IRC logs for #openttd on OFTC at 2019-04-08
⏴ go to previous day
00:14:30 <TrueBrain> glx: minor request, please don't use hashes for 'fixup'. Just mention something out of the commit message :) That would really help me out :D
00:14:38 <TrueBrain> and tnx for fixing :)
00:37:42 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7453: Remove AutoDeleteSmallVector and AutoFreeSmallVector https://git.io/fjtO6
00:45:25 *** Corns343434 has joined #openttd
01:01:55 *** Corns343434 has joined #openttd
01:08:31 *** Corns343434 has joined #openttd
01:53:22 <peter1138> And someone wants a Windows 9x build... o_O
01:58:54 <Samu> if i multiply a double with a uint, what do I get? a uint or a double?
01:59:45 <Samu> line 33, avg * active_scripts
02:00:20 <Samu> active_scripts is the only uint there
02:02:46 <Samu> GL_RATE is 33.3333333333
02:03:30 <Samu> maybe all > GL_RATE isn't triggering yet
02:04:33 <Samu> over 30.00 ms is already displayed as red, but I'm only doing subtractions at 33.33
02:05:57 *** Corns343434 has joined #openttd
02:08:14 <Samu> oh, interesting, target is actually 30.0000000
02:08:30 <Samu> i thought it was the same as GL_RATE
02:08:37 <Samu> where or how is this 30 coming from
02:09:55 <peter1138> MILLISECONDS_PER_TICK?
02:10:34 <Samu> ah, I see, looks like that's it
02:11:10 <Samu> so what I want is MILLISECONDS_PER_TICK, not GL_RATE
02:13:13 <Samu> MILLISECONDS_PER_TICK is a uint, hmmm
02:14:15 <Corns343434> What places can i ask somebody to help me draw a tool icon?
02:17:45 *** Thedarkb-X40 has joined #openttd
02:28:54 <Samu> strange, the parameters aren't loaded as they were set
02:29:53 *** tokai|noir has joined #openttd
02:29:53 *** ChanServ sets mode: +v tokai|noir
02:34:15 <Samu> random deviation is broken
02:39:01 <Samu> when i start a game, from main menu, no random occurs
02:39:25 <Samu> if i save the game and then load it back, those settings are randomized on load
02:46:52 <Samu> I think AnchorUnchangeableSettings should anchor CONFIG_RANDOM settings
03:31:33 <Corns343434> How do i go about adding my own custom sprites to the game? Eg if i was making a pull request to trunk
03:33:53 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtsp
03:39:31 *** Corns343434 has joined #openttd
03:45:20 <Samu> there's still another problem, related
03:48:00 *** Supercheese has joined #openttd
03:52:47 *** Corns343434 has joined #openttd
03:52:57 <Corns343434> nnyby thank you :)
04:44:42 <nnyby> my ships keep saying they're lost but they're not >_<
04:50:10 <nnyby> these ships are making a profit.. vehicles can recover from getting lost, but then get lost again following the same path? hmm..
04:57:51 <Samu> from what I know, the "lost ship feature" was removed on master
04:58:21 <Samu> it was a distance check between ship position and destination
04:59:46 <Samu> it was done to prevent calling the pathfinder every tile
05:00:03 <Eddi|zuHause> tried changing the pathfinder?
05:00:19 <nnyby> i am playing on master. i havent tried changing the pathfinder
05:00:46 <nnyby> i did see peter removed the ship max order distance, which is nice
05:01:24 <Samu> this AnchorUnchangeableSettings fix is probably the wrong fix, I found another way
05:10:34 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtsp
05:11:49 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtGj
07:55:29 *** andythenorth has joined #openttd
08:18:17 *** Supercheese has joined #openttd
08:33:09 *** Supercheese has joined #openttd
10:26:16 <planetmaker> I... got a reply by DaleStan on my Factorio bug report oO
10:26:30 <planetmaker> the world is tiny
10:27:38 <peter1138> andythenorth, what's different?
10:27:56 <andythenorth> peter1138: pixels
10:28:55 <andythenorth> what shall I draw next?
10:28:55 <peter1138> nnyby, wrong advice about adding custom sprites to the game, though.
10:29:07 <andythenorth> I have 3 big newgrf projects
10:29:19 <peter1138> Ah yes, those round buttons are better.
10:29:20 <andythenorth> and drawing a little bit for each keeps them nicely far away from finished
10:29:48 <andythenorth> sounds familiar? o_O
10:29:57 <Xaroth> andythenorth: you know this perhaps, but tech trees possible with newgrf (plus/minus usage of GS)
10:30:11 <Xaroth> Or is that wishful thinking
10:31:26 <andythenorth> tech tree would be best handled by OpenTTD
10:31:39 <andythenorth> newgrf could do it, but it would be specific to each newgrf, so quite weird
10:31:57 <andythenorth> GS can't do it, because GS is hopeless
10:32:18 <andythenorth> tech trees would answer the majority of daylength wishes, afaict
11:04:11 <peter1138> What would a tech-tree look like in OpenTTD?
11:05:08 *** andythenorth has joined #openttd
11:11:41 <LordAro> peter1138: like a low res jpeg screenshot of civ3's tech tree
11:12:19 <peter1138> Ok, look wasn't the right word :P
11:12:25 <peter1138> How would a tech tree fit in?
