Config
Log for #openttd on 6th May 2019:
Times are UTC Toggle Colours
00:08:25  *** chomwitt has quit IRC
00:18:18  *** Fuco has quit IRC
00:35:37  *** Samu has quit IRC
00:51:13  *** Xaroth has quit IRC
00:59:45  *** glx has quit IRC
01:32:34  *** HerzogDeXtEr has quit IRC
02:01:34  *** Flygon has joined #openttd
02:08:01  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjn9A
02:16:06  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHf
02:21:02  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHU
02:23:11  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHk
02:30:51  *** debdog has joined #openttd
02:34:13  *** D-HUND has quit IRC
02:39:45  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHq
03:09:34  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
03:10:31  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHc
03:12:30  *** Supercheese has quit IRC
03:12:51  *** Supercheese has joined #openttd
03:14:46  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHC
03:27:11  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
03:31:50  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHu
03:37:33  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
03:49:49  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
04:54:51  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHM
05:00:38  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnH5
05:01:21  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHF
05:02:52  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnHN
05:06:22  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
05:35:22  <DorpsGek_II> [OpenTTD/OpenTTD] YJSoft commented on pull request #7572: GDI engine for font glyph rendering as a replacement for FreeType (take 2) https://git.io/fjnQe
05:50:09  *** andythenorth has joined #openttd
05:52:11  <peter1138> Is it lunch time yet?
05:54:07  <andythenorth> soon
05:54:19  <andythenorth> no frosch yet?
05:54:21  <andythenorth> bit early
06:06:06  <peter1138> Ah well, I have a cup of tea instead then.
06:08:46  <andythenorth> I have not had coffee yet
06:12:44  <andythenorth> 1 or 2? https://dev.openttdcoop.org/attachments/download/9433/curtains_for_you_1.png  https://dev.openttdcoop.org/attachments/download/9434/curtains_for_you_2.png
06:13:14  <andythenorth> it doesn't matter, but IRL is https://paulbartlett.zenfolio.com/img/s/v-3/p1399971904-3.jpg
06:15:12  <dwfreed> it would be a better comparison if the screenshots had the same bounding box :P
06:15:36  <dwfreed> I kind of like 2, simply because it's more noticeable
06:15:48  <dwfreed> 1 you really only notice at that high resolution
06:16:19  <peter1138> andythenorth, in between.
06:16:38  <peter1138> andythenorth, curtains in the edges, but not 3 pixels of them.
06:17:15  <peter1138> You have 3 pixels in the middle so 1.5 pixels on the edges makes sense.
06:17:54  <andythenorth> so it's quite common with symmetrical trains to get sprites that look like Space Invader enemies
06:17:59  <andythenorth> and 1 has that effect
06:18:29  * andythenorth proceeds to draw 1.5 pixels
06:19:19  <peter1138> If you were drawing at 2x, it would be 6 and 3...
06:19:33  <andythenorth> I shall remember that
06:26:59  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnQn
06:31:10  * andythenorth times optimised nml compiles
06:33:46  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7572: GDI engine for font glyph rendering as a replacement for FreeType (take 2) https://git.io/fjnQC
06:36:09  <andythenorth> if I spend about a day on it, I can probably get 10-15s off each compile run
06:36:25  <andythenorth> they're about 60-65s currently, depending on cache state
06:37:05  * andythenorth coffee
06:37:13  <peter1138> Is 60-65s faster than it was?
06:37:17  <peter1138> (Did something change)
06:38:13  <andythenorth> it's about the same for Horse
06:38:27  <andythenorth> bit slower than last week, I added a few thousand more sprites
06:39:54  <peter1138> Nice.
06:40:20  <andythenorth> removing all the cargo sprites knocks about 25s off
06:41:14  <andythenorth> the cargo sprites are comped into the spritesheets using python currently
06:41:32  <andythenorth> I could move them to layers
06:41:59  <andythenorth> then figure out the reduced set of unique cargo sprites needed to cover all wagon types and lengths
06:42:21  <andythenorth> then add another layer for the mask to draw bits of vehicle back over the cargo
06:43:30  <andythenorth> @calc 72 * 6
06:43:30  <DorpsGek> andythenorth: 432
06:43:33  <peter1138> Layers?
