IRC logs for #openttd on OFTC at 2021-08-15
⏴ go to previous day
00:12:17 *** Progman has quit IRC (Remote host closed the connection)
02:20:31 *** Wormnest has quit IRC (Quit: Leaving)
02:40:29 *** roadt__ has quit IRC (Ping timeout: 480 seconds)
02:40:42 *** roadt__ has joined #openttd
02:56:34 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:54:54 *** Tirili has quit IRC (Quit: Leaving)
03:58:53 *** APTX has quit IRC (Quit: Farewell)
06:31:28 *** Progman has joined #openttd
07:17:19 *** andythenorth has joined #openttd
08:26:26 <Eddi|zuHause> i hope no train set... :p
08:42:19 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9346: Feature: [NewGRF] Maximum curve speed modifier for rail vehicles https://git.io/JRjjs
08:42:32 <andythenorth> TrueBrain do you have the image in a gimp or psd file with layers?
08:43:13 <andythenorth> is the screenshot tightly cropped, or could you move it down to centre the farm vertically?
08:43:31 <TrueBrain> the GUI is in relation to the farm :P
08:43:50 <TrueBrain> but that is just a matter of making a new screenshot
08:44:07 <TrueBrain> but putting the farm in the center also looked really weird :P
08:44:11 * andythenorth assumed separate layers
08:44:25 <TrueBrain> I just made a screenshot ingame :)
08:44:31 <andythenorth> I would move the GUI and the logo in so they have 40-70px of margin left / right
08:44:34 <andythenorth> otherwise it's done
08:44:46 <andythenorth> these images can be over-thought :)
08:45:02 <TrueBrain> and as long as nobody else picks it up, they will keep looking this quirky
08:45:03 <andythenorth> nobody really interrogates them, they're just "oh pixture"
08:45:11 <TrueBrain> someone on Discord tried to make a better one .. it had the same issue :P
08:45:36 <andythenorth> the content is fine, the elements aren't quite in proportion currently
08:46:54 <TrueBrain> it is super misaligned everywhere, but I have a hard time centering stuff in GIMP :D
08:47:13 <andythenorth> oh yeah the text isn't centered
08:47:18 <TrueBrain> I hate GIMP, basically :)
08:47:38 * andythenorth considers scripting a screenshot make
08:47:59 <andythenorth> at work we do all design in Figma now, it's 100% collaborative, so awesome
08:48:05 * andythenorth was sceptical, but totally converted
08:48:16 <andythenorth> watching design happen live is so much better
08:48:28 <andythenorth> figma probably kills our need for an office
08:49:06 <TrueBrain> I just want to tell GIMP: center this element
08:49:09 <TrueBrain> but I cannot find where :P
08:49:41 *** gelignite has joined #openttd
08:50:07 <TrueBrain> okay, the top text was centered, as it turns out
08:50:09 <TrueBrain> did not expect that :P
08:50:45 <TrueBrain> it was the bottom one that wasnt :P
08:53:09 <TrueBrain> btw, for #9346, doing floats like this is common for others part of the NewGRF spec?
08:54:50 <andythenorth> I spotted one missing . I commented
08:55:45 * andythenorth hovering over 'approve' button :P
08:56:52 <andythenorth> also floats, what now? :)
08:56:57 <andythenorth> integer maths innit
08:57:03 <TrueBrain> prop 2E, is basically a float
08:57:06 <TrueBrain> "decimal precision"
08:57:10 <TrueBrain> what-ever you want to nick it :)
08:58:36 <TrueBrain> it allows a value of 12.1313 :P
08:58:37 <andythenorth> "value 1.0 (0x0100) would double the vehicle's speed in curves"
08:58:46 <andythenorth> yeah this is .... unusual
08:58:57 <andythenorth> using 1.0 as double is odd
08:59:07 <TrueBrain> basically it is in units of 1/256th increase
08:59:07 <andythenorth> but might reflect the original spec, I didn't look
08:59:23 <andythenorth> fractional floats are not common afaik, but might make sense here
08:59:41 <TrueBrain> no, it seems it is common .. just the way of wording seems new
08:59:45 <TrueBrain> but lot of values are in 1/8th
08:59:54 <TrueBrain> which is just the same thing :P
09:00:27 <andythenorth> yeah that's a pattern
09:00:36 <Rubidium> prop 2E is far from a float ;)
09:00:39 <TrueBrain> here the precision is just higher, basically
09:00:48 <TrueBrain> Rubidium: it is not an int, so call it what-ever you want
09:00:50 <TrueBrain> I really don't care :)
09:01:06 <TrueBrain> sure, what-ever floats your boat :)
09:01:10 <andythenorth> it's fun that 15 bits strikes again
09:01:17 <andythenorth> so the prop can do double the range of the cb result
09:01:25 <TrueBrain> question stands, is that common or not :)
09:02:16 <andythenorth> so I could set the prop to 128, then cb result to 64 for [some condition], then I can't return to 128
09:02:24 <andythenorth> so your train can get slower in curves :P
09:03:02 <TrueBrain> I see similar stuff used for cargos, and for some other things
09:03:09 <TrueBrain> but it is the lesser used method, by the looks