11:16:40 <andythenorth> some kind of epoch / era per vehicle
11:16:49 <andythenorth> and then the game advances them at a default rate
11:16:52 <andythenorth> which triggers intro
11:17:11 <andythenorth> default rate is obvs. 1 tech tree year = 1 year
11:17:24 <andythenorth> then let GS manipulate the current state of tech
11:17:42 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7431: Fix aa7ca7fe6: Linkgraph node index order must be maintained due to other references. https://git.io/fjtW1
11:17:43 <andythenorth> bonus points: let industries and houses also understand it
11:19:38 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7452: Fix 6fc60d8c4f: forgot to update API changelog https://git.io/fjtWy
11:20:05 <peter1138> Ok, I'm going to stop commenting.
11:23:30 <peter1138> So there's 3 commits marked for backport that have not been merged into master yet.
11:25:37 <peter1138> Marking things for backport when they don't affect 1.9 is not particularly useful :/
11:28:49 <peter1138> planetmaker, so please don't do that ;)
11:47:18 *** Flygon_ has joined #openttd
11:48:30 <andythenorth> so given the orders of a train, we should be able to predict statistically the route it will take?
11:48:46 <andythenorth> so we'd know the maximum grade?
11:49:07 <andythenorth> so we could calculate HP-per-ton and the resulting speed over the grade?
11:49:10 <andythenorth> we probably shouldn't :P
11:49:16 <andythenorth> sledgehammer -> nut
11:50:14 <peter1138> Which nut are you trying to crack?
11:50:50 <andythenorth> I just run a train over the route, and add more engines if it's too slow :P
12:05:05 <planetmaker> peter1138, yes, not yet. But when they are, they are bug fixes which need backporting
12:05:26 <planetmaker> so ... as it's a thing often overlooked I thought it a good idea to mark them straight away
12:05:58 <peter1138> Backporting to what?
12:06:30 <planetmaker> to 1.9.x. Whether that is 1.9.1 or later. In the assumption that the suggested fixes get merged soonish
12:06:52 <peter1138> But you marked things for backport which don't affect 1.9.
12:09:20 <peter1138> #7471, #7456, #7452, #7431
12:09:39 <peter1138> #7435 and #7419 are okay though :)
12:16:03 <peter1138> Yeah, I'm evaluating my choices here :p
12:16:20 <peter1138> It's good to have things marked as otherwise we would miss important fixes.
12:16:42 <peter1138> And it was easy for _me_ know the ones which are not relevant.
12:16:58 <planetmaker> I looked at what they fixed... but obviously I was not exactly awake
12:21:19 <peter1138> Hmm, testing out a 32" screen. Somewhat large, aren't they?
12:23:19 * andythenorth lives in a 13" world
12:23:31 <peter1138> It's not HiDPI though, so I can't test that 4k issue ;)
12:24:08 <planetmaker> hm, 32" and only 2560x1440?
12:24:38 <planetmaker> One of my monitors quit on me the other day... now I'm pondering to replace both... by larger ones
12:24:47 <peter1138> Definitely doesn't need to be on a stand to raise it.
12:25:14 <planetmaker> or just taking one big one... but hm
12:25:18 *** planetmaker has left #openttd
12:25:21 <peter1138> Yes, pixel size is pretty good though, about equivalent to 23" 1080p.
12:25:23 *** planetmaker has joined #openttd
12:25:23 *** ChanServ sets mode: +o planetmaker
12:25:26 <peter1138> Yes, pixel size is pretty good though, about equivalent to 23" 1080p.
12:25:45 <peter1138> If I was buying one for myself I'd want 4k though.
12:26:31 <peter1138> Or 5k? I think Apple have 5k screens.
12:27:28 <andythenorth> they sell an LG or something
12:28:29 <peter1138> I think 32" is too big!
12:28:37 <peter1138> I'm going to get neck strain.
12:30:14 <planetmaker> move your chair backwards :)
12:30:46 <planetmaker> will be a total win... moving so far back that the bigger screen has the same viewing angle as the old, smaller one :P
12:31:13 <peter1138> 1080 -> 1440 is much more of an improvement than I'd imagined, though.
12:31:36 *** planetmaker has left #openttd
12:31:41 *** planetmaker has joined #openttd
12:31:41 *** ChanServ sets mode: +o planetmaker
12:32:49 <planetmaker> don't browse and chat :P
12:33:47 <peter1138> So my issue is I currently have 3x 22" 1680x1050 monitors at home.
12:34:09 <peter1138> If I get wider monitors, I can only fit 2x in.
12:34:12 <planetmaker> upright or landscape?
12:34:22 <planetmaker> that's pretty wide :)
12:34:33 <peter1138> It's pretty good for driving games :-)
12:34:48 <peter1138> Also good for developing OpenTTD.
12:34:51 <planetmaker> for openttd 1024 x 256 maps :P
12:35:11 <peter1138> I tend to run things in half-width full-height windows.
12:35:15 <planetmaker> see everything concurrently... it's interesting
12:35:24 <peter1138> So I get 6 views concurrently.
12:35:44 <peter1138> Usually a source view, a compilation view, or git/tig output.
12:35:45 <planetmaker> yo, for dev many screens is gold
12:35:48 <peter1138> IRC on another screen :)
12:36:10 <peter1138> But yeah, if I dropped back to 2 screen, I'd have a horrible bezel right in the middle.
