IRC logs for #openttd on OFTC at 2021-04-24
⏴ go to previous day
01:02:04 *** Wormnest has quit IRC (Quit: Leaving)
01:42:34 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
02:15:24 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
02:55:15 *** ChanServ sets mode: +v tokai
03:02:19 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
03:14:27 *** supermop_Home_ has quit IRC (Ping timeout: 480 seconds)
04:38:24 *** snail_UES_ has quit IRC (Quit: snail_UES_)
05:02:31 *** didc has quit IRC (Remote host closed the connection)
05:24:56 *** lobstarooo has joined #openttd
05:31:34 *** lobster has quit IRC (Ping timeout: 480 seconds)
05:31:36 *** lobstarooo is now known as lobster
06:01:53 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9057: Codechange: Use std::string in console commands/aliases registration, and std::map instead our sorted linked list https://git.io/JOyl6
06:02:58 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9080: Feature: Do not default the player name to "Player" https://git.io/JOXqC
06:05:58 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 approved pull request #9057: Codechange: Use std::string in console commands/aliases registration, and std::map instead our sorted linked list https://git.io/JOy8k
06:34:04 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOyRW
06:35:51 *** Progman has joined #openttd
06:45:52 *** Speedyn has quit IRC (Read error: No route to host)
06:46:05 *** andythenorth has joined #openttd
06:56:38 *** tokai|noir has joined #openttd
06:56:38 *** ChanServ sets mode: +v tokai|noir
07:03:31 *** tokai has quit IRC (Ping timeout: 480 seconds)
07:04:59 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOyzf
07:07:57 *** andythenorth has quit IRC (Quit: andythenorth)
07:14:04 *** andythenorth has joined #openttd
07:26:59 <peter1138> Ah, when you add an override function but the other functions aren't tagged with override...
07:36:11 <peter1138> MSVC's compiler apparently don't care about missing override.
07:37:31 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #8982: Feature: text filter for the industry directory https://git.io/JOyax
07:41:52 <didac> I think override can be enforced with #pragma warning(default:4263) #pragma warning(default:4266)
07:59:25 *** didac has quit IRC (Remote host closed the connection)
08:02:55 *** sla_ro|master has joined #openttd
08:04:04 *** ChanServ sets mode: +v tokai
08:10:53 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
08:38:38 <peter1138> Oof, another Garmin update.
08:42:09 <andythenorth> hmm what shall we do today?
08:50:15 <TrueBrain> Take over the world, duhhhh
08:53:23 <peter1138> You can do all the exercise I need to do.
08:55:23 <peter1138> In return, I will do all the eating you need to do.
09:03:47 <Wolf01> pinkythenorth: I was thinking about an engineless train set, but not knowing how grfs work I don't know if it's possible. Maybe it's more like a engine only train set, with refitable cargo to all and not allowing standard rolling stock, only the same engine with low power which can be added multiple time
09:11:04 <andythenorth> isn't that NUTS?
09:11:11 <peter1138> Offsets strike again :/
09:11:20 <andythenorth> I think there's some ultimate train thing in NUTS
09:11:23 <andythenorth> slugs or something
09:12:10 <Wolf01> I remember slugs, but I can't remember if they were just a multiple unit
09:13:43 <Wolf01> I was trying to figure out a slow speed high capacity autonomous service... almost eyecandy
09:32:45 <peter1138> Oh, this code works but only because it's wrong. Hmm.
09:32:54 *** andythenorth has quit IRC (Quit: andythenorth)
09:45:25 <Wolf01> Next suggestion: make town not follow roads when expanding :P
09:49:28 *** Speedy` has joined #openttd
09:50:59 <TrueBrain> peter1138: did you just invent a label? :D
09:56:31 <TrueBrain> we have another label for similar tickets, so I changed it to that one, if you don't mind :)
09:57:18 <TrueBrain> its funny how nobody likes the default label names GitHub creates :P
09:58:03 <peter1138> I don't know what the defaults are, heh.