06:44:01  <andythenorth> https://newgrf-specs.tt-wiki.net/wiki/Action2/Vehicles#Composing_vehicles_from_multiple_sprites
06:44:11  <peter1138> Don't defer combining sprites to OpenTTD.
06:44:37  <peter1138> You're asking OpenTTD to do more work, all the time, to save a few seconds of compile-time.
06:44:52  <andythenorth> yes
06:44:55  <peter1138> Never optimize compile-time over runtime-time.
06:45:13  <andythenorth> but but :P
06:45:20  <andythenorth> this is exactly what layers were added to do :P
06:45:33  <andythenorth> it's not that I disagree btw
06:45:43  <peter1138> I don't know why they were added. I'm sure there was a specific reason.
06:46:02  <andythenorth> because it avoids people like me doing python based composition
06:46:05  <peter1138> Same with V453000 using callbacks all the time instead of static properties.
06:46:52  <andythenorth> so errr....
06:47:11  <peter1138> Although that's less of a issue.
06:47:11  <andythenorth> Horse would be much faster to compile if I split it into separate grfs
06:47:23  <peter1138> Just get a faster CPU :D
06:47:40  <andythenorth> I have seriously considered buying a PC tower for compiling
06:47:41  <peter1138> Didn't frosch say something about a parallelizing build at some point?
06:47:54  <peter1138> Wait, you're using a laptop? o_O
06:48:04  <andythenorth> yes
06:48:16  <andythenorth> get a PC tower, put it in the loft, compile over ethernet
06:48:32  <andythenorth> network IO might eat any speed gain though
06:48:49  <andythenorth> how fast do CPUs go single-threaded now?
06:50:19  * andythenorth can't understand the Dell site
06:51:06  <andythenorth> meh high core count, low clock, not what I want
06:54:04  <peter1138> Rewrite nmlc in C++ :p
06:55:38  * andythenorth looking for i9-9900 prices
07:00:28  <andythenorth> £1700 for an over-clocked i9 tower
07:01:43  <andythenorth> about 25% faster than my CPU
07:04:02  <peter1138> You reckon?
07:04:08  <peter1138> What's your CPU now?
07:04:37  <andythenorth> i7 8559U
07:05:00  <andythenorth> geekbench has it at 5140 single threaded
07:05:20  <andythenorth> i9 9900 is averaging 6216
07:07:40  <andythenorth> 18% faster
07:08:54  <peter1138> Sounds wrong.
07:09:33  <peter1138> The base clock is 33% higher to start with.
07:09:59  <peter1138> + more cache.
07:10:09  <peter1138> + better thermal properties, being not-a-laptop.
07:10:25  <peter1138> But still, £1700 to make Horse compile a few seconds faster is a bit much :D
07:10:29  <peter1138> Crowd fund it!
07:11:01  <LordAro> i think it would be more worthwhile to speed £1700 on speeding up nml
07:11:30  <andythenorth> I had another idea about detecting when only the sprites have changed
07:11:59  <andythenorth> then switching to grfcodec
07:12:14  <andythenorth> nml can output nfo, and often that would be unchanged
07:12:49  <LordAro> not sure how feasible that is with makefiles
07:13:08  * andythenorth suspects there's a way
07:13:14  <andythenorth> grfcodec compile is 3s
07:14:40  <andythenorth> @calc 3 * 0.82
07:14:40  <DorpsGek> andythenorth: 2.46
07:14:53  <andythenorth> would be 2.46s if I put that i9 in the loft
07:15:33  <peter1138> Yeah, hot dusty lofts are the best places for computers.
07:16:41  <andythenorth> I could liquid cool it, and use it to keep my hot tub warm?