12:36:58 <planetmaker> doesn't matter for dev... but yes, car racing :D
12:39:29 <peter1138> To be fair, since I got back into OpenTTD coding, I really haven't done anything much else.
12:39:37 *** andythenorth has left #openttd
12:39:47 <planetmaker> it quickly eats a lot of time
12:44:14 <peter1138> Hmm, maybe I should test #7484 on this screen
12:44:39 <peter1138> Maybe I should do some work.
13:09:28 <DorpsGek_II> [OpenTTD/OpenTTD] JGRennison opened pull request #7488: Fix: Crash due to use of invalid iterator in ClientNetworkContentSocketHandler https://git.io/fjt8H
13:20:57 <peter1138> Github is quite nice at 1280x1440 :-)
13:35:56 <peter1138> Hmm, not sure on #7460, two ways to fix it, one like in the PR, and the other is as in my comment.
13:40:14 <milek7> does it work? there should be one server available
13:42:07 <V453000> where is our train photo specialist Mr. North
13:45:10 <peter1138> It loaded but the music was loud in the office so I had to close it :p
13:47:43 *** planetmaker has left #openttd
13:47:49 *** planetmaker has joined #openttd
13:47:49 *** ChanServ sets mode: +o planetmaker
13:48:26 <V453000> didn't try that, already closed :d
13:59:52 <pnda> Why do I feel like a lot of devs have strange monitors
14:01:56 <pnda> Like peter said 1280x1440. Is that not 3:4?
14:03:04 <V453000> well you usually say x first and that's not 4:3 I think :)
14:03:14 <V453000> so suspicion is in place :D
14:05:41 <peter1138> No, it's not 4:3 :p
14:06:05 <peter1138> It's also half the screen.
14:06:25 <pnda> So what is the total resolution?
14:07:02 <peter1138> Not excessive for 32"
14:07:18 <pnda> 1440p is quite normal at 32"
14:07:28 <pnda> I'm getting a 1440p with 29"
14:07:52 <peter1138> Colours are very vibrant. Hmm.
14:08:11 <pnda> So I guess a IPS screen?
14:08:22 <peter1138> Viewsonic VX3211-2K-mhd
14:09:37 <pnda> Ah that one.. relatively mid-level but good
14:11:13 <pnda> With my next update I'll have 3 screens, 1x 2560x1440, 1x 2560x1080, 1x 1920x1080... Yes a quite weird setup
14:11:16 <V453000> that's what I suspected :)
14:12:29 * peter1138 has 3x 1680x1050 at home. Very old-school.
14:12:53 <pnda> You're at work? huh and yes that's quite an oud
14:13:31 <LordAro> it still displays stuff, doesn't it?
14:14:04 <peter1138> And I'm not using VGA with any of them, so that's a plus.
14:14:24 <Xaroth> I'm using 2x 2560*1440p, the amount of space it gives is awesome :P
14:15:14 <pnda> I love 1440p, but I sadly only have the money for one high-end 1440p monitor so I decided to keep my 2 current ones and mount the current main above to the wall
14:15:16 <peter1138> I might consider moving my computer elsewhere so I can upgrade.
14:15:35 <peter1138> Or just drop back to 2 screens.
14:16:04 <LordAro> 2x1080p, 1x1440p at home
14:28:02 <Samu> startai command in the console is tricky
14:30:39 <peter1138> Ah, hitting the sprite cache limit kills performance, heh
14:36:36 <Samu> do you remember that last random deviation fix?
14:37:44 <Samu> for some reason, when start date is a value other than 0 and the ai is random ai, no random deviation occurs
14:38:33 <Samu> looks like last fix, fix one thing, but broke another
14:51:13 *** Lejving_ has joined #openttd
14:54:25 <Samu> this is so complex to follow
15:34:02 <pnda> Do fields from farms have any special name? sub-industries?
15:40:25 <pnda> Well they could be unified
15:40:37 <peter1138> The tile type is the same as a clear type.
15:40:42 <peter1138> The tile type is the same as a clear tile.
15:40:57 <peter1138> Unified with what and why?
15:41:36 <pnda> Why not let a industry, from a GRF, specify a special type of tile that works just like fields and gets placed randomly in a radius of the industry
15:43:13 <pnda> Ahhhh wasn't this already a thing with industry behaviour?
15:43:36 <LordAro> pnda: you might find that a lot of the answers to your questions are "because no one has done it yet"
15:44:11 <LordAro> oil rigs are the other obvious example
15:47:25 <Samu> field fences are the 2nd most cpu intensive from tileloop
15:51:14 <planetmaker> but it's related :P
15:52:03 <planetmaker> That callback 3B does not exactly exist (yet)
15:52:29 <planetmaker> to the extend described there
16:02:58 <_dp_> ffs asrock, 2h driver download :(
16:03:16 <_dp_> modern file sizes dont mix well with 90s speeds
16:03:28 <V453000> it's called rock made out of ass for a reason :P
16:04:01 <V453000> idk what is it with motherboard manufacturers and ass obsession
16:04:16 <V453000> I will see myself out.
16:04:38 <_dp_> it probably mean something cool in chinese :)
16:06:14 <V453000> something along the lines of ass
16:06:46 <pnda> ffs asus giving me a broken gpu... I think that's worse than a long driver download ^^
16:07:43 <pnda> planetmaker: does frosch still work on CB 3B & is whats done available in trunk/master?