09:58:40 <TrueBrain> Enhancement, is one of them
09:59:08 <peter1138> We need a tag for "googling it is fine"
09:59:15 <LordAro> i think adding GS-ability to change "growth centre" would probably be an interesting idea
09:59:40 <peter1138> LordAro, yes, I was also thinking that it should perhaps be GS rather than built-in.
09:59:41 <LordAro> as for a core-game setting? probably not
10:02:52 <TrueBrain> oof, an icon for "host" in 10x10-ish ... that is hard :D
10:04:51 <_dp_> may also need to change town zone centre
10:05:40 <_dp_> and adjusting zone radiuses
10:06:36 <peter1138> My screen is too large for OpenTTD 0.4.5.
10:06:59 <_dp_> also having multiple growth centres would be cool as well
10:07:42 <LordAro> yeah, the correct behaviour for a growth centre being "out of town" might be tricky to get right, depending what you want to do
10:07:53 <LordAro> but then satellite town zones might be fun too
10:07:54 <peter1138> You could do that by letting a GS switch over time, or indeed just list multiple and have the game cycle.
10:08:14 <peter1138> (Yes, I am comparing some GUI stuff against 0.4.5)
10:08:26 <TrueBrain> it was better in those days? :D
10:08:53 <_dp_> switching over time would work weirdly with house replacement
10:09:06 <peter1138> Some of the weirdness with scaling has been around a while :-)
10:09:07 <_dp_> while it grows in one place others slowly die out
10:09:50 <peter1138> By "over time" i mean in terms of game days, not years.
10:09:59 <TrueBrain> [ 3%] Generating openttd.grf
10:10:00 <TrueBrain> Error on sprite 1729.
10:10:03 <TrueBrain> well, that is helpful
10:10:17 <LordAro> better than silently failing :p
10:10:43 <TrueBrain> only by a very very very small amount :D
10:10:45 <TrueBrain> as cmake just continues
10:10:49 <TrueBrain> which is a bug on its own, honestly
10:10:52 <_dp_> peter1138, you'll just get a mix of central zone and outskirts houses
10:10:56 <TrueBrain> I only notice it failed because ingame it was not loading :P
10:11:21 <LordAro> isn't that what glx fixed yesterday though?
10:11:30 <TrueBrain> no, he fixed validation of the executables
10:11:34 <TrueBrain> not the return codes of executing them
10:13:08 <Rubidium> #9089 is indeed quite interesting, as it already (implicitly) states that it will never be implemented by the author of the issue
10:13:33 <LordAro> yeah, strictly speaking we should just close it as a suggestion
10:15:42 <Rubidium> hmm, languages need some way to better "group" words as the sentence I wrote could ofcourse be interpreted in multiple ways
10:15:47 *** Strakaty has joined #openttd
10:16:50 <Rubidium> the point is: the author of the issue states that the actual result (the setting not being there) will be the case for all future versions
10:17:32 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9085: Fix: [CMake] Validate grfcodec and nforenum executables https://git.io/JOyQT
10:20:09 *** WormnestAndroid has quit IRC (Remote host closed the connection)
10:20:22 *** WormnestAndroid has joined #openttd
10:21:01 <LordAro> well, they're just trying to fill out the template that was not intended for this usecase
10:21:12 <TrueBrain> as it is intended for .. issues ! :D
10:21:15 <Rubidium> I've also got a eints related question; if strings get renamed, should that PR then also change the names in the translations so eints can pick those up? Or will eints not pick up those translations? And I reckon eints would reorder the strings in the next translation push, right?
