56:28 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8500: Change: move "give money" from client-list to company window https://git.io/JLNUZ
20:58:50 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNG3
20:59:30 <DorpsGek> [OpenTTD/OpenTTD] mattkimber commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNGn
21:00:15 <TrueBrain> sorry andythenorth , that doesn't do it for me :(
21:00:33 <Wolf01> Those binoculars need some depth :P
21:00:59 <andythenorth> I tried rotating them, but nah
21:03:37 *** Gustavo6046 has joined #openttd
21:03:57 <Samu> if (this->industries.SortType() != IDW_SORT_BY_TRANSPORTED) {
21:04:32 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNGD
21:06:33 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNG7
21:06:45 <frosch123> andythenorth: too big, you made it 11x11
21:07:01 <Wolf01> OMG, I tried to draw something in paint... it's... hideous
21:07:14 <andythenorth> TrueBrain the map pin?
21:07:29 <andythenorth> they're really hard to do without vectors
21:07:37 * andythenorth has spent a year working on map stuff
21:08:33 <andythenorth> hmm looks like a toy balloon
21:09:13 * Timberwolf likes the target and snowflake-target options.
21:10:48 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNZU
21:10:57 <andythenorth> ^ too heavy at the moment
21:11:09 <andythenorth> also relies on grey pixels for anti-aliasing, dislike
21:11:23 <TrueBrain> if that can be made a bit nicer in such a small dimension, that might be obvious to everyone
21:11:37 <LordAro> that is a lot of buttons
21:11:39 <TrueBrain> but now it indeed looks odd :D
21:12:05 <TrueBrain> either way, I am fine with X. It aint perfect maybe, but it is clear once you pressed it once :D
21:12:58 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNZY
21:12:59 <andythenorth> it's a trivial improvement later if someone can do better
21:13:22 <andythenorth> ooh the movie camera
21:13:40 <andythenorth> CCTV camera is most appropriate eh
21:13:41 <frosch123> yay, someone recognised it :)
21:13:57 <TrueBrain> but it is currently not centered :P
21:14:03 <TrueBrain> (in both X and Y :P)
21:14:10 <Timberwolf> Yeah, I think the film camera has potential.
21:14:19 <frosch123> Y won't happen, X is possible
21:14:21 <Timberwolf> I didn't think of doing one without the tape reels.
21:14:39 <andythenorth> ok well that was more fun than other things
21:14:49 <TrueBrain> owh, I meant x-axis and y-axis, not icon X and Y, to be clear :D
21:14:59 <frosch123> i understood that :)
21:15:44 <andythenorth> can we draw a predator tracking drone in 10x10? :P
21:16:05 <TrueBrain> can we make that if you look really close, you see another universe?
21:16:49 <andythenorth> TrueBrain there are fractal videos on YouTube, try those :P
21:18:47 <DorpsGek> [OpenTTD/OpenTTD] Wolfolo commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNZa
21:19:50 <TrueBrain> I kinda like the way the camera looks of Wolf01
21:20:17 <TrueBrain> too much pixels ... E_PIXEL_OUT
21:20:29 <Wolf01> I also fixed the binoculars, they looked like a cat at the first attempt
21:22:32 <DorpsGek> [OpenTTD/OpenTTD] mattkimber commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNZD
21:23:27 <TrueBrain> that is a nice camera too
21:24:45 <FLHerne> It's a nice-looking camera, but it makes me think 'screenshot' more than 'jump to position'
21:24:58 <TrueBrain> yeah ... the map pin still is most clear for me
21:25:03 <TrueBrain> but making that look good is HARD
21:25:24 <FLHerne> Yeah, the Google-esque one is a bit anachronistic for a '90s interface
21:25:25 <andythenorth> I think 'pin' and 'pin' is confusing
21:25:44 <TrueBrain> everyone knows the mappin by now :P
21:26:18 <Wolf01> FLHerne, yeah, the eye suits better for that, they had strange taste back in those years :P
21:26:51 <LordAro> andythenorth: i hadn't considered that they're both "pins"
21:26:55 <andythenorth> also the eye echoes the icon we're replacing
21:27:03 <FLHerne> What about inward-pointing arrows?