16:08:10 <planetmaker> I don't think so. It's a collection of ideas
16:08:36 <planetmaker> master allows industries to plant fields. Or trees. But you knew that.
16:08:42 <planetmaker> that is done via CB 38
16:08:50 <Samu> after furher investigation, looks like random deviation fix didn't break anything, just tested without that fix, and it still happens. It was already there
16:08:55 *** HerzogDeXtEr has joined #openttd
16:09:58 <Samu> the bug was already present even before the fix
16:11:16 <Samu> random deviation doesn't occur for start_date even when it's not 0
16:22:12 *** WWacko1976-work has quit IRC
16:23:20 *** sla_ro|master has joined #openttd
16:23:33 <Samu> i fail at understanding what the heck SettingValueList::const_iterator is doing
16:24:13 <Samu> but this is the thing where start_date is sometimes pulled out, and some other times not
16:26:50 <Samu> this->settings[stredup((*it).first)] = (*it).second;
16:27:15 <Samu> stredup((*it).first) is "start_date"
16:27:23 <Samu> but this->settings has a size of 0
16:27:27 <Samu> something doesn't feel right
16:28:36 <Samu> what would this->settings["start_date"] be getting if this->settings is size=0 ?
16:34:53 *** Wormnest has joined #openttd
16:36:24 <LordAro> Samu: it's a map, it automatically creates new entries
16:37:49 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fjt06
16:43:55 <Samu> there's ScriptConfigItemList::const_iterator
16:44:08 <Samu> and there's SettingValueList::const_iterator
16:44:27 <Samu> too easy to get confused
16:45:41 <nielsm> nothing to be confused about, they're different types
17:03:22 *** synchris has joined #openttd
17:16:48 <Samu> ScriptConfigItemList doesn't contain "start_date" on Random AIs, so when random valuator is called, it doesn't do anything about it
17:37:52 <peter1138> Do we do a RC for 1.9.1?
17:39:22 <nielsm> if it's just bug fixes and it looks safe imo not
17:39:27 <pnda> If at all only one I'd say
17:39:49 <nielsm> if it turns out to introduce new bugs then fast-track 1.9.2
17:40:10 <peter1138> Would be good to release tonight as it's 1 week after 1.9.0 ;)
17:41:30 <pnda> Would all backport-requested PRs then get merged?
17:42:54 <Samu> map() : _Mybase(key_compare()) { // construct empty map from defaults
17:43:22 <nielsm> it's part of the C++ standard template library
17:43:34 <peter1138> pnda, only if they're already merged into master.
17:44:00 <Samu> or maybe a better question, where does it get the defaults?
17:44:09 <peter1138> Hmm, doing the changelog is awkward. The commit messages by themselves usually lack context and make no sense for a public list.
17:45:46 <peter1138> Ah, #7483 could be backported too.
17:53:17 <Samu> - config_list { size=??? } std::list<ScriptConfigItem,std::allocator<ScriptConfigItem> >
17:53:21 <Samu> - config_list 0xcdcdcdcdcdcdcdcd { size=??? } std::list<ScriptConfigItem,std::allocator<ScriptConfigItem> > *
17:53:38 <Samu> there are 2 config_lists
17:53:50 <Samu> and then there still settings
17:54:01 <Samu> 3 lists with pretty much the same thing
18:14:43 <V453000> ofc colours and textures not final
18:24:14 *** Hobbyboy|BNC has joined #openttd
18:24:28 *** Hobbyboy|BNC is now known as Hobbyboy
18:28:42 <DorpsGek_II> [OpenTTD/OpenTTD] JGRennison commented on issue #7481: Fast forwarding very fast can cause window.cpp/kdtree.hpp crash https://git.io/fjtuN
18:29:46 *** Alberth has joined #openttd
18:29:46 *** ChanServ sets mode: +o Alberth
18:29:48 <Samu> line 77 of script_config.cpp is failing to do what's supposed to be doing
18:30:12 <Samu> this->PushExtraConfigList(); should add the start_date to the config_list
18:30:35 <Samu> virtual void PushExtraConfigList() {};
18:31:45 <Alberth> it's virtual, a derived class may be doing something
18:32:31 <Alberth> although I fail to see how a function PushExtraConfigList() would do anything with start_date
18:32:37 *** Progman has joined #openttd
18:32:54 <Samu> this->config_list->push_back(_start_date_config);
18:33:18 <Samu> but that one is void AIConfig::PushExtraConfigList()
18:35:22 <peter1138> Oh my god my 1680x1050 screens feel tiny now :(
18:36:38 <Samu> new AIConfig(_settings_newgame.ai_config[c]); this is the root function
18:39:06 <Alberth> got a bit spoiled peter :)
18:40:31 <V453000> 3440 x 1440 master race <3
18:40:46 <nielsm> I kind of want a non-wide screen for coding on...
18:41:24 <nielsm> (someone is going to respond "get a gigantic display and split it in halves")
18:42:16 <peter1138> If it's big enough, that does work really well.
18:43:12 <Alberth> rotate it 90 degrees niels?
18:43:43 <milek7> hm, emscripten sdl2 video output is slow on higher resolutions, 20ms+
18:45:41 <nielsm> Alberth that tends to become too tall for comfort
18:45:58 <peter1138> I find the subpixel arrangement irritating if the screen is rotated.