09:03:15 <TrueBrain> at least not completely out of line with the rest
09:03:25 <andythenorth> I would like to see that PR included for $reasons
09:03:30 <andythenorth> but there are no usable docs
09:03:51 <andythenorth> (tell me if I'm being an ass) - 99% it will end up landing on me to write the docs otherwise
09:03:53 <TrueBrain> refit cost, uses 1/32th precision
09:04:07 <andythenorth> and I don't quite understand the feature, and the PR isn't enough for authors to succeed in using it
09:05:19 <TrueBrain> andythenorth: guess you didn't read my reply :P
09:05:40 <andythenorth> I read it without reading it :P
09:05:47 <andythenorth> my life would be better if I could stop doing that
09:06:03 <andythenorth> also if I could listen to my wife at certain critical points
09:06:40 <TrueBrain> reason 0x0100 doubles, is because it does max_speed += max_speed in that case
09:06:47 <TrueBrain> I have no clue if that makes sense or not
09:06:59 <andythenorth> it doesn't really matter as long as it works
09:07:01 <TrueBrain> so 0 means "no change"
09:07:02 <andythenorth> newgrf is full of odd
09:07:03 *** sla_ro|master has joined #openttd
09:07:14 <TrueBrain> yeah, I made the mistake to see if there is some consensus in these kind of things
09:07:29 <andythenorth> the consensus recently is the frosch consensus
09:07:55 <andythenorth> which is 'wherever possible use callbacks over properties' and 'make callback results sensible'
09:08:17 <andythenorth> all the frosch API extensions are very good, which is why we default to frosch, but that's not a great pattern :)
09:08:35 <andythenorth> was there an nml patch for that?
09:08:45 <TrueBrain> so there just aren't any specs for it, basically
09:09:28 <TrueBrain> okay, so the fraction part seems to be common, just not in this high of a resolution
09:09:31 <andythenorth> it's not easy to know what defines 'good' in newgrf spec, so people are a bit loathe to touch it
09:09:44 <andythenorth> and newgrf PRs tend to get 5 different opinions in reviews
09:09:49 <TrueBrain> the name is slightly misleading, as it is not a curve-speed-mod, but a max-speed-mod. Not sure that matters
09:10:35 <TrueBrain> (it is not a modification to the curve-speed, but to the max-speed)
09:10:49 <andythenorth> but only modifies in context of GetCurveSpeedLimit() no?
09:11:00 <TrueBrain> ah, yes, you are right
09:11:10 <andythenorth> stopped clock, twice a day
09:11:25 <andythenorth> newgrf reviewing needs to be a more fun game :P
09:11:40 <TrueBrain> so the variable there is just a bit odd duck :)
09:12:03 <andythenorth> no test case, no examples
09:12:15 <TrueBrain> so I guess the only "choice" here is to have it an addition, instead of a multiplier
09:14:08 <TrueBrain> I really have zero opinion on addition vs multiplier ..
09:14:58 * andythenorth doesn't understand the implementation enough
09:15:23 <andythenorth> it's hard to make the connection between -ve values and 'multiplier < 1' but I assume it's bitmaths stuff
09:15:39 <TrueBrain> you keep saying that, and I keep not understand wtf you mean :P
09:15:51 <andythenorth> "negative numbers decrease" in the PR description
09:16:06 <TrueBrain> but what does "-ve" mean?
09:16:26 <andythenorth> it's some UK engineering convention, not sure why
09:16:29 <TrueBrain> not in a million years ..
09:16:45 <TrueBrain> but tnx :) I really did not understand that :P
09:17:00 <andythenorth> oh it's medical I think, from doctors notes
09:17:07 <andythenorth> e.g. blood type O -ve
09:17:10 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #9346: Feature: [NewGRF] Maximum curve speed modifier for rail vehicles https://git.io/JZcqh
09:17:37 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9346: Feature: [NewGRF] Maximum curve speed modifier for rail vehicles https://git.io/J0eIQ
09:17:47 <TrueBrain> yeah, okay, I get why he went for addition over multiplication
09:17:56 <TrueBrain> basically, any positive value increases the curve speed
09:18:00 <TrueBrain> any negative value decreases it
09:18:21 <TrueBrain> michi_cc: awesome :D
09:18:41 <andythenorth> shall I put the yellow on it :)
09:19:18 <michi_cc> TrueBrain: Can I write version 12 for it?
09:19:36 <TrueBrain> will update changelog now
09:19:46 <TrueBrain> so a value of like 0xFF00 means the curve-speed becomes zero, I think
09:20:05 <TrueBrain> and 0xFE00 makes it -(curve-speed)
09:20:56 <andythenorth> this makes sense
09:21:05 <michi_cc> Well, almost. Code clamps the result to 2 internal units (to make sure trains don't get permanently stuck).
09:21:42 <TrueBrain> michi_cc: true .. so negative values are not all that useful?
09:22:06 <TrueBrain> the range is like -1 .. 63.(255/256th), I guess?