21:27:18 <Wolf01> <andythenorth> also the eye echoes the icon we're replacing <- was it a "I"?
21:27:26 <andythenorth> it's a big eye :P
21:27:56 <Wolf01> Nice, but it might look like an iron cross at 10x10 :S
21:28:36 <TrueBrain> I am all for these kind of changes LordAro , but not a C++ expert, so ... no clue what the drawbacks are :P
21:29:06 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNnt
21:29:42 <LordAro> and removing a load of casts
21:30:03 <TrueBrain> and changing some const in constexpr :P
21:30:05 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #8468: Fix #8316: Make sort industries by production with a cargo filter possible https://git.io/JLHah
21:30:19 <TrueBrain> so yeah, PR it, and we make michi_cc review it :P
21:30:48 <TrueBrain> Q, T or X for me :)
21:30:51 <Wolf01> frosch123, put in all the buttons with the same function and let players push the one they feel like it's the right one?
21:32:33 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNn2
21:32:43 <andythenorth> make it a setting!
21:32:52 <andythenorth> if only the sprites were in a grf somewhere :P
21:33:16 <TrueBrain> andythenorth: hard no on that last addition :P
21:33:35 <andythenorth> you can't even tell what it is?
21:34:04 <Timberwolf> I had that same problem with inward arrows, they either don't fit or look too light.
21:38:01 <Timberwolf> Heh snap, I tried the same. Can't decide if I like it or not :/
21:38:38 <DorpsGek> [OpenTTD/OpenTTD] mattkimber commented on pull request #8455: Unified GUI buttons for renaming and scrolling to entities https://git.io/JLNnA
21:40:12 <Wolf01> Make default UI 2x and draw better icons? XD
21:40:59 <frosch123> ottd will already use the extra zoom icons
21:41:12 <frosch123> ogfx has a few, but not all
21:41:20 <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #8502: Codechange: Remove min/max functions in favour of STL variants https://git.io/JLNcq
21:41:30 <LordAro> and apparently i already need to rebase it
21:41:57 <TrueBrain> LordAro: very tempted to say: if CI passes, blind approve, and go go go :P
21:42:03 <FLHerne> Without those holes in the top/bottom arrows it's a bit confusing, and with them it looks silly
21:44:22 <LordAro> TrueBrain: very brave :p
21:44:34 <LordAro> i at least want to check for cast warnings on all the platforms
21:45:30 <Samu> quick question: cargos or cargoes?
21:45:35 <TrueBrain> std::max({src->r, src->g, src->b}); <- wow, this works? Cool ..
21:45:38 <andythenorth> potato or potatoes?
21:46:05 <frosch123> TrueBrain: you can even skip the ( )
21:46:20 <frosch123> "std::max{src->r, src->g, src->b};" also works
21:46:45 <frosch123> no, i think it is considered uncool now
21:47:08 <frosch123> c++20 fixes stuff in that area, so better not touch it
21:48:06 <TrueBrain> I see LordAro did apply some ... "coding style" at some places :P
21:48:10 <andythenorth> Samu dictionary disagrees with me and says cargoes
21:48:20 <andythenorth> but in all my native English speaking life I would never have written that
21:48:32 <TrueBrain> OpenTTD has no more instances of cargos since a few weeks ago :P
21:49:25 <TrueBrain> 3 places LordAro split a long line with newlines :D OpenTTD has no max-width you know!! :P
21:49:31 <TrueBrain> I am fine with it, like .. #care :P
21:49:54 <andythenorth> mostly when I search for rules about 'cargo or cargoes' I get 'cookies'
21:49:58 <andythenorth> the internet is fucked eh
21:50:07 <andythenorth> most of it is useless, just these banners across every page
21:50:30 <andythenorth> there's some browser plugin that gets rid of them somewhere
21:51:42 <Wolf01> I once activated some banners on one site I was working on, and I needed some time to figure out why I couldn't see them
21:52:54 <LordAro> TrueBrain: you got me to split the long comments just yesterday!