18:46:11 <nielsm> 2048x1536 would be a good display size
18:46:35 <Alberth> 2nd hand shops perhaps
18:47:40 <peter1138> If you're lucky you might find 1600x1200.
18:47:41 <Alberth> and nobody said you have to use the entire height :)
18:48:27 <nielsm> at work I do have one display in vertical
18:49:09 <nielsm> rather expensive, but then it is eizo
18:49:40 <peter1138> Probably an old model anyway.
18:50:41 <nielsm> "Date First Available 20 Oct. 2013"
18:51:05 <peter1138> Probably missing quite a chunk of cost-saving tech, though.
18:51:17 <nielsm> now if just they would begin making OLED panel computer displays
18:51:32 <nielsm> instead of reserving the tech for 10 inch tablets and 53 inch televisions
18:51:36 <nielsm> and nothing in-between
18:54:24 *** Smedles_ has joined #openttd
18:54:44 <peter1138> Minor improvement already on ProZone 13.
18:54:53 <peter1138> Which doesn't have too much going on.
18:55:37 <V453000> nice game to test with :P
18:55:37 <peter1138> Graphics rendering drops about .75ms
18:56:07 <peter1138> Emperor Jake's NetTrans is good for this patch, I think
18:56:31 <peter1138> If you can find all the NewGRFs...
19:02:28 <peter1138> Still only about 5ms.
19:02:41 <peter1138> Hmm, less now it's settling.
19:06:07 <planetmaker> only thing it's missing is some shading on vehicles
19:06:12 <Eddi|zuHause> you mean the jpeg artifacts, right?
19:06:35 <peter1138> Shading. Detail. Correct sizing...
19:06:39 <peter1138> Quite a lot missing.
19:06:48 <peter1138> It's a shame I can't draw for shit.
19:06:57 <planetmaker> no. That's just not your graphics style
19:07:01 <Eddi|zuHause> is that a donkey or a horse on the plaza?
19:07:40 <planetmaker> judging all graphics by "is it sufficiently similar to ttd" won't cut it. It's a bit ... pop art style
19:07:42 <nielsm> isn't the statue a spider?
19:08:05 <nielsm> (in original graphics)
19:08:21 <planetmaker> here it is a fossil dinosaur :P
19:08:57 <Alberth> I was thinking mule :)
19:09:23 <Eddi|zuHause> so something between "both" or "neither"?
19:09:43 <peter1138> I'm just judging it by "does it look good"
19:09:50 <peter1138> And that is, IMHO, no.
19:10:55 <Eddi|zuHause> that's the weird thing, some things don't need to be good to find their target audience
19:12:02 <Eddi|zuHause> is that actually happening now?
19:12:13 <peter1138> Hmm, well, 7484 is marginally faster on this game, so that's something.
19:12:51 <Alberth> it may be simpler to have another election about it, and it gets voted 'no' :p
19:13:49 <peter1138> Did you hear one of the Hard-brexiters in the government is now... calling for a revote... because we know more now... and the outcome would be different...
19:13:52 <peter1138> But not on Brexit, no.
19:14:03 <peter1138> On the no-confidence motion in their leader.
19:14:11 *** andythenorth has joined #openttd
19:15:00 <Alberth> planetmaker: unsticky the titlescreen competition thread?
19:16:41 <peter1138> Hmm, now this patch is about 4K resolutions being slow.
19:16:56 <peter1138> I'm on 1680x1050, so kinda... not comparable.
19:17:38 <Eddi|zuHause> try zooming out?
19:18:02 <peter1138> Yeah, it's about 4K resolutions being slow when zoomed out...
19:18:14 <Eddi|zuHause> try zooming out further? :p
19:18:50 <V453000> I did notice massive slowdown when zooming to 0.5
19:19:12 <Eddi|zuHause> V453000: thanks for volunteering as test subject :p
19:19:34 <peter1138> That might just be sprite cache size.
19:19:37 <V453000> I'd love to test, but you'll need to show me how to build openttd on windows first :P
19:19:45 <peter1138> This particular issue is with lots of overlapping sprites, stressing the sprite sorter.
19:20:59 <Eddi|zuHause> so, like, bridges?
19:21:15 <Eddi|zuHause> trucks in towns with large buildings
19:21:28 <peter1138> Well, not overlapping.
19:21:34 <peter1138> Just lots of sprites. They all need sorting.
19:22:03 <Eddi|zuHause> i'm assuming non-overlapping is harder to get into worst-case sorting?
19:22:27 <peter1138> I guess lhttps://github.com/OpenTTD/OpenTTD/pull/7484
19:27:07 <peter1138> Hmm, it was FIRS that broke before.
19:28:17 <peter1138> Nope, still broken.
19:28:25 <andythenorth> FIRS did what to you now?