07:16:58  <andythenorth> big heat sink :P
07:28:44  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7380: Feature: Multi-tile docks. https://git.io/fjv8j
07:36:18  *** Supercheese has quit IRC
07:52:04  *** Fuco has joined #openttd
09:07:20  *** chomwitt has joined #openttd
09:18:37  <_dp_> i9 is kinda meh
09:18:56  <_dp_> sure, it's fast single thread but amd totally tramples it on cores
09:19:02  <_dp_> and new ryzen is coming
09:19:11  <V453000> the single threaded performance is amazing
09:19:29  <V453000> great for Factorio & openttd computer torture
09:19:37  <V453000> for other things yeah :D
09:19:49  <V453000> but idk, a lot of the tasks I do still rely heavily on single threaded stuff
09:20:51  <V453000> like for example rendering in Blender - it's cute that all CPU benchmarks are on multi threaded when you render from the CPU, but you generally never want to do that, you want to render on the GPU. But the preparations before the data is sent to the GPU is from a large part single-threaded. :) And surprisingly that amount of time to synchronize is not insignificant
09:21:17  <V453000> but yeah the new AMDs are incredible as well in their own
09:21:25  *** yegjog[m] has left #openttd
09:22:06  <V453000> btw _dp_ I saw your zombies :D looked interesting but your humanity was quite wiped out
09:25:31  <peter1138> _dp_, laughable. Intel trounce AMD. More cores is great for servers, but don't make up for low single-core performance.
09:25:46  <peter1138> (i.e. gaming)
09:26:22  <_dp_> V453000, well, it restarts when they destroy everything
09:26:52  <_dp_> V453000, if you start early there you have some time to build up
09:27:19  <_dp_> V453000, but it's just an experiment at this point, not a complete gameplay
09:27:21  <reldred> I dunno peter1138, some of the new vidja james are getting a lot better at smashing cores
09:27:23  <V453000> yeah that's what I thought
09:27:45  <V453000> well generally buying a big good CPU for games is kind of pointless,no ? Just dump it all into a GPU and go nuts
09:27:47  <LordAro> some of them, not all
09:27:48  <dwfreed> I mean, look at factorio
09:27:50  <V453000> for "general games"
09:27:56  <V453000> factorio is quite s pecific
09:28:04  <reldred> Yeah depends on the games
09:28:06  <LordAro> quite a lot of what games do isn't very parallelisable
09:28:12  <dwfreed> yeah, but more games are going to multithreading
09:28:40  <_dp_> peter1138, openttd is the only st thing I use :p
09:28:41  <LordAro> i'm thinking of the physics simulation for KSP
09:28:47  <reldred> I've started to find I'm getting CPU constrained in a lot of games now. Didn't always used to be that way, but I am still on AM3.
09:28:49  <LordAro> there's nothing that can be done about that
09:28:50  <V453000> for example I'm getting +100% (!) performance on factorio between 7820X with 3000MHz CL15 vs 9900K 4266MHz CL17
09:29:12  <LordAro> V453000: that is surprising
09:29:12  <_dp_> peter1138, but now I can have stitton of browsers open, up to 4 openttd sometimes and not care at all
09:29:14  <andythenorth> nml is single threaded :P
09:29:17  <dwfreed> LordAro: independent objects can be parallelized there
09:29:20  <V453000> and I don't undersand why
09:29:25  <andythenorth> nml is all that matters :P
09:29:25  <_dp_> peter1138, also clean-compile openttd in < 30 sec ;)
09:29:36  <LordAro> dwfreed: until they have to interact with each other
09:29:43  <V453000> I blame the HEDT vs gaming grade CPU/motherboard but WTF
09:29:54  <dwfreed> LordAro: then they stop being independent objects
09:30:08  <dwfreed> LordAro: but there's a fixed point in time in the simulation where that happens
09:30:15  <LordAro> yes, and that's the point that people start complaining about performance :p
09:33:44  <peter1138> Basically you want more MHz AND more cores :p
09:34:02  <LordAro> except AMD cores are generally clocked higher than Intel :p
09:34:11  <peter1138> AMD cores are generally shitter.
09:34:33  <LordAro> exactly
09:34:39  * peter1138 ponders using guids as primary keys.
09:34:40  <LordAro> MHz isn't comparable
09:35:04  <peter1138> Yeah, I meant within the same architecture.
09:36:19  <reldred> Yeah I was running stuff that required a lot of threads at the time I bought my 8 core am3.