10:21:38 <peter1138> Apologies, I messed it up by replying in the first case :(
10:21:49 <TrueBrain> peter1138: its fine! :)
10:22:11 <TrueBrain> what we can do btw, on GitHub, is like on OpenTTD/teams make more templates, 1 of them being: suggestions? which link to the forums, for example
10:23:01 <LordAro> might not be a bad thing to try
10:23:03 <TrueBrain> and maybe make a "Crash" template, which asks for crashdumps, and a "This is wrong" template, for other bugs
10:23:30 <TrueBrain> full of those; also with many bad ideas, your job to filter them :P :P
10:24:49 <TrueBrain> LordAro: I can do it later today, but if you have time, feel free to make a PR :) Pretty sure you will manage to word these things a lot better ;)
10:32:00 <TrueBrain> yeah, maybe not be like this ;)
10:32:08 <TrueBrain> tat hurtful .. bit more constructive doesn't hurt :)
10:32:15 <peter1138> That's always how I feel when I start gimp, tbh.
10:32:40 <_dp_> yeah, gimp ui on 4k is way worse than openttd xD
10:33:56 <TrueBrain> owh, tooltips are a lot harder to add than I expected, lol
10:34:02 <peter1138> I'm going to have to get myself (at least) one these 4K screens aren't I?
10:34:22 <peter1138> TrueBrain, they are really easy, except fixed for one widget...
10:34:36 <TrueBrain> yeah ...... and I basically made one huge custom widget
10:34:49 <TrueBrain> might still be easier to implement code to have dynamic widgets :P
10:35:12 <TrueBrain> I have no information on which rows these icons are drawn, lol
10:38:19 * peter1138 ponders a standing desk.
10:38:33 <TrueBrain> ugh, and you cannot write information during drawing ofc
10:39:21 <LordAro> peter1138: upgrade all the things!
10:42:23 *** andythenorth has joined #openttd
10:43:50 <peter1138> Dell UltraSharp UP3218K
10:44:00 <peter1138> That'll do, 7680x4320.
10:44:30 <peter1138> 5120x1440 is being labelled as 5K :(
10:44:59 <peter1138> But it's just ultra-wide 2560x1440 * 2.
10:46:03 <peter1138> LG 27MD5K-B, genuine 5K, 5120x2880.
10:47:38 <peter1138> At 7680x4320, you might need more than 4x gui scale :D
10:57:40 <Strakaty> Hi, Have anyone experience with the Bug in City Builder Script? During the game my Capital City station stop accepting Goods even goods are requested and was delivered this station before.
10:58:16 <TrueBrain> it works! But what an ugly code to make it work ..
11:02:06 <peter1138> Urgh, I wish TTF clearly defined the top and baseline.
11:02:35 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 updated pull request #8546: Doc: provide a visual representation of the company colour ranges https://git.io/JLhCP
11:04:11 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #8546: Doc: provide a visual representation of the company colour ranges https://git.io/JOybp
11:09:07 <FLHerne> Strakaty: It's possible the buildings in the catchment area have changed
11:10:08 <FLHerne> Strakaty: When you use the 'info' tool on a building, it'll say something like "Cargo accepted: (3/8 Passengers, 1/8 Goods)"
11:10:41 <FLHerne> Your station will only accept a cargo if the buildings in catchment sum to at least 8/8
11:11:42 <FLHerne> Town buildings are demolished and replaced over time, so acceptance can change if it's marginal
11:19:17 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8546: Doc: provide a visual representation of the company colour ranges https://git.io/JLhCP
11:19:21 <_dp_> huh? since when does crown give a bias toward males?
11:20:17 <TrueBrain> kings have these kind of crowns
11:20:21 <TrueBrain> queens have different types
11:21:11 <LordAro> i think that's nonsense
11:21:44 <TrueBrain> if everyone thinks that, my problem with it is not important
11:23:45 <TrueBrain> funny, it strongly depends on the country (read: cultural) if and what the difference is
11:24:20 <TrueBrain> there are countries where there is no difference between the genders
11:24:22 <TrueBrain> and ones with clear differences
11:30:57 <TrueBrain> right, lets not delve too much in the cultural history of crowns .. rabbit hole is deep enough :D
11:31:59 *** HerzogDeXtEr has joined #openttd
11:32:20 <LordAro> i'm sure you can find some nutjob on tumblr complaining about them if you look hard enough
11:32:40 <TrueBrain> I think you are to dismissive of this ;)
11:32:54 <TrueBrain> some countries really have a strong difference between what a king has and what a queen has
11:33:04 <TrueBrain> I don't mind if we say: we are fine with this, don't get me wrong
11:33:13 <TrueBrain> but it doesn't mean that for some countries there isn't a gender bias
11:33:25 <TrueBrain> but, ironically, I found countries that have this type of crown for kings
11:33:31 <TrueBrain> but also countries where it is for queens
11:33:56 <TrueBrain> so I suspect it strongly depends on your country if this is a problem at all :)
11:39:52 <peter1138> I'm sure there'll be someone objecting to the use of a crown at all...