19:28:53 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7484: Codechange: Sprite sorting optimization https://git.io/fjtgJ
19:29:34 <peter1138> Well that solves that dilemma :p
19:30:55 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtsp
19:31:20 <Samu> well, that PR fixes 2 things now :|
19:32:58 <Samu> took me a while to understand where the problem was
19:37:47 <Samu> this->info == NULL aka I'm a random AI, i think
19:38:50 <Samu> ScriptConfig constructor thing doesn't construct the start_date in random AI
19:39:04 <Samu> so there was no setting at all to deviate
19:40:25 <Samu> PushExtraConfigList calls the AI version of the function which adds the "start_date"
19:40:41 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7460: Fix #6222: Advanced sprite layout sometimes showed incorrect railtype ground tile. https://git.io/fjtgO
19:40:55 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7460: Fix #6222: Advanced sprite layout sometimes showed incorrect railtype ground tile. https://git.io/fjtg3
19:41:10 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7460: Fix #6222: Advanced sprite layout sometimes showed incorrect railtype ground tile. https://git.io/fjIRU
19:41:17 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN closed issue #6222: Stations: using property 1A sometimes shows incorrect railtype ground tile https://git.io/fjtgZ
19:41:49 <Samu> AddRandomDeviation then calls the AI version of the function which already does the right thing, so that's why it doesn't run the 2nd part of the code
19:44:17 <peter1138> Well, I can't backport things which are merged yet. Hmm.
19:44:38 <Samu> does this PR deserve backporting?
19:45:45 <Samu> start_date = 0 isn't a thing in 1.9
19:47:12 *** frosch123 has joined #openttd
19:47:47 <Samu> makes no sense if it is because start_date = 0 isn't
19:48:14 <peter1138> Samu, all your crap was already reverted for 1.9
19:51:04 *** gelignite has joined #openttd
19:51:49 <peter1138> You're the only person using AIs like that anyway.
19:51:55 <peter1138> So it hardly matters.
19:52:45 <peter1138> So are we merging #7480 or not?
19:54:30 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7480: Fix #6666: arabic town directory strings https://git.io/fjtgB
19:59:40 <peter1138> Do I reset APPV_BUILD from 4 to 0 for 1.9.1?
19:59:54 <pnda> peter1138 Would the PR not have to be renamed if those two strings are fixed?
20:00:05 <peter1138> pnda, I have no idea.
20:00:25 <peter1138> Doesn't necessarily need to be fixed there, just mentioned it as related.
20:00:44 <pnda> I can fix it there, would be easier to oversee
20:05:49 <LordAro> peter1138: might as well
20:07:07 <peter1138> 6 PRs backported now o_O
20:07:10 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtg6
20:07:48 <andythenorth> anyone drawn the Carbon Black Plant for FIRS yet? o_O
20:09:31 <peter1138> "Action 4 has feature 48 'original strings'"
20:09:39 <peter1138> Doesn't really translate well to a changelog :p
20:10:17 <LordAro> "fix: newgrf station strings did not display"
20:10:24 <nielsm> "Action 4 feature 48 strings were not being loaded from NewGRFs causing various bugs"
20:10:27 <peter1138> Not just stations :-)
20:10:37 <LordAro> some newgrf strings...
20:10:38 *** chomwitt has joined #openttd
20:11:19 <peter1138> Missing NewGRF strings due to Action 4 feature check skipping pseudo-feature 48
20:12:27 <frosch123> "make newgrf older than 2010 work again"
20:14:23 <peter1138> There's some open PRs tagged for backport. Anyone dare approve them?
20:14:59 <peter1138> Might be getting into 1.9.2 territory.
20:15:11 <peter1138> They're not exactly critical thouhg.
20:15:46 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7483: Fix #7478: Don't remove NewGRF objects on company take-over. https://git.io/fjtg5
20:16:58 <nielsm> Fix: [Windows] Error in error handling could lead to erroneous error boxes
20:17:30 <nielsm> Fix: [Windows] Incorrect error handling could lead to cascading error boxes
20:18:47 <peter1138> - Fix: [Windows] Incorrect error handling could lead to cascading error windows (#7482)
20:20:31 <peter1138> Is #7478 safe enough to backport? Seems quite simple to me.
20:20:46 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN closed issue #7478: Newobject disapears if company owner get buyed by another company https://git.io/fjLQf
20:20:47 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7483: Fix #7478: Don't remove NewGRF objects on company take-over. https://git.io/fjLp6
20:26:48 <andythenorth> 'probably fine' is the best
20:28:14 <pnda> Seems I fixed the other arabic strings
20:28:20 <pnda> Are there any other known ones?
20:30:56 <pnda> ok these are still broken...
20:39:50 <pnda> How the hell did this happen though
20:40:18 <LordAro> well those look like strgen errors
20:40:24 <LordAro> so the lang file has been mangled in some way
20:40:35 <pnda> I know... but it isn't shown by VS
20:41:38 <LordAro> aiui VS does like to add BOMs sometimes
20:41:52 <pnda> Only touched this file with VS
20:44:27 <TrueBrain> sorry peter1138, seems I am incapable of switching branches before hitting 'git push' :(
20:44:33 <pnda> Now why thell is it skipping building the .exe
20:45:46 <TrueBrain> change rename the PR: stuff for 1.9.1, and you will be fine :P
20:47:24 <pnda> IT'S STILL NOT FIXED WHYYYYYYYYY
20:47:58 <andythenorth> it's TrueBrain !
20:48:59 <peter1138> I've got the "release 1.9.1" PR with the changelog done as well.
20:49:21 <TrueBrain> yeah, my commits should have gone in that one I guess :D
20:50:32 <peter1138> The backport_1.9 stuff also included the translations, and they are backports, so it's all good.