09:22:16 <TrueBrain> (effectively, that is)
09:23:08 <michi_cc> Negative does reduce curve speed, it's makes sure a train can never to completely stuck (just like the code that always allows 1 mph no matter the slope).
09:24:00 <TrueBrain> yeah, but the PR description says -128 to 127.996, but effectively it is just -1 to 127.996, if I understand it correctly
09:24:46 <TrueBrain> and I get what you mean andythenorth , via the prop you can set a wider range than via the callback :)
09:24:52 <TrueBrain> this 15bit "words" are funny :)
09:26:46 <michi_cc> It 8 fraction bits, so 0x8000 becomes -128 effectively.
09:27:32 <TrueBrain> yes, but it is clamped after that, not? So doing "maxspeed += (maxspeed * -128)" doesn't give an other final result than "maxspeed += (maxspeed * -1)"
09:27:40 <TrueBrain> (sorry, trying to understand NewGRF, and my head hurts a bit :P)
09:28:20 <TrueBrain> after -2, negative values are all the same
09:28:35 <TrueBrain> (before -2? What is the proper English there? :P)
09:29:11 <michi_cc> Ah, yes of course. The usable negative range are only the fractions from 0 to -1, but having a different fractional part for positive and negative would be even more confusing :)
09:29:26 <TrueBrain> yeah, I don't mind this way of implementing it at all :)
09:29:46 <TrueBrain> just might be worth noting in the specs :D
09:30:10 <TrueBrain> no, not -2, I am confused again .. it really is "just above -1" :P
09:30:21 <TrueBrain> right, chapter closed, continuing on :)
09:32:35 <andythenorth> TrueBrain how all newgrf PRS end :P 'moving on'
09:32:53 * andythenorth looking if GS could read grf params
09:32:57 <andythenorth> grf can read grf params
09:33:01 <andythenorth> human can read grf params
09:33:09 <andythenorth> why this apartheid against GS!
09:33:32 <TrueBrain> you just have been given the power to know GRFs exist, and now you want more?
09:33:37 <TrueBrain> you are just like Cookiemonster :)
09:33:53 <andythenorth> well it's Russian dolls :)
09:34:02 <andythenorth> now I know the grf, but I still have to do hacks to detect the params
09:34:33 <TrueBrain> so, do the reasonable thing, and make an issue to state why you want that :P
09:34:40 <andythenorth> reading the code first
09:34:41 <TrueBrain> wait, you should do that on forums by workflow, not?
09:35:27 <andythenorth> 'parameter' is a horrible search term for src
09:36:46 <andythenorth> it's Action D 0xFE but eh
09:37:39 <andythenorth> maybe this is the right place
09:37:42 <andythenorth> "/** Action 0x0D: Set parameter */"
09:39:25 <andythenorth> GS could also do that probably?
09:39:35 <andythenorth> I guess we'd have to make a method or something instead of copy-pasting
09:40:18 <michi_cc> Okay, NFO wiki done.
09:42:31 <michi_cc> And I just found a bug in the PR :) >> on signed numbers is implementation defined (even if most implementations do what you'd expect).
09:42:50 <TrueBrain> did not expect that
09:42:58 <michi_cc> I'll make a PR to replace it with / 256 :)
09:43:11 <TrueBrain> but we use that trick too to sign-extend the value earlier
09:43:17 <TrueBrain> (the 15-bit to 16-bit)
09:43:25 <michi_cc> "For negative a, the value of a >> b is implementation-defined (in most implementations, this performs arithmetic right shift, so that the result remains negative)."
09:43:41 <TrueBrain> don't we have many more places with this issue then? :D
09:44:17 <TrueBrain> many places in the code do ">>" with the assumption the value remains signed, if I take a quick look in newgrf.cpp :)
09:44:26 <michi_cc> Sane implementations probably do an arithmetic shift. No idea if there's even any popular platform that doesn't.
09:44:47 <TrueBrain> res = (int32)src1 >> -(int32)src2;
09:44:57 <TrueBrain> guess the intention there is the same :)
09:45:39 <TrueBrain> but also means GetEngineProperty should be addressed
09:45:59 <TrueBrain> these kind of quirks in the C++ specs are just annoying :P
09:46:12 <TrueBrain> case DSGA_OP_SAR: return (int32)(S)last_value >> ((U)value & 0x1F);
09:46:22 <_dp_> using signed types in C is like walking on a minefield
09:46:56 <TrueBrain> michi_cc: okay, it is not as bad as I assumed, those are the only cases I can find in the NewGRF code
09:47:33 <andythenorth> hmm what was I supposed to be doing?
09:47:40 <andythenorth> before all this reviewing DISTRACTION
09:47:50 <andythenorth> oh I was watching YouTube
09:49:07 <michi_cc> Those two are actually harder to change as you'd have to calculate the proper dividend first.