09:36:42  <reldred> Not so much these days
09:36:43  <LordAro> yeah, we use threadripper at work for our build servers
09:37:09  * andythenorth ponders multi-threaded nml :P
09:38:27  <andythenorth> map-reduce on the parse step :P
09:38:47  <LordAro> has anyone actually done any profiling on nml?
09:39:16  <LordAro> trying to optimise something without knowing what's actually slow is a recipe for disaster
09:42:35  <Artea> bah
09:42:41  <Artea> bad morning
09:43:03  <andythenorth> LordAro: frosch has afaik
09:43:08  <andythenorth> basic timings are easy to get
09:43:09  <andythenorth> https://paste.openttdcoop.org/prudxgwg7/1a0m81/raw
09:43:25  <andythenorth> a chunk of that is my stuff, but the nml timings are obvious there
09:43:45  <andythenorth> bbl
09:43:52  *** andythenorth has quit IRC
10:11:45  *** andythenorth has joined #openttd
10:48:35  * peter1138 ponders setting up Elite Dangerous again.
10:48:43  <peter1138> It lost all my binds in the last update :(
10:59:09  *** spnda has joined #openttd
12:01:03  *** nonki[m] has left #openttd
12:01:45  *** caubb has joined #openttd
12:02:19  *** tu2116v has joined #openttd
12:03:10  *** caubb has quit IRC
12:17:40  <DorpsGek_II> [OpenTTD/OpenTTD] stale[bot] commented on pull request #7025: Add #6887: Option to show zone inside local authority boundary of towns https://git.io/fjnFx
13:03:01  *** Flygon has quit IRC
13:08:57  *** supermop_work has joined #openttd
13:09:20  <supermop_work> yo
13:11:44  *** Samu has joined #openttd
13:19:48  *** TinoDidriksen has quit IRC
13:24:18  *** TinoDidriksen has joined #openttd
13:24:55  *** TinoDidriksen is now known as Guest1894
13:24:59  *** Guest1894 is now known as TinoDidriksen
13:35:06  *** gareppa has joined #openttd
13:40:32  *** gareppa has quit IRC
13:41:25  *** nielsm has joined #openttd
14:29:20  *** Alberth has joined #openttd
14:29:20  *** ChanServ sets mode: +o Alberth
14:42:55  <Samu> round 16 almost over
14:43:01  <Samu> 1 game remaining
14:46:27  *** sla_ro|master has joined #openttd
14:57:36  *** Gumle2 has joined #openttd
15:03:34  <peter1138> Hmm
15:05:08  * Artea on Doom OST (SC55) - I Sawed The Demons (Extended) @ https://www.youtube.com/watch?v=u5wZmjZ0yG0
15:05:46  <Artea> good song ::]
15:07:26  *** Gumle2 has joined #openttd
15:09:34  *** Wormnest has joined #openttd
15:13:01  <peter1138> "song"
15:26:52  *** Gumle2 has quit IRC
15:52:24  *** moll has joined #openttd
15:53:21  <moll> Hey! Are there any OpenTTD wiki admins here, too? I'd like to ask them to turn MediaWiki's API on for a wiki backup. It'd be so much faster and less load on the server to grab the wikitext through that way rather than via less machine-readable ways. :)
15:57:24  <Samu> round 16 complete!
15:59:25  <LordAro> moll: there's an awful lot of things that we'd like to do to the wiki :)
15:59:43  <LordAro> but the guy who normally deals with these things has been away for a while
16:00:05  <LordAro> i wouldn't worry too much about our server load though, so feel free to scrape it the slow way
16:02:09  <moll> Thx. To be honest, I'm not actually sure whether the Wikiteam's backuper does support other means... https://github.com/WikiTeam/wikiteam
16:02:25  <moll> I'm in no hurry, so maybe I can wait until that fellow returns.
16:02:35  <moll> Who are we waiting for? ^_^
16:02:48  <LordAro> @seen Truebrain
16:02:48  <DorpsGek> LordAro: Truebrain was last seen in #openttd 3 weeks, 0 days, 22 hours, 1 minute, and 42 seconds ago: <TrueBrain> if that is the excuse, I am done
16:03:09  <moll> Gotcha. Thanks!