20:50:43 <TrueBrain> yeah, I mostly did everything from a single branch :P
20:50:52 <pnda> whyyyyyyy is the VS formatting soooooo bad
20:52:27 <pnda> 6>Project not selected to build for this solution configuration
20:53:39 <pnda> lol don't even have notepad++
20:53:40 <Samu> no, to avoid those weird vs additions
20:53:56 <pnda> they aren't happening now
20:56:46 <Samu> i need an AI that makes use of random deviation in one of its settings
20:56:55 <Samu> eh, might as well be mine, brb
20:59:35 <pnda> thiiissss iss annoying meeeeeeeee
21:03:55 <peter1138> All checks passed...
21:05:13 <pnda> Whyyyy do I have to rebuild each time
21:05:40 <pnda> I seem to have a fix but that requires to have {BLACK} inbetween each part of the string
21:05:45 <TrueBrain> just don't forget to update the labels or something peter1138 :)
21:09:25 <peter1138> Was another page :D
21:10:05 <andythenorth> shall we ship the website?
21:11:55 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7489: Update: Changelog for 1.9.1 and prepare for release https://git.io/fjtak
21:12:36 <peter1138> So... I hope I got everything :p
21:13:06 <TrueBrain> hopefully LordAro likes your changelog :D
21:13:39 <milek7> TrueBrain: have you seen new emscripten build, with network? :)
21:14:16 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7489: Update: Changelog for 1.9.1 and prepare for release https://git.io/fjtak
21:14:31 <peter1138> OS2 installer. Probably doesn't even work, mind you...
21:15:08 <TrueBrain> orudge promised to look at it :P
21:16:51 <TrueBrain> that is no patch :)
21:17:16 <peter1138> Maybe it needs no patching? :LD
21:18:02 <peter1138> So pressing the "Exit" button stops the game but leaves the music running.
21:18:52 <TrueBrain> and that works without some proxy?
21:19:00 <TrueBrain> wasm allows raw network sockets?
21:19:59 <TrueBrain> how does that look like?
21:20:15 <TrueBrain> most of that patch can be upstreamed without much issue, so that is nice
21:21:30 <peter1138> Ah yeah, sockets only work to the server hosting it, right?
21:22:18 <nielsm> so you could still make a wrapper (docker?) that runs an ottd-dedicated and serves a wasm version up
21:22:33 <peter1138> So that `content.openttd.org` -> `milek7.pl` change... ought to be improved.
21:22:58 <milek7> you can use other servers, but must be websocket and https
21:23:58 <peter1138> I was just going to say, is the proxy available or some proprietory glue that you keep to yourself to be evil ;D
21:24:25 <peter1138> Did I do it right? :p
21:24:31 <milek7> ordinary websocket proxies don't work because it must redirect to proper tcp/udp from single websocket port
21:24:59 <TrueBrain> nice work milek7 :)
21:25:11 <TrueBrain> I don't see any real blockers to move this to openttd.org tbh :)
21:26:04 <milek7> sdl video output annoys me, it gets much slower at higher resolutions
21:26:08 <milek7> i need to look into it
21:27:28 <TrueBrain> I still think that would be cool to add to the site .. :D
21:27:33 <TrueBrain> but does need cloud storage tbh :P
21:28:18 <milek7> saving in browser storage isn't enough? ;p
21:29:37 <TrueBrain> what is the average savegame size, give or take?
21:30:40 <peter1138> #7489 for review? ;)
21:31:23 <TrueBrain> if someone else can look at the changelog; I am fine with the rest
21:34:11 * peter1138 sips a bit of birthday Whiksy.
21:34:43 <pnda> Ok I can actually not fix these arabic strings :(
21:35:33 <milek7> ^ so there's something that couldn't be upstreamed?
21:35:33 <pnda> also annoying I have to rebuild the solution each time
21:35:35 <nielsm> you'll probably need to LRE marks for that pnda
21:35:50 <pnda> didn't have to for the town directories though
21:35:56 <V453000> _dp_: how the hell did you get into a circumstance of making a map with that many stations? :D some script went nuts?
21:35:58 <nielsm> and huh? you shouldn't need to rebuild anything other than a single language file when not changing english.txt
21:36:19 <pnda> it skipps creating the .exe
21:36:31 <nielsm> yeah that one doesn't need updating
21:36:37 <peter1138> V453000, pretty sure it's a script, and deliberate :-)
21:36:38 <nielsm> the language data are in external files
21:37:05 <V453000> deliberate WHY :D WTF (:
21:37:23 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain approved pull request #7489: Update: Changelog for 1.9.1 and prepare for release https://git.io/fjtaS
21:37:41 <pnda> Ok I found a way to fix this. But this requires removing one part of the string.... hmm
21:37:52 <TrueBrain> to avoid waiting for hours :P
21:38:01 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7489: Update: Changelog for 1.9.1 and prepare for release https://git.io/fjtak
21:38:21 <TrueBrain> GitHub -> Releases -> Create -> 1.9.1
21:38:21 <pnda> Just having قياسي) نوع المحرك) results in that bug, if I remove the first part its normal
21:38:24 <peter1138> LordAro was here! Just hiding!
21:38:32 <TrueBrain> he didnt want to read some changelogs
21:38:36 <LordAro> peter1138: i'm always watching
21:38:56 <TrueBrain> select the right commit
21:39:49 <peter1138> So... "1.9.1" @ "Target: release/1.9" ?
21:40:31 <peter1138> Hmm, draft or "publish release" ?
21:41:29 <TrueBrain> now it is just: wait :)
21:41:32 <peter1138> DorpsGek_II, y u no notification
21:41:46 <TrueBrain> yeah ... that needs fixing
21:41:52 <peter1138> Btw, in my 15 years of OpenTTDing, this is my first release.