09:49:36 <TrueBrain> return ((int32)(S)last_value >> ((U)value & 0x1F)) & (last_value & 0x8000);
09:51:23 <TrueBrain> michi_cc: clang, gcc and MSVC do what we expect btw
09:51:54 <DorpsGek> [OpenTTD/OpenTTD] michicc opened pull request #9480: Fix 2183fd4d: [NewGRF] Use divide instead of right shift for signed numbers. https://git.io/J0es3
09:52:01 <michi_cc> You may deny the PR if you want :)
09:52:19 <TrueBrain> nah, it is not wrong :)
09:52:30 <TrueBrain> I was more checking in how much world of hurt we really are :P
09:52:39 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9480: Fix 2183fd4d: [NewGRF] Use divide instead of right shift for signed numbers. https://git.io/J0esR
09:53:22 <TrueBrain> trying random other compilers on godbolt to see if any breaks :P
09:53:56 <michi_cc> Other CPU targets is probably the more interesting case (like rest in peace PowerPC or so :)
09:54:27 <TrueBrain> I cannot imagine a CPU without a proper bitshift support :P
09:54:36 <TrueBrain> even the custom CPU I build in uni did the right thing :)
09:55:11 <TrueBrain> (writing microcode was fun .. I miss those simple days :P)
09:56:19 <TrueBrain> PowerPC with GCC 6.3 also does the right thing
09:57:18 <TrueBrain> icc too, even djgpp
09:57:32 <TrueBrain> they should just change the spec :P
09:58:24 <TrueBrain> now I do wonder one thing michi_cc ... why ever did you came up to check the specs for this? :D
09:58:26 <michi_cc> Well, C still assumes that somewhere out there sign-magnuted still lives on.
09:58:34 <_dp_> looking at #9480 I'm wondering where is / behavior defined
09:59:37 <michi_cc> TrueBrain: Literally a random thought like "Hmm, is >> actually an arithmetic or a logic operator?"
09:59:47 <_dp_> everything I see is mentioning c++11 _draft_
09:59:56 <TrueBrain> michi_cc: amazing how a mind sometimes can make such jumps :D
10:00:22 <_dp_> michi_cc, IO operator :p
10:00:41 <TrueBrain> and now we wait for the CI to finish validating #9480 ... boooorrrriiinnnggggg
10:03:10 <michi_cc> "For integral operands, it yields the algebraic quotient."
10:03:30 <michi_cc> "Note: Until C++11, if one or both operands to binary operator % were negative, the sign of the remainder was implementation-defined, as it depends on the rounding direction of integer division. The function std::div provided well-defined behavior in that case."
10:04:03 <michi_cc> TrueBrain: msys is just too slow :)
10:06:28 <TrueBrain> luckily enough, doesn't require changelog updates
10:11:11 <TrueBrain> I still cannot believe I was always bitching about MacOS :P
10:11:18 <TrueBrain> MSYS was like: hold my beer
10:14:57 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9480: Fix 2183fd4d: [NewGRF] Use divide instead of right shift for signed numbers. https://git.io/J0es3
10:25:02 * andythenorth can't approve 9477 :P
10:28:55 <TrueBrain> lets see how much stuff breaks
10:29:10 <TrueBrain> lets send that to staging already too
10:30:15 <TrueBrain> time to get lunch .. we will see how much of a boom it says in like 15 minutes :P
10:34:32 <_dp_> mp:sp being about 1:4 matches other data as well
10:35:43 * _dp_ no idea what all those people do in sp
10:36:32 <_dp_> fill map with objects :p
10:36:37 <TrueBrain> ugh .... guess nobody addressed a problem we knew we had? ugh .....
10:36:41 <TrueBrain> so yeah, this tag will fail to build
10:37:29 <TrueBrain> always a bit sad, that we know something isn't working, but nobody takes the time to fix it :D Before it says boom, ofc ;)
10:40:15 <TrueBrain> at least lets see if it is the only target that no longer works :)
10:46:40 <TrueBrain> ugh, fixing this cleanly is annoying
10:46:44 <TrueBrain> can I just drop the target? :P
10:46:51 <TrueBrain> LordAro would get mad at me :P
10:50:24 <TrueBrain> cmake doesn't want to see gcc-8
10:54:52 <TrueBrain> guess I just delete the tag again for now? :)
10:56:48 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9481: Fix: [Actions] Ubuntu 18.04 by default has a compiler too old for OpenTTD https://git.io/J0eEY
10:59:46 <TrueBrain> okay, my testing shows that should address the issue, but you can never be really sure till you run it :)
11:00:50 <TrueBrain> owh, and I should run eints
11:01:10 <TrueBrain> lot of strings to remove :D
11:01:25 <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #9481: Fix: [Actions] Ubuntu 18.04 by default has a compiler too old for OpenTTD https://git.io/J0euC
11:06:23 <DorpsGek> - Update: Translations from eints (by translators)
11:09:01 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9481: Fix: [Actions] Ubuntu 18.04 by default has a compiler too old for OpenTTD https://git.io/J0eEY
11:09:03 <TrueBrain> right, lets try this again shall we
11:11:44 <TrueBrain> out of workers, lol
11:11:50 <TrueBrain> owh well, will be another 30 minutes :D
11:12:16 <TrueBrain> at least I updated the Steam translations in the meantime :)
11:26:10 *** Hazzard has joined #openttd
11:26:47 *** Hirundo has joined #openttd
11:28:17 *** planetmaker has joined #openttd
11:28:17 *** ChanServ sets mode: +o planetmaker
11:28:32 *** ^Spike^ has joined #openttd
11:28:47 *** Terkhen has joined #openttd
11:28:47 *** ChanServ sets mode: +o Terkhen
11:29:17 *** XeryusTC has joined #openttd
11:29:17 *** V453000 has joined #openttd
11:35:39 <TrueBrain> universal MacOS builds ....... sssslllllloooooowwwwwwwww (and yes, I am impatient :P)
11:41:39 *** jottyfan has joined #openttd
11:47:17 <TrueBrain> over 30 minutes ... boy owh boy ..