11:40:50 <Strakaty> FLHerne : weird is , my station cover such building with Goods required, Even cargo car transport nearby and another train station in City doesn't accept Goods .
11:41:14 <TrueBrain> peter1138: in general, we can safely say that someone will object to <insert icon we use> :D
11:41:28 <_dp_> TrueBrain, funnily in chess it's a queen's crown
11:42:12 <TrueBrain> yeah .. in general, this is funny
11:42:24 *** WormnestAndroid has quit IRC (Remote host closed the connection)
11:42:27 <TrueBrain> some countries point crown == king, rounded crown == queen
11:42:32 <TrueBrain> and some countries exactly the opposite
11:42:44 <TrueBrain> there seems to be no agreement between countries on these things
11:43:24 <TrueBrain> which highlights my point, that the use of a crown can be offensive (as we ship worldwide :P)
11:43:55 <TrueBrain> but again, I am personally fine with it, mostly as the crown is already used in OpenTTD
11:44:11 <TrueBrain> drawing the round one is harder, I would guess :P
11:44:27 <LordAro> TrueBrain: if you want to go that route, then you run into all sorts of existing stuff
11:44:30 *** WormnestAndroid has joined #openttd
11:44:41 <TrueBrain> LordAro: doesn't mean we can dismiss it :)
11:44:50 <_dp_> actually, may not be harder, I almost made one when trying to draw another xD
11:44:53 <TrueBrain> cultural sensitivity is something that exists ;)
11:45:13 <LordAro> for sure, but i have very little time for making issues where there are none
11:45:17 <TrueBrain> and you see we already removed a few pieces
11:45:35 <TrueBrain> LordAro: that is your choice; just don't make it my problem :) I just mention it is something I noticed and I would prefer something else instead
11:45:57 <TrueBrain> just a preference :) No need to make a big deal out of it :)
11:46:03 <TrueBrain> if we are all fine with a crown, lets go with it :)
11:48:43 <TrueBrain> okay, I worded that terribly. Let me try again: I just wanted to mention it, and if we as OpenTTD are fine with it, lets not talk about it any further
11:48:50 <TrueBrain> I didn't want to make a big deal out of it :)
12:06:53 *** frosch123 has joined #openttd
12:12:23 <frosch123> Rubidium: eints uses stringids as keys. if you change stringids, eints thinks the old string was deleted, and a new unrelated is added
12:12:48 <frosch123> if yuo want to keep the translations, you have to change the stringids in the translations as well
12:13:17 <frosch123> otoh, if you want to delete translations, you can change the stringid to make eints drop the translations
12:14:01 <frosch123> (deletings translations while keeping the stringid does nothing, since eints will consider the old translation better than no translation, and just readd it)
12:14:59 <frosch123> wrt. reordering strings. eints will always sort all translations the same way as english.txt
12:15:10 <frosch123> so it's enough if you change english.txt, eints will adapt the translations
12:16:39 <LordAro> frosch123: eints surely can't distiguish between a delete & add vs a move though?
12:16:58 <LordAro> unless it's a pure reorder without any renaming, i suppose
12:17:04 <LordAro> but why would you do that?
12:17:42 <Rubidium> lovely, that was roughly what I expected. I just did not want to cause major mayhem for "fixing" #8986 the long way
12:17:43 <frosch123> "move" as in "reorder strings" is fine. "move" and in "change stringid" is "delete string + add unrelated new string"
12:17:51 <andythenorth> _date seems to be used in files without declaring it as a var?