21:42:10 <peter1138> Oh now you tell me lol
21:42:29 <TrueBrain> it helps to tell what to do after release, at least, for me
21:42:36 <TrueBrain> I always forget something .. like IRC :P
21:42:44 <peter1138> @openttdnews. Hmm..
21:43:12 <TrueBrain> basically, poke planetmaker
21:43:15 <TrueBrain> which most likely will forget :P
21:43:48 *** DorpsGek changes topic to "0 1.9.1"
21:43:59 <TrueBrain> I knew I would fuck that up some day :P
21:44:13 <TrueBrain> meh .. this client doesnt say what the topic was
21:44:52 *** DorpsGek sets mode: +o TrueBrain_
21:44:59 *** TrueBrain_ changes topic to "1.9.1 | 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"
21:45:10 *** DorpsGek sets mode: -o TrueBrain_
21:46:39 <DorpsGek_II> [OpenTTD/OpenTTD] jmakovicka updated pull request #7484: Codechange: Sprite sorting optimization https://git.io/fjLh3
21:47:54 <peter1138> Hmm, so, news message on the website :/
21:48:36 <_dp_> just thought of ultimate color challenge: sorting socks after washing
21:48:42 <Samu> this->info != NULL also means random ais are being ignored, and they have start_date as a valid deviator
21:48:44 <_dp_> way more difficult than that xrite thing
21:48:53 * _dp_ doesn't get perfect score on socks :(
21:49:18 <DorpsGek_II> [OpenTTD/OpenTTD] jmakovicka commented on pull request #7484: Codechange: Sprite sorting optimization https://git.io/fjtVT
21:51:03 <pnda> When is the website going up for 1.9.1?
21:53:24 <DorpsGek_II> [OpenTTD/OpenTTD] michicc commented on pull request #7453: Remove AutoDeleteSmallVector and AutoFreeSmallVector https://git.io/fjtVm
21:54:27 <pnda> Also, for the arabic strings: If I have ({ArabicWord}) {ArabicWord2}, the first bracket will ALWAYS go to the end of the string, due to it not being attached to the arabic words.
21:54:50 <pnda> If I were to remove the {ArabicWord2}, the string would work fine.
21:58:24 <pnda> But I guess I'll have to this tomorrow.... So good night
22:09:15 <TrueBrain> heffer: we release 1.9.1 :) Just as a friendly poke :)
22:12:11 <milek7> it ignores dirty areas and copies always whole frame ;/
22:18:15 <DorpsGek_II> [OpenTTD/OpenTTD] stale[bot] commented on pull request #7350: Change: Add pid, file, line number & function/procedure name to debug messages https://git.io/fjtwe
22:35:17 <peter1138> Got water pipe issues :(
22:38:02 <peter1138> Great, my VM won't boot.
22:39:04 <peter1138> Alright, now it's running. Odd :p
23:07:23 <Samu> 1.9.0 was short lived :(
23:07:50 <LordAro> does everything make you sad?
23:15:33 <LordAro> random thought, should it link to latest.html, or 1.9.1.html ?
23:17:28 <planetmaker> I'm torn. But in a release posting probably to the specific version?
23:18:00 <LordAro> but equally we don't want people looking at old release posts and downloading old versions
23:18:06 <planetmaker> and... orudge already replyed very friendly
23:18:06 <peter1138> I copied the last one, so...
23:28:07 <Xaroth> "there are no bugs left"
23:31:15 <Samu> random deviance is computed twice for start_date in some occasions, but the 2nd time it is done, it no longer matters
23:32:09 <Samu> doesn't matter which value it gets after it is started
23:33:37 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7490: Add: Press ctrl to build diagonal rivers in scenario editor. https://git.io/fjtrz
23:34:42 <Samu> saving game and loading it back with a different value than that it was used to start with
23:35:25 *** Supercheese has joined #openttd
23:36:17 *** andythenorth has left #openttd
23:43:37 <peter1138> "For rail types a number of callbacks are used to define rail type graphics"
23:43:47 <peter1138> Why does the NML spec call it a "callback"? o_O
23:44:33 <Supercheese> Hysterical raisins
23:47:36 <_dp_> damn, wanted to make first 1.9.1 server but someone beat me to it %)
23:48:18 <_dp_> still the first open one :p
23:52:28 * peter1138 installs OpenTTD 1.9.1
23:52:55 <peter1138> Damn, TrueBrain is a massive hero. This CI and deployment stuff is so cool.
23:53:44 <Samu> this start_date parameter is utterly difficult to track down
23:54:16 <peter1138> _dp_, "If you're new to your servers please read" ...
23:54:38 <peter1138> Okay that map is far to boring and flat for me.
23:54:43 <Samu> I wonder, would it be a good idea to get rid of it?
23:54:49 <peter1138> And not enough water for ships to be fun.
23:55:33 <Samu> there's too many corner cases where it doesn't work as it should
23:59:27 <_dp_> peter1138, you're the first one to read that message I guess xD
23:59:29 <Samu> ok, here's a new issue: when a Random AI is set up, starting a new game will do a random deviation for start_date. Then when the AI actually starts in game, random deviation is done again for start_date + any other settings the randomly chosen AI has
23:59:48 <Samu> how to avoid doing it a 2nd time?
continue to next day ⏵