11:47:33 <TrueBrain> universal binary was just a terrible idea :P
11:48:01 <andythenorth> end user experience was ok :P
11:49:30 <TrueBrain> so much wasted bandwidth :P
11:49:49 <TrueBrain> think about the CO2!
11:52:15 <TrueBrain> "Waiting for results from Apple. This can take minutes to hours."
11:52:19 <TrueBrain> I love that that is not a lie :)
11:53:24 <TrueBrain> right, time to see if there is any other part of our infra that expected major.minor.patch :D
12:02:15 <TrueBrain> guess I found one issue :D
12:05:53 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9482: Fix: [Actions] changelog assumed major.minor.patch versioning https://git.io/J0eXs
12:06:01 <TrueBrain> lets manually fix it for now :)
12:08:49 *** WormnestAndroid has quit IRC (Remote host closed the connection)
12:09:07 <TrueBrain> @topic set 1 "1.11.2 / 12.0-beta1"
12:09:07 *** DorpsGek changes topic to "1.11.2 / 12.0-beta1 | Website: *.openttd.org (source: github, translator: translator, server list: servers, wiki: wiki) | Don't ask to ask, just ask | 'Latest' is not a valid version, 'Most recent' neither | English only"
12:10:02 *** WormnestAndroid has joined #openttd
12:10:13 <TrueBrain> I think I did everything
12:12:34 <TrueBrain> checklist says: yes
12:13:27 <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #9482: Fix: [Actions] changelog assumed major.minor.patch versioning https://git.io/J0e1K
12:14:17 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9482: Fix: [Actions] changelog assumed major.minor.patch versioning https://git.io/J0eXs
12:16:14 *** ChanServ sets mode: +v tokai
12:16:25 <TrueBrain> and V453000 , time for that NUTS release? :P
12:19:05 <TrueBrain> the NML PR needs examples :)
12:20:28 <glx> nml needs a release too :)
12:24:19 <TrueBrain> right .. lets see if we get some people testing :D
12:25:00 <andythenorth> nml changelog, needs updating though
12:28:37 <TrueBrain> hmm, can't find any issues with 12.0 naming .. seems everything just said: sure, okay :) (code-wise)
12:29:08 <andythenorth> it's so exciting though
12:29:10 <andythenorth> having a bigger number
12:29:13 <andythenorth> this one goes to 11!
12:30:01 <andythenorth> 13 is unlucky though :P
12:31:43 * andythenorth reading nml commits
12:33:03 <andythenorth> except my birthday is 13
12:33:19 <TrueBrain> you skip that too? :P
12:34:09 <_dp_> this commment seems wrong
12:50:05 <TrueBrain> 42 STUN (IPv4) connections, and 1 fallback to TURN
12:51:47 <glx> _dp_: it was mostly correct 10 years ago, persistant storage was an industry only thing
12:53:41 <glx> ah no, airports were concerned too
12:59:03 <glx> "1.11.1 is followed up by 12.0" <-- and nobody noticed the "lie" (1.11.1 is followed by 1.11.2 ;) )
13:00:31 <glx> it's part of optimisations
13:02:53 <glx> but 1.11.1 then 12.0 looks so much better
13:03:45 <TrueBrain> so we leave it :P And we say we did it on purpose :D
13:06:36 <andythenorth> glx FLHerne ^ I have to go out for a bit soon, but that's the ones I thought were relevant
13:14:48 <glx> hmm server listing doesn't seem to work
13:15:01 <glx> there should be 2, I see none
13:15:10 <TrueBrain> E_MESSAGE_TOO_GENERIC :)
13:15:33 *** jottyfan has quit IRC (Quit: jottyfan)
13:15:46 <glx> I started the beta in steam, open multiplayer window, pressed search internet
13:15:53 <glx> all I get is an empty list
13:16:12 <TrueBrain> that is really unusual
13:16:21 <TrueBrain> as it also shows pre 12.0-beta1 servers etc
13:16:30 <TrueBrain> start with -dnet=6 and see where it goes wrong?
13:16:34 <glx> oh maybe my dual stack is playing again
13:16:49 <_dp_> hm, I see a lot of timeout messsages on a server...