12:18:22 <LordAro> andythenorth: it's an extern
12:18:52 <LordAro> more specifically, it'll be declared as an extern in some header file
12:18:52 <frosch123> Rubidium: ah yes, i explicitly choosed not to update the stringid, when i moved the setting to the mapgen gui :)
12:21:20 <frosch123> ah crown is biased towards males? what do queens and pricesses wear?
12:21:36 <andythenorth> subject is silly
12:21:46 <frosch123> but you have a queen
12:22:04 <andythenorth> I sign contracts with The Crown
12:22:11 <Rubidium> but isn't she also the king?
12:22:23 <andythenorth> 'this is complicated'
12:22:30 <TrueBrain> frosch123: not "crowns", that type of crown :D
12:23:29 <frosch123> well, i don't live in a monarchy, i would not have recognised that icon as "male crown"
12:23:53 <TrueBrain> yeah .. as we said in chat a bit above, it is even more funny: depending on your country, the "male crown" looks different
12:24:04 <TrueBrain> pointy crowns vs curved crowns, so to say
12:24:12 <TrueBrain> it is a complete mess in the world :P
12:24:29 <TrueBrain> so I guess it depends on where you live if you get offended or not :D
12:27:28 <andythenorth> I haven't made a test newgrf yet, but eh
12:27:39 <TrueBrain> anyway, I think I did everything for the Network GUI I can do ..
12:27:48 <andythenorth> ^^ do I break any style rules?
12:32:50 <TrueBrain> LordAro: they have those 2 types of crowns, not?
12:32:56 * andythenorth wonders if that date will return as days
12:32:59 <TrueBrain> one pointy, one curly?
12:33:00 <andythenorth> or an actual date
12:33:18 <LordAro> andythenorth: probably the date struct, which i can't imagine will work all that well
12:33:31 <Rubidium> andythenorth: style wise it looks okay to me. You could add [NewGRF] between the : and vehicle in the commit message, and maybe consider using last_station_departure_date so the comments in vehicle_base.h can stay aligned
12:33:52 <Rubidium> andythenorth: it will return the days since 1-1-0 (it's an integer)
12:34:05 <andythenorth> same as last depot visit
12:34:56 <peter1138> Whew, nasty hack to determine where the ttf sprite offset should be.
12:35:42 <TrueBrain> peter1138: owh no you didn't?
12:35:55 <peter1138> What you think it is? :D
12:36:27 <TrueBrain> Confirmed OpenTTD 1.12 needs RTX!
12:36:42 <Rubidium> andythenorth: maybe you want to set the initial date in the constructor to INVALID_DATE, so there is a distinction between a vehicle leaving at 1-1-0 and a new vehicle
12:37:06 <peter1138> I look at the outline of a particular letter and go "err, no mate, that pixel there is the top"
12:37:32 <peter1138> THEN I have to reload the font in a different size to adjust for the difference.
12:37:46 <peter1138> And then the top of the font lines up with the top of the original font.
12:38:10 <TrueBrain> but what can you do?
12:38:12 <TrueBrain> we cannot fix all the fonts
12:38:28 <peter1138> If there was some sort of standard for ttfs... :/
12:38:35 <TrueBrain> LordAro: were you looking into issue templates, or shall I?
12:38:48 <LordAro> TrueBrain: i'm about to disappear, you go for it
12:38:54 <peter1138> The idea of looking at a particular character is actually what Adobe does (did?).
12:39:15 <TrueBrain> LordAro: will do; and I will await your return :D
12:46:29 *** virtualrandomnumber has joined #openttd
12:46:36 <andythenorth> hmm this var only makes sense on lead engine
12:46:52 *** virtualrandomnumber has quit IRC ()
12:47:48 <TrueBrain> sneaky Zuu , tssk :P
12:48:04 <TrueBrain> he just knew I would revert the PR after merging as it was no longer true :P :P
12:49:16 <Zuu> James found a bug in my PR. :-)
12:49:24 <TrueBrain> and I didn't, that is the worst part :P
12:49:26 <peter1138> Oh god damnit. My motivation to fix this alignment issue was mostly the drop down settings list with the ticks.