21:53:03 <LordAro> but yes, i did decide some lines were *too* long :p
21:57:18 <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLNCt
21:58:41 <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLNCt
22:01:12 <frosch123> does that really check for "compatible"? i.e. can you depower an electric train?
22:01:40 <LordAro> i'll be honest, i didn't get as far as testing it
22:01:54 <LordAro> i tend to assume JGR's stuff works :p
22:02:03 <LordAro> but i will test it shortly
22:02:27 <andythenorth> I would have assumed Gwyd uses JGR
22:02:36 <andythenorth> all the active discord users seem to use JGR
22:04:08 <TrueBrain> it has a lot of good things
22:04:21 <TrueBrain> but while backporting this patchset today, I got reminded about the quality of some patches on the forum
22:04:47 <supermop_Home> andythenorth drawing real hotels is less fun
22:05:41 <andythenorth> supermop_Home quit and do game art :P
22:05:45 <andythenorth> jobs are over-rated?
22:06:50 <andythenorth> "Now think that OpenTTD is RCT1 and JGR patchpack is RCT2. "
22:07:18 *** rptr_ has quit IRC (Ping timeout: 480 seconds)
22:07:19 <andythenorth> TrueBrain what's your favourite forum patch? :)
22:07:34 <TrueBrain> no clue; never indexed it sufficiently to have an opinion
22:11:23 <supermop_Home> andythenorth not sure which industry has least healthy hours
22:12:34 <supermop_Home> just reading architecture theory criticism of Cyberpunk 2077 issues
22:12:38 <frosch123> hmm, in my memory preview was faster
22:12:53 <TrueBrain> faster to build, or to use?
22:13:10 <TrueBrain> slowest target we have atm :D
22:14:57 <supermop_Home> though reports in media of CDPR crunch as "they had to work 6 day weeks in the months leading up to release" fall on unsympathetic ears here
22:17:27 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLNWt
22:20:19 <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #8502: Codechange: Remove min/max functions in favour of STL variants https://git.io/JLNcq
22:21:48 <frosch123> never noticed "mouse_over_start_stop", is that new?
22:25:54 <TrueBrain> frosch123: from C++ perspective, any reason not to do the std::min/max?
22:25:55 *** jottyfan has quit IRC (Quit: jottyfan)
22:26:10 <TrueBrain> CI fails LordAro :(
22:26:29 <TrueBrain> no member named 'min' in namespace 'std'
22:26:52 <LordAro> i somewhat expected as such
22:27:11 <TrueBrain> src\os\windows\crashlog_win.cpp(222): error C3861: 'min': identifier not found
22:27:27 <TrueBrain> and MacOS fails too :)
22:27:57 <TrueBrain> funny how only linux passes
22:28:00 <TrueBrain> how does that happen?
22:28:11 <frosch123> TrueBrain: the c++ one is exactly the same as our template
22:28:18 <frosch123> no idea why we ever had a custom one
22:28:32 <frosch123> minu() is special, but LordAro took care of that
22:28:33 <TrueBrain> RichK, or what was his name?
22:28:45 <frosch123> no idea what magic is behind our min(int, int) overload
22:29:26 <frosch123> kudr is yapf and weird misc/*hpp classes. richk is tgp, airports, and first newairports disaster
22:29:41 <TrueBrain> one of the two just added a LOT of C++ magic templating bla in core/
22:29:46 <TrueBrain> was his own framework
22:30:16 <frosch123> most of it is still in misc/
22:30:18 <TrueBrain> believe nobody could really review it, and it was one of those: what-ever, now it is in there :P
22:30:23 <frosch123> no idea where core is from
22:30:55 <Samu> please someone re-review #8468, glx
22:31:04 <Samu> going to bed, cyas goodnight
22:31:05 <TrueBrain> ah, yes, core is skidd
22:31:17 *** Samu has quit IRC (Quit: Leaving)
22:31:46 <TrueBrain> KUDr was misc indeed
22:31:49 <TrueBrain> owh, the history :P
22:32:08 <TrueBrain> "header files with miscellaneous template classes (smart pointers, blob, array, hashtable, etc.) moved from src/yapf to src/misc as they can now be used anywhere."