13:16:55 <TrueBrain> as long as either IPv4 or IPv6 work, it should work fine
13:17:42 <_dp_> and mp window looks fine for me
13:18:37 <_dp_> and if I connect to a server it times out on downloading the map
13:19:06 <_dp_> though may be a server issue
13:19:43 <TrueBrain> glx: or did you still have entries in your hosts file? :D
13:20:36 <TrueBrain> for me it works for both IPv4 and IPv6 .. so I am really curious what your debug log says :)
13:22:08 <glx> OTTD_COORDINATOR_CS coordinator.openttd.org:4976 <-- this can be the issue
13:23:30 <TrueBrain> downside of testing :P
13:23:49 <glx> I though I only set it locally
13:25:22 <glx> ok restarting openttd doesn't change anything, let's also restart steam then
13:29:49 <_dp_> hm, can anyone connect to my server (one with numbers)?
13:29:58 <_dp_> all I ssee is timeout messsages and no connections
13:30:42 <_dp_> oh, yeah, lol, I guess it's toyland xD
13:31:00 <_dp_> just started some random settings I had left from testing
13:31:41 <glx> youhou 2 newgrf added to my long list
13:32:12 <TrueBrain> glx is collecting them all! :P
13:32:43 <glx> any server I test has newgrf I don't have
13:32:51 <andythenorth> I could make some special rare grf editions
13:33:13 <TrueBrain> glx: ghehe, I know what you mean :D
13:33:25 <TrueBrain> but I just wipe my folder from time to time :D
13:35:31 <TrueBrain> omg, the CPU of the Game Coordinator is spiking!
13:35:37 <TrueBrain> it went from 0.1% to 0.5%
13:36:20 <TrueBrain> well, that is to say: not since 12.0-beta1
13:36:27 <peter1138> #9483 "Try this" "I've done this other thing"
13:36:29 <TrueBrain> there still is some memleak I think, as it has slowly been climbing
13:36:50 <TrueBrain> it started at 48% memory usage 4 weeks ago
13:37:03 <glx> should we tell gpu is not used in 1.10 ?
13:37:04 <TrueBrain> so I guess it crashes in ... 12 weeks or so
13:37:08 <TrueBrain> we redeploy every month anyway :)
13:37:25 <TrueBrain> glx: go for it .. pretty sure it is an unfixible problem anyway :)
13:39:19 <TrueBrain> no clue if the OpenGL on a Pi4 is recent enough
13:39:55 <glx> the good old saw teeth graph :)
13:40:10 <TrueBrain> finding this memleak will be tricky :D
13:40:22 <TrueBrain> also not sure I really care, honestly
13:42:44 <_dp_> hm... if you update GC will it disconnect all routed clients?
13:47:29 *** virtualrandomnumber has joined #openttd
13:47:45 *** virtualrandomnumber has quit IRC ()
13:52:09 *** andythenorth has quit IRC (Quit: andythenorth)
15:00:18 *** andythenorth has joined #openttd
15:01:46 *** sla_ro|master has quit IRC ()
15:32:37 <Eddi|zuHause> "add variable for 'town_index' (var 0x41)" <-- needs specifying the feature the var applies to
15:36:57 <andythenorth> do we have a small font for grf? :P
15:37:06 * andythenorth needs some hint text for industry window
15:39:07 *** jottyfan has joined #openttd
15:43:20 <michi_cc> Do we need to plan a multiplayer livestream now?
15:43:50 <LordAro> i feel like i've missed some things
15:44:30 <michi_cc> A release with multiplay-just-works (or so they say :)
15:47:37 <TrueBrain> Haha, not a bad idea :p
15:48:54 <andythenorth> NoCarGoal 7 years
15:49:02 * andythenorth probably can't right now
15:50:14 <andythenorth> glx FLHerne ^^ :)
15:55:01 <andythenorth> how about a new name?
16:05:12 *** Wormnest has joined #openttd
16:06:42 <andythenorth> but now I can't approve that
16:06:47 <andythenorth> I could have approved frosch's
16:06:52 <andythenorth> but frosch's is out of date
16:07:22 <spnda> I might PR my updated visual studio code syntax generator then
16:09:15 *** qwebirc80692 has joined #openttd
16:09:35 *** qwebirc80692 has quit IRC ()
16:10:22 <andythenorth> oh that's merged :)
16:10:49 <andythenorth> wait why is that not in the commit history
16:11:06 <andythenorth> oh it's the OpenTTD commit that's merged, Github is confusing
16:11:21 <michi_cc> No idea if NML property definition in this PR is okay for signed (negative) numbers.
16:12:03 <andythenorth> other unit_conversion values are 255
16:12:45 <spnda> (no real need to review this one, as it's mainly just changes of the syntax)
16:13:05 <michi_cc> 255 should be okay, i.e. 1.0 should really be 256 in the GRF. I just have no clue if the NML property handling works for negative numbers.