12:49:33 <peter1138> As the tick does not align with the text.
12:49:45 <peter1138> This change also moves the tick up, so it's also no longer aligned. :/
12:50:25 <Zuu> TrueBrain: I think I can trust you not to put me back as you are also in the inactive bracket :-)
12:51:19 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOStT
12:51:25 <TrueBrain> Zuu: SSSTTTTTT, don't tell anyone :)
12:52:31 <andythenorth> var 4E seems to work
12:54:32 <andythenorth> should I set var 4E somehow for all vehicles in the consist when departing station?
12:54:59 <andythenorth> for newgrf purposes it's irrelevant parent/self, unless we add shunting
12:55:43 <Rubidium> I'd do it the same as for servicing
12:56:04 <FLHerne> TrueBrain: There's a lot of "Player Transport (Company 1)" in those screenshots
12:56:22 <FLHerne> Would "Player Transport (#1)" be any better? It seems obvious in context
12:56:29 <FLHerne> Or tbh just leave it out...
12:56:46 <peter1138> Also, re that newgrf desync, can more stats be refreshed once cargo is finished loading (except length :D)
12:56:54 <FLHerne> I'm not sure why anyone needs to know, except server admins who can get it by console
12:56:57 <TrueBrain> FLHerne: I picked the default way we present company strings
12:57:07 <FLHerne> *except people using console commands
12:57:20 <FLHerne> given you added buttons for most of the things
12:57:41 <TrueBrain> I personally think we can leave it out everywhere, tbh
12:57:49 <TrueBrain> (the "(Company 1)" stuff)
12:59:15 <TrueBrain> and yeah, over time, I hope this makes the console unused by most admins
12:59:28 <TrueBrain> as I think it is really shitty you ahve to use the console to do basic administrative jobs :P
12:59:48 <TrueBrain> (mainly as I can never remember .. delete_company? reset_company? company_reset? )
13:01:18 <TrueBrain> still cracks me up :)
13:02:57 *** lobstarooo has joined #openttd
13:03:22 <peter1138> Hmm, I guess I should restart my Minecraft world. Incompatible world changes indeed :/
13:05:26 <andythenorth> hmm I need some loop over vehicles
13:05:37 <andythenorth> but only if it's a train consist or articulated RV
13:05:51 <peter1138> Yes, to be expected with snapshots. But nobody actually expects it :D
13:06:25 <TrueBrain> I have to say, (Company 1) etc made it more clear they were companies
13:07:27 <peter1138> Damn, a 6x8 abandoned portal near my new spawn.
13:08:29 *** lobster has quit IRC (Ping timeout: 480 seconds)
13:08:38 *** lobstarooo is now known as lobster
13:09:03 <FLHerne> I see your point, but it does look less cluttered
13:09:11 <TrueBrain> I agree .. so I am +0 on both :D
13:09:24 <TrueBrain> what do others think? With or without "(Company 1)" ?
13:15:33 <DorpsGek> [OpenTTD/website] TrueBrain opened pull request #209: Fix: [Actions] prevent multiple deployments at the same time https://git.io/JOSYR
13:17:49 <TrueBrain> FLHerne: lets go without "(Company 1)" for now
13:17:52 <TrueBrain> less text == better
13:20:01 <DorpsGek> [OpenTTD/OpenTTD] glx22 merged pull request #9057: Codechange: Use std::string in console commands/aliases registration, and std::map instead our sorted linked list https://git.io/JOu1a
13:24:32 <DorpsGek> [OpenTTD/BaNaNaS] TrueBrain opened pull request #94: Fix: [Actions] prevent multiple reloads at the same time https://git.io/JOS3a
13:24:41 <TrueBrain> I have to add ^^ in many repos, but these 2 are the most important ones :)
13:25:18 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOS3i
13:28:24 <andythenorth> hmm so what is a vehicle in the game?