22:33:01 <TrueBrain> it is where I checked out from understanding those parts of the code btw :)
22:33:10 <TrueBrain> I avoided it like the plague
22:33:34 <LordAro> frosch123: the only "potentially weird" thing is that our one is inline, whereas STL is not
22:34:26 <TrueBrain> it was the months of: if we do it ourself, it is better than std:: :D
22:34:40 <TrueBrain> at least nobody added boost
22:34:49 <frosch123> LordAro: compilers can figure out the "inline" themself
22:36:05 <frosch123> TrueBrain: can i add the http client from boost? :p
22:36:54 <andythenorth> I am hard -1 to doing the work
22:37:12 <andythenorth> if it turned up 100% working patch with docs, I would be like...shrug, ship it
22:37:42 <andythenorth> like nmlc, newgrf wiki...everything
22:37:48 <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLNCt
22:37:50 <LordAro> frosch123: usually, yes
22:37:59 <LordAro> though it's a bit trickier if the function is in a different library
22:38:22 <frosch123> yes, but that does not apply to header templates
22:38:41 <andythenorth> so is it bedtime?
22:39:34 <LordAro> frosch123: oh derp, true
22:40:33 <glx> min/max needs <algorithm> (I had that in grfcodec)
22:41:04 <LordAro> glx: yeah, i was hoping i could get away with it (like i can with gcc)
22:41:12 <LordAro> probably will need to add it to algorithm
22:41:28 <LordAro> probably will need to add <algorithm> to stdafx.h *
22:41:34 <frosch123> TrueBrain: i have no idea either. yesterday we compared it to town authority. it's some weird restriction, that players have to work around. however, there was a similar patch on the forums in 2005, which also disallowed water wells and other. this one exclused liquid
22:41:46 <frosch123> so, the industry selection is opinionated :)
22:42:04 <TrueBrain> so, if it is done, via NewGRFs, which requires spec and blablablabla, lot of work
22:42:59 <frosch123> newgrf also do not work here. industry authors should not care about player playstyle
22:43:18 <frosch123> and forking "firs with restrictions" from "firs without restrictions" is insane
22:43:34 <TrueBrain> right ... so it is even a deeper rabbit hole
22:44:01 <frosch123> if we had multiple gs, which can check player actions, that would be awesome, but is utopical :)
22:44:19 <TrueBrain> but something like that would be needed, indeed
22:44:26 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLNl1
22:44:36 <TrueBrain> I am close to some specs for something like that btw ... I just need to validate how Squirrel VMs work again :D
22:44:55 <TrueBrain> if creating multiple VMs is an option, or that it has to be 1 VM with multiple instances
22:45:00 <TrueBrain> (Squirrel instances, that is)
22:45:04 <TrueBrain> but that tomorrow :)
22:45:42 <frosch123> every ai has a separate VM, so it is an option. but you probably already knew that :p
22:46:00 <TrueBrain> yeah, but 1 VM per town, for example
22:46:03 <TrueBrain> not sure if that holds up :D
22:46:14 <TrueBrain> ~20k VMs on 4kx4k? :D
22:46:25 <frosch123> no, and i don't think you want that.
22:46:35 <frosch123> that puts gs into the same position as newgrf: no global state
22:46:35 <TrueBrain> I doubt that is efficient :)
22:47:01 <TrueBrain> global state is either way something we need to talk about .. as can there be state between 2 GSes
22:47:09 <frosch123> if gs wants per-town state, it can do that itself
22:47:22 <TrueBrain> but okay .. I will write something down first, makes talking a bit easier, I hope :D
22:47:35 * andythenorth thinks sleeping time
22:47:38 <frosch123> if you want state between 2 gs, then go for messagepassing. no shared memory :)
22:47:45 <orudge> So, from January - November last year, we had a surplus for the year of around £45. After TrueBrain's post in December, we then had £412 of donations that month! Taking out the new OVH servers, that led to a 'profit' of £337 last year :)
22:47:51 <frosch123> gs can send json to each other
22:47:53 *** andythenorth has quit IRC (Quit: andythenorth)
22:47:57 <orudge> Now, let's see what the first AWS bill looks like :|
22:48:02 <frosch123> there is some draft for json passing between gs and newgrf somewhere
22:48:14 <TrueBrain> orudge: haha :D Tnx for the update!!