16:13:24 <michi_cc> Damn, **256** of course
16:16:42 <andythenorth> I would like to see a test case
16:16:46 <andythenorth> but eh, bureacracy :P
16:17:18 <michi_cc> TrueBrain: I'm free the whole week, but then I firmly fall on the "SP without competitors" side of the reddit poll :P
16:17:48 <TrueBrain> I also don't really have time the next few weeks .. :(
16:36:28 *** Smedles has joined #openttd
17:04:25 <Samu> my ai was trying to build heliport to heliport routes
17:40:12 <spnda> can we pls get #233 merged?
17:42:03 * andythenorth reads the diff for eval()
17:42:20 <andythenorth> does *anyone* have any idea on this one, or shall I just yolo it?
17:42:51 <andythenorth> spnda do you want to un-draft it?
17:43:43 <andythenorth> there's no coin miner, so if checks pass I'm approving it
17:45:16 *** jottyfan has quit IRC (Quit: jottyfan)
17:50:17 <spnda> 8m 25s codeql that's so excessive lmao
17:52:52 <spnda> i don't want to push, but can we pls get it merged?
17:53:03 <spnda> i'm just waiting on it right now before I can finish writing this action
17:53:19 <FLHerne> I don't know the VS syntax, but it looks reasonable to me
17:53:48 <glx> hmm looks like I need to fix some actions
17:53:56 <spnda> i made the original and this one and I already used that new syntax locally for the newest versions of my extension
17:54:25 <spnda> I just decided to put it up so I can write a action that runs every few months and automatically updates it from that script
18:01:22 <andythenorth> checks passed, I'll merge it
18:01:41 <andythenorth> hmm, I might have relinquished merge rights
18:01:53 <andythenorth> yeah I have approve but not merge
18:03:01 <andythenorth> FLHerne have you got merge rights?
18:15:54 <andythenorth> shall we call it for 0.6.0 then?
18:24:38 *** tokai|noir has joined #openttd
18:24:38 *** ChanServ sets mode: +v tokai|noir
18:31:29 *** tokai has quit IRC (Ping timeout: 480 seconds)
18:33:58 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:34:16 <spnda> Also might it be possible to integrate a linter into NMLC? Say some flag and then it knows not to output any GRF or NFO code and simply look for syntax errors or unknown symbols.
18:43:03 <michi_cc> andythenorth: If I modify the train example, the resulting NFO seems to be good, i.e. -0.5 is correctly translated to 2E \wxFF81
18:43:37 <DorpsGek> [OpenTTD/nml] michicc commented on pull request #222: Feature: Support for rail vehicle property 0x2E, curve_speed_mod https://git.io/J0ffs
18:48:57 <DorpsGek> - Update: Translations from eints (by translators)
18:51:53 <DorpsGek> [OpenTTD/nml] andythenorth approved pull request #222: Feature: Support for rail vehicle property 0x2E, curve_speed_mod https://git.io/J0fJq
18:54:29 <andythenorth> someone can merge pls?
18:55:12 <spnda> Imo you can leave "- Update: VS generation script syntax (#233)" out of the changelog if you want to
18:55:22 <spnda> It's not really relevant so :shrug:
18:58:06 <FLHerne> spnda: Does it fix something?
18:58:17 <FLHerne> Or just newer and better?
18:58:46 <spnda> Basically just adds better comment support and better support for vscode themes
18:59:07 <spnda> not comments, numeric and strings*
19:01:43 <DorpsGek> [OpenTTD/nml] FLHerne approved pull request #222: Feature: Support for rail vehicle property 0x2E, curve_speed_mod https://git.io/J0fUa
19:01:54 <DorpsGek> [OpenTTD/nml] FLHerne merged pull request #222: Feature: Support for rail vehicle property 0x2E, curve_speed_mod https://git.io/JZySd
19:02:35 <FLHerne> Are there any breaking changes that are worth making that wouldn't be allowed in 0.6.1 ?
19:03:09 <FLHerne> I vaguely wanted to deprecate random_switch because it's hardly ever the right idea
19:03:16 <FLHerne> but grf authors might complain
19:03:27 <spnda> Well random switches are quite essential, no?
19:03:53 <FLHerne> Someone wanted to deprecate only independent random_switches which might be more realistic
19:04:49 <FLHerne> but this is too late for that proposal so nvm
19:08:01 <andythenorth> there can be more releases
19:08:04 <FLHerne> spnda: Using a normal switch on getbits(random_bits, blah) is less prone to nmlc saying "no" for no reason
19:08:06 <andythenorth> releasing is not expensive
19:08:44 <FLHerne> and sometimes you can combine it with a normal switch instead of chaining normal + random_
19:08:51 <andythenorth> we have a lot of stuff stored up that only I'm using probably
19:09:02 <andythenorth> most people probably don't build their own nmlc
19:09:18 <spnda> do language files have commas?