16:03:13  <LordAro> we're, ah, hoping it's not permanent
16:05:58  <moll> Well, I hope the bus-factor of that wiki/server isn't one. :P
16:06:18  <LordAro> i don't think it's quite that bad :)
16:32:43  <peter1138> Hi
16:39:39  *** frosch123 has joined #openttd
16:43:02  *** Gumle2 has joined #openttd
16:45:40  *** TelepathicVampire has joined #openttd
16:45:56  *** andythenorth has joined #openttd
16:49:02  * Artea on UT'99 - Foregone Destruction
16:49:38  <TelepathicVampire> hey guys, im trying to compile 1.9.1 with visual studio 2017 (openttd_vs141) using the openttd useful libraries as explained in the compilation guide on the wiki. But i get a lot of link errors. Do i have to use nuget packages instead?
16:50:37  <nielsm> that's no longer correct
16:50:45  <nielsm> see the readme file for correct instructions
16:50:45  <nielsm> https://github.com/OpenTTD/OpenTTD/tree/1.9.1/#70-compiling
16:52:01  <TelepathicVampire> thanks, thats what i was looking for
16:59:11  *** HerzogDeXtEr has joined #openttd
17:06:41  *** andythenorth has quit IRC
17:09:38  *** andythenorth has joined #openttd
17:11:50  <TelepathicVampire> sorry, i'm not very experienced with those packages: i tried to find the best match for each library even though they sometimes don't include static keyword or even x86, they install fine but compiler says cannot find "png.h", "zlib.h" and so on
17:12:15  <andythenorth> yo
17:18:39  <Alberth> o/
17:19:09  *** Laedek_ has joined #openttd
17:19:10  <andythenorth> ohai Alberth :)
17:19:53  <LordAro> TelepathicVampire: i'm not super experienced with MSVC compilation, but i believe all the dependencies do need to be static
17:20:13  <LordAro> and match the architecture of whatever you're building as (probably should be x64)
17:20:21  <Alberth> wouldn't complain about .h then, would it?
17:20:42  <LordAro> depends how the include paths are set up, maybe?
17:21:20  <TelepathicVampire> i think my mistake was to assume i could use the nuget packages when the manual said to use vcpkg
17:21:24  <Alberth> possibly, it may be smart and ignore libraries if they don't fulfil all the requirements
17:22:09  <TelepathicVampire> sometimes i ignore explicit instructions for some reason^^
17:22:40  <andythenorth> so are we profiling nmlc then? o_O
17:23:58  *** Laedek__ has joined #openttd
17:24:04  *** Laedek has quit IRC
17:25:08  <Alberth> we are?
17:25:08  <Alberth> I tried that once, and concluded it collects memory, and does too much optimization iirc
17:25:09  <Alberth> so it does something wrong somewhere :p
17:26:16  *** gelignite has joined #openttd
17:28:55  <Alberth> I have no idea how to fix that tbh
17:29:00  *** Laedek_ has quit IRC
17:29:28  * andythenorth wonders how fast it _should_ be
17:29:35  <Alberth> would be nice if you could draw conclusions like "it never modifies an existing EXpression object
17:30:02  <Alberth> but that's not quite possible due to the dynamic nature of the language
17:30:23  <Alberth> except for generating images, I would say bloody fast
17:30:48  <Alberth> as far as I know, it doesn't do much computation at all
17:31:50  <Alberth> I also tried speeding up the parser by converting it to C++, but that had pretty much no impact, as time isn't spend there
17:32:10  <LordAro> profile before optimisation ;)
17:32:20  <Alberth> yep :p
17:32:20  <andythenorth> 10s out of 50s is spent in parser for Horse
17:32:27  <andythenorth> I suspect it might vary by grf
17:32:40  <Alberth> was a nice experiment to do anyway :)
17:33:05  <Alberth> likely
17:33:13  <Alberth> how much time for generating pixels?