22:48:31 <TrueBrain> frosch123: JSON parsing is nasty ... but yeah, it has to be an API of sorts
22:48:35 <TrueBrain> no shared memory no :)
22:48:52 <frosch123> we already have a json parser somewhere
22:49:08 <frosch123> gs <-> admin port probably
22:50:13 <orudge> TrueBrain: Bank balance has increased overall in the past 4 years by just £102, which is fine really, as there's still a fair pile of money sitting there from previous years!
22:50:40 <TrueBrain> okay, the bill more than halved indeed by having those 2 VPSes doing BaNaNaS
22:50:46 <TrueBrain> current estimate is ~230 dollar this month
22:51:14 <frosch123> we need weak $, strong £
22:51:19 <orudge> OK. Well, at that rate, we can fund AWS for a couple of years at any rate :)
22:51:39 <TrueBrain> funny, EC2 instances now are the main cost
22:51:42 <TrueBrain> I can get those down a bit
22:51:43 <orudge> well, not "at any rate", that's a confusing expression for me to use when I already said "at that rate" :D
22:51:57 <TrueBrain> owh, and Tax ended up on the first of Jan, which is a bit odd ..
22:52:21 <orudge> If funds look like they're running low at any point, then we can perhaps do a fundraiser, but I think we should be OK for a while yet
22:54:48 <TrueBrain> yeah, and I will fiddle a bit with AWS to see what services I can reduce cost of
22:55:00 <TrueBrain> as I now see for example we are over EBS quota .. that is most likely unintended :D
22:55:12 <TrueBrain> we did 100k IO operations in the last 5 days :D
22:55:39 <orudge> [22:50:57] <frosch123> we need weak $, strong £ <-- it's looking a bit healthier these days, almost back to where it was pre-Brexshit
22:55:47 <orudge> Not quite as nice as when it was £1=$2 though
22:56:12 <TrueBrain> 500k DNS requests in 5 days, lolz
22:56:18 <TrueBrain> people love those DNS queries
22:57:13 <TrueBrain> but yeah, let's see how this month goes :)
23:00:16 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
23:01:21 *** sla_ro|master has quit IRC ()
23:02:22 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8439: Feature: Make town bridge max length a function of its population https://git.io/JLN8D
23:05:40 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLN8h
23:05:43 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLFsM
23:06:07 *** DasPoseidon has quit IRC (Remote host closed the connection)
23:08:24 *** DasPoseidon has joined #openttd
23:09:26 <LordAro> hmm, i'm not sure my Ensure* functions are being called at all
23:09:33 <LordAro> my print statements aren't doing anything, anyway
23:10:53 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8495: Feature: Prevent tunnels from being built under mines https://git.io/JLN4t
23:12:20 <TrueBrain> for some reason I have a hard time seeing the first, the second does work fine :P
23:15:25 <milek7> with multiple track layers :P
23:15:44 <milek7> (so more real bridges and tunnels)
23:15:59 <TrueBrain> too bad I removed devs.openttd.org .. I always link a an mp4 I think SmatZ once made, when people say these things :)
23:16:05 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8505: Feature: Allow converting track under trains when compatible with the new rail type https://git.io/JLN4z
23:16:57 <TrueBrain> lol @ "highway" PR .. for most of the time reading it, I was like: wtf is "highway" .. but I guess a NewGRF :D
23:20:18 *** gelignite has quit IRC (Quit: Stay safe!)
23:33:10 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:49:44 <LordAro> TrueBrain: i use fprintf(stderr by default to avoid stdout weirdnesses :p
continue to next day ⏵
![https://i.imgur.com/vhRiqAc.png[/img]](https://i.imgur.com/vhRiqAc.png%5B/img%5D){kind=link}