19:13:54 *** jottyfan has joined #openttd
19:15:45 <DorpsGek> [OpenTTD/nml] glx22 opened pull request #234: Fix: [Actions] CodeQL can now run on the merge commit https://git.io/J0fkM
19:23:30 <DorpsGek> [OpenTTD/nml] LordAro approved pull request #234: Fix: [Actions] CodeQL can now run on the merge commit https://git.io/J0fIM
19:25:26 <DorpsGek> [OpenTTD/nml] glx22 merged pull request #234: Fix: [Actions] CodeQL can now run on the merge commit https://git.io/J0fkM
19:29:29 <DorpsGek> [OpenTTD/nml] LordAro commented on pull request #231: Change: add string pluralisation examples to example industry https://git.io/J0fLu
19:49:44 <andythenorth> 231 might get closed
19:49:57 <andythenorth> maybe just paste it into wiki as example
19:52:30 <glx> no, 231 is a good idea, it tests strings to ensure nmlc is not broken
19:52:42 <DorpsGek> [OpenTTD/nml] andythenorth commented on pull request #231: Change: add string pluralisation examples to example industry https://git.io/J0fmk
19:54:09 <glx> because I don't think any of regression checks use plural
19:54:22 <LordAro> just make sure the nfo changes are actually correct :p
19:54:46 <glx> hehe, from the fail log it seems ok
19:56:44 <andythenorth> ok changelog done? :)
19:56:54 <andythenorth> there's always something else we _could_ include
19:57:03 <andythenorth> "real artists ship" :)
19:57:22 <LordAro> andythenorth: "- Add: Vehicle variable 'tile_has_catenary'.' (#183)"
19:58:02 <andythenorth> oh that's odd isn't
19:58:12 <LordAro> ideally the curve_speed_mod line would include the PR number as well
19:58:22 <andythenorth> weird typo I did
19:58:32 <glx> #230 ideally should be merged before release too
19:58:56 <DorpsGek> [OpenTTD/nml] LordAro approved pull request #230: Add: Options to disable optimisation and/or deprecation warnings https://git.io/J0fmd
19:59:21 <andythenorth> oh yeah, those warnings :)
19:59:34 <DorpsGek> [OpenTTD/nml] glx22 merged pull request #230: Add: Options to disable optimisation and/or deprecation warnings https://git.io/JRcOc
20:00:23 <glx> (many new "annoying" warnings since 0.5.3)
20:01:28 <andythenorth> add 230 note to changelog
20:02:13 <andythenorth> glx should tag :)
20:02:26 <andythenorth> as a major contributor to this one :)
20:02:59 <LordAro> it will close automatically when 232 is merged
20:05:58 <andythenorth> slow checks are slow :)
20:06:19 <glx> yeah codeQL takes time, but it's very useful
20:10:25 <glx> I guess andythenorth can't merge
20:10:43 <andythenorth> I gave up my rights, didn't like the responsibility :)
20:13:15 <glx> hoping the magic still works :)
20:13:25 <glx> 0.5.3 was a long time ago
20:26:50 <andythenorth> does that automatically publish to pip?
20:26:50 <glx> funny how release workflows finished way before #232 merge workflow
20:30:07 <DorpsGek> [OpenTTD/nml] glx22 commented on issue #229: Not enough random bits for the town name generation (40 needed, 32 available) https://git.io/JRUIc
20:30:10 <DorpsGek> [OpenTTD/nml] glx22 closed issue #229: Not enough random bits for the town name generation (40 needed, 32 available) https://git.io/JRUIc
20:34:27 *** gelignite has quit IRC (Quit: Stay safe!)
20:42:06 *** Gustavo6046 has quit IRC (Remote host closed the connection)
20:42:16 *** Gustavo6046 has joined #openttd
21:07:52 <andythenorth> should I forum post then :P
21:26:51 <DorpsGek> [OpenTTD/nml] andythenorth commented on issue #105: Document procedures for switches and random switches https://git.io/Jftrs
21:32:18 *** Samu has quit IRC (Read error: Connection reset by peer)
21:33:43 *** Progman has quit IRC (Remote host closed the connection)
21:38:02 *** andythenorth has quit IRC (Quit: andythenorth)
21:39:24 <_dp_> sometimes it takes years of C++ experience to know when to look for a missing brace
21:40:08 <_dp_> freaking compiler just complained about some incomplete type 1000 lines later...
21:41:19 *** sla_ro|master has joined #openttd
21:49:25 *** sla_ro|master has quit IRC (Ping timeout: 480 seconds)
21:52:19 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:00:38 *** spnda has quit IRC (Remote host closed the connection)
22:13:56 <glx> at least when it happens multiple times you know it may be in a header :)
22:18:39 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:34:45 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:46:21 <TrueBrain> Daily stats: 45 TURN and 226 STUN connections
22:46:51 <TrueBrain> seems to agree with the playtest we did in a small group :)
22:57:47 <_dp_> and no public 12.0 servers :(
23:06:46 *** WormnestAndroid has quit IRC (Remote host closed the connection)
23:06:59 *** WormnestAndroid has joined #openttd
23:17:57 <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #9484: [Bug]: Currency setting desync between ISO code and `locale_currencies` setting label https://git.io/J0fgV
23:21:03 <glx> and probably from a long time ago
23:48:30 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
23:49:53 *** WormnestAndroid has joined #openttd
23:57:55 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