17:33:38  *** Progman has joined #openttd
17:33:42  <Alberth> fr0sch rewrote that to c++, which helped a lot
17:34:51  <LordAro> i wonder whether something like https://github.com/uber/pyflame would reveal anything obvious
17:36:01  <andythenorth> Alberth: Horse-specific timings https://paste.openttdcoop.org/prudxgwg7/1a0m81/raw
17:37:42  <frosch123> the conversion to C only did factor 2
17:38:00  <frosch123> the major effect came from sorting the sprites before encoding and opening every source image only once
17:38:18  <frosch123> instead of opening/closing/reopening files as they appeared in the nml source
17:39:24  <andythenorth> maybe as a bandaid, I should stop invalidating the cache
17:39:27  <frosch123> LordAro: the main problem is that nml sources as generated by pyandynml and cppnml are huge
17:39:41  <andythenorth> can't be too hard to compare the pngs I need during generation
17:40:05  <Alberth> only one way to find out I guess :p
17:40:07  <frosch123> i think you can only speed up nml by adding higher level syntax, like functions/templates/whatever directly in nml
17:40:43  <LordAro> frosch123: how huge is huge?
17:40:54  <frosch123> megabytes
17:41:00  <LordAro> ah
17:41:44  <LordAro> it does seem like nml could do with some higher level syntax
17:41:44  <frosch123> 9945600 for firs.nml
17:41:59  <LordAro> given it seems people are only able to use it with a layer on top
17:42:02  <andythenorth> is the parser immune to parallelisation?
17:42:06  <Alberth> nml was good relative to nfo
17:42:37  <LordAro> mm
17:42:41  <DorpsGek_II> [OpenTTD/OpenTTD] agentw4b opened issue #7574: Damaged Newgrf PIPE train sprites in Openttd 1.9.1 in combination with another train set (V 1.8.0 was not the problem) https://git.io/fjnji
17:42:41  <Alberth> nobody could think beyond nml
17:43:06  <LordAro> grfcodec & nml is equivalent to an assembler and c compiler, i guess
17:43:13  <LordAro> one is always going to be slower than the other
17:43:25  <frosch123> the funny thing is, nml is way better at being grfcodec :p
17:43:55  <frosch123> that is, the encoding part (the only thing that grfcodec does) is faster in nml
17:44:08  <LordAro> heh
17:44:15  <frosch123> (about same speed for 8bpp, way faster for 32bpp)
17:44:16  * andythenorth wonders what to do next
17:44:46  <andythenorth> I could stop invalidating the cache
17:44:55  <andythenorth> that might save 5s
17:45:11  <andythenorth> I could switch to nfo, if the nml is unchanged (only pngs changed)
17:45:14  <andythenorth> that would save about 50s
17:45:24  <andythenorth> the biggest payoff seems to be switching to nfo
17:45:25  <Alberth> pyflame doesn't know about python 3.7 :(
17:45:33  <LordAro> ha
17:45:44  <andythenorth> a lot of the time I am tweaking sprites, and code is unchanged
17:46:37  <Alberth> so can't even compile it
17:47:55  <LordAro> Alberth: https://github.com/uber/pyflame/pull/153 :p
17:49:46  <Alberth> so they're stuck for not being able to press "merge" ;p
17:50:17  <LordAro> given there's been no activity since june
17:53:56  <andythenorth> looks like a fork waiting to happen :P
17:55:46  <DorpsGek_II> [OpenTTD/OpenTTD] agentw4b commented on issue #7574: Damaged Newgrf PIPE train sprites in Openttd 1.9.1 in combination with another train set https://git.io/fjnjx
17:56:02  <DorpsGek_II> [OpenTTD/OpenTTD] agentw4b closed issue #7574: Damaged Newgrf PIPE train sprites in Openttd 1.9.1 in combination with another train set https://git.io/fjnji
18:32:06  *** Gumle2 has quit IRC
18:36:19  <peter1138> Huh
18:36:33  <peter1138> I must have imagined that #7516 was merged...
18:52:20  *** Gumle2 has joined #openttd
18:56:11  *** glx has joined #openttd
18:56:11  *** ChanServ sets mode: +v glx
18:59:04  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcej
19:00:00  *** Alberth has left #openttd
19:10:55  *** Gumle2 has quit IRC
19:11:21  *** tokai|noir has joined #openttd
19:11:21  *** ChanServ sets mode: +v tokai|noir
19:18:04  *** tokai has quit IRC
19:32:19  *** glx_ has joined #openttd
19:32:19  *** Arveen has joined #openttd
19:32:37  *** Lejving_ has joined #openttd
19:32:39  *** sla_ro|master2 has joined #openttd
19:33:18  *** tycoondemon2 has joined #openttd
19:33:21  *** spnda has quit IRC
19:34:04  *** Taede_ has joined #openttd
19:34:13  *** m1crOman has joined #openttd
19:35:40  *** gnu_jj_ has joined #openttd
19:35:43  *** Execthts has joined #openttd
19:35:54  *** Wormnest has quit IRC
19:36:19  *** milek7_ has joined #openttd
19:36:23  *** MC8 has joined #openttd
19:36:38  *** orudge` has joined #openttd
19:36:48  *** berndj-blackout has joined #openttd
19:36:50  *** SpComb^ has joined #openttd
19:37:07  *** glx has quit IRC
19:37:07  *** sla_ro|master has quit IRC
19:37:07  *** TinoDidriksen has quit IRC
19:37:07  *** Fuco has quit IRC
19:37:07  *** VoId23[m] has quit IRC
19:37:08  *** gnu_jj has quit IRC
19:37:08  *** tycoondemon has quit IRC
19:37:08  *** Arveen2 has quit IRC
19:37:08  *** dag[m] has quit IRC
19:37:08  *** twom[m] has quit IRC
19:37:08  *** yur3shmukcik[m] has quit IRC
19:37:08  *** olmvnec[m] has quit IRC
19:37:08  *** godva[m] has quit IRC
19:37:08  *** iarp[m] has quit IRC
19:37:08  *** cawal[m] has quit IRC
19:37:08  *** Ammler has quit IRC
19:37:08  *** Yexo has quit IRC
19:37:08  *** XeryusTC has quit IRC
19:37:08  *** ^Spike^ has quit IRC
19:37:08  *** Hirundo has quit IRC
19:37:08  *** Lejving has quit IRC
19:37:08  *** milek7 has quit IRC
19:37:08  *** berndj has quit IRC
19:37:08  *** kgz has quit IRC
19:37:08  *** goodger has quit IRC
19:37:08  *** zuzak has quit IRC
19:37:08  *** _dp_ has quit IRC
19:37:08  *** DorpsGek has quit IRC
19:37:08  *** SpComb has quit IRC
19:37:08  *** m1cr0man has quit IRC
19:37:08  *** ericnoan has quit IRC
19:37:08  *** michi_cc has quit IRC
19:37:08  *** Exec has quit IRC
19:37:08  *** KenjiE20 has quit IRC
19:37:08  *** Taede has quit IRC
19:37:08  *** orudge has quit IRC
19:37:08  *** m1crOman is now known as m1cr0man
19:37:54  *** michi_cc has joined #openttd
19:37:54  *** ChanServ sets mode: +v michi_cc
19:37:56  *** Hirundo has joined #openttd
19:38:26  *** Ammler has joined #openttd
19:38:56  *** Yexo has joined #openttd
19:39:26  *** XeryusTC has joined #openttd
19:39:32  *** TinoDidriksen has joined #openttd
19:39:56  *** ^Spike^ has joined #openttd
19:40:03  *** dP has joined #openttd
19:40:03  *** TinoDidriksen is now known as Guest1978
19:40:04  *** dP is now known as _dp_
19:40:06  *** KenjiE20 has joined #openttd
19:41:26  *** DorpsGek` has joined #openttd
19:41:26  *** Fuco_ has joined #openttd
19:41:26  *** goodger has joined #openttd
19:41:26  *** dag[m] has joined #openttd
19:41:26  *** cawal[m] has joined #openttd
19:41:26  *** yur3shmukcik[m] has joined #openttd
19:41:26  *** olmvnec[m] has joined #openttd
19:41:26  *** ericnoan has joined #openttd
19:44:16  *** DorpsGek` has quit IRC
19:44:29  *** DorpsGek has joined #openttd
19:44:29  *** ChanServ sets mode: +o DorpsGek
19:46:52  <DorpsGek_II> [OpenTTD/OpenTTD] ldpl commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcfk
19:48:03  <DorpsGek_II> [OpenTTD/OpenTTD] Milek7 updated pull request #7510: Emscripten support https://git.io/fjm0r
19:52:42  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcf3
19:52:44  *** godva[m] has joined #openttd
19:55:23  *** gelignite has quit IRC
19:56:58  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
20:00:46  *** iarp[m] has joined #openttd
20:06:08  *** twom[m] has joined #openttd
20:09:31  *** arikover has joined #openttd
20:11:31  <Samu> wow, weird victories
20:11:37  <Samu> WmDOT beats CivilAI
20:12:02  <Samu> CluelessPlus beats WormAI
20:42:38  *** Guest1978 is now known as TinoDidriksen
20:43:38  *** VoId23[m] has joined #openttd
20:50:48  <argoneus> good evening train friends
20:51:09  <argoneus> I'm wondering - any idea what the openttd music was made in? e.g. "plain simple relax song"
20:51:16  <argoneus> is it a tracker or
20:54:09  *** berndj-blackout is now known as berndj
20:59:37  <FLHerne> argoneus: According to http://bundles.openttdcoop.org/openmsx/releases/LATEST/readme.txt that was made by Tistou Blomberg
21:01:09  <argoneus> I was wondering about the specific software
21:02:36  <FLHerne> Yeah, I was hoping he had a blog or a forum post or something
21:02:40  <FLHerne> But I can't find one
21:03:07  <DorpsGek_II> [OpenTTD/OpenTTD] michicc commented on pull request #7572: GDI engine for font glyph rendering as a replacement for FreeType (take 2) https://git.io/fjcJK
21:03:17  <DorpsGek_II> [OpenTTD/OpenTTD] michicc updated pull request #7572: GDI engine for font glyph rendering as a replacement for FreeType (take 2) https://git.io/fjngD
21:04:13  *** andythenorth has quit IRC
21:07:08  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN requested changes for pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcJy
21:08:23  <FLHerne> argoneus: So far the only information I can find for those particular tracks is https://www.tt-forums.net/viewtopic.php?f=29&t=46479&p=863101#p863101
21:08:28  <FLHerne> (which is useless)
21:08:35  <argoneus> ;;
21:09:01  <FLHerne> If you don't care which track, there's some discussion of other ones in that thread
21:09:34  <argoneus> any is fine, I'm just wondering if it's some old tracker software
21:10:16  <FLHerne> e.g. https://www.tt-forums.net/viewtopic.php?f=29&t=46479&p=860777&hilit=software#p860777
21:10:56  <Eddi|zuHause> you don't need tracker software to write midi music
21:11:07  <FLHerne> Anvil, Sibelius, Guitar Pro (?)
21:11:45  <argoneus> what would you use?
21:11:47  <Eddi|zuHause> muse score is a current FOSS one
21:12:19  <Eddi|zuHause> (unlikely that that was around when openmsx was made)
21:14:39  *** arikover has quit IRC
21:21:44  *** Progman has quit IRC
21:25:12  *** Wormnest has joined #openttd
21:29:48  *** frosch123 has quit IRC
21:30:55  *** sla_ro|master2 has quit IRC
21:31:47  *** nielsm has quit IRC
21:39:32  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
21:40:53  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcU3
21:49:24  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcUB
22:04:03  *** Wormnest has quit IRC
22:18:10  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
22:18:44  <peter1138> Nearly
22:19:18  <Eddi|zuHause> CI needs retriggering...
22:19:45  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko commented on pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjcUd
22:21:29  <Eddi|zuHause> wtf... github decided to not let me post anymore?
22:21:49  <Eddi|zuHause> "You can't comment at this time — your comment At least one email address must be verified to do that." wtf does that even mean?
22:22:04  <peter1138> Probably something broke.
22:23:00  <Eddi|zuHause> i'm not going to bother with github tonight then...
22:26:56  <DorpsGek_II> [OpenTTD/OpenTTD] dorobouNeko updated pull request #7573: Fix #7561: Remove assumption between power and cost https://git.io/fjnyI
22:27:41  <peter1138> Well, it's trying now.
23:07:48  *** Heiki has quit IRC
23:43:31  *** Fuco_ has quit IRC
23:48:50  *** HerzogDeXtEr has quit IRC

Powered by YARRSTE version: svn-trunk