13:32:50 <andythenorth> sometimes v->foo works and sometimes v->foo doesn't?
13:33:00 <andythenorth> or rather, sometimes it's a vehicle, and sometimes it's a consist?
13:33:19 <FLHerne> TrueBrain: How about 'New Company' as orange button, but in the list under spectators?
13:33:44 <TrueBrain> FLHerne: a button really looked odd, how ever I tried
13:33:54 <TrueBrain> but maybe you can mock something that does look nice?
13:34:11 <peter1138> v is always a vehicle. Sometimes it's the first vehicle in the chain.
13:34:45 <FLHerne> I think it might be font-dependent :-/
13:35:05 <TrueBrain> what we have currently looks nice style-wise, but I am not sure users will be able to find it
13:35:08 <TrueBrain> that is my only real issue here
13:35:11 <FLHerne> In pixel-font, I think a button would be the same size as the icon ones
13:35:24 <FLHerne> But in yours they end up taller
13:36:20 <FLHerne> Which version is 'currently'?
13:36:54 <FLHerne> I really don't like the "(New company)" in the list, that doesn't look much like an action
13:37:13 <TrueBrain> maybe we should name it (empty company)
13:37:52 <FLHerne> Too many screenshots, and the last one doesn't have 'new company' visible at all, so I'm not sure which is actually current
13:37:58 <FLHerne> I could compile it :p
13:38:01 <TrueBrain> the ones in the PR starting thing
13:38:07 <TrueBrain> are current (if not noted as outdated)
13:38:20 <_dp_> company id's are kinda useful for admins
13:38:48 <TrueBrain> FLHerne: so you can ignore all the replies, and just look at the top post, basically :)
13:38:50 <_dp_> as there can be several companies named the same
13:39:34 <_dp_> I think there are even some extra ids somewhere in cmclient for that
13:40:06 <TrueBrain> well, with most admin actions moving to the admin button, I guess for admins that do need it, it is fine to look in the company list from the toolbar?
13:40:24 <TrueBrain> not sure that having information that is only used by a very small portion of the players is useful to have always visible .. but I really do not know if that is true :D
13:40:59 <_dp_> not sure how will it work with new ui as well
13:42:10 <_dp_> for me personally the biggest issue with admin part of that ui is that admin is not a "host" on dedicated server
13:42:31 <_dp_> so he won't even get the ui
13:42:36 <michi_cc> Ah, fun times today. There's an anarcho leftist punk "concert" right in at the corner where I live that's supposed to go until the evening.
13:42:36 <TrueBrain> no .. rcon support needs to be added :)
13:43:08 <_dp_> dunno, I don't like rcon in general
13:43:42 <_dp_> but I just have a server console on the website so don't rly need it
13:47:16 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSZL
13:48:19 <andythenorth> is there anyway I can delegate v->foo to lead of consist?
13:48:25 <andythenorth> instead of setting it for every vehicle?
13:48:41 <frosch123> andythenorth: use the PARENT scope in the newgrf
13:48:42 * andythenorth has run out of copy-pasteable vars to copy
13:48:52 <andythenorth> frosch123 I mean in openttd
13:48:53 <michi_cc> andythenorth: I have a consist branch, but it's not finished at all :)
13:49:02 <frosch123> andythenorth: don't invent problems, that don't exist
13:50:24 <FLHerne> "as there can be several companies named the same"
13:50:47 <FLHerne> That sounds like a Bad Feature
13:51:20 <_dp_> there can even be clients with the same name ;)
13:51:26 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSZM
13:56:16 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSnC
13:56:59 <DorpsGek> [OpenTTD/BaNaNaS] glx22 approved pull request #94: Fix: [Actions] prevent multiple reloads at the same time https://git.io/JOSnz
13:57:10 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSnV
13:58:07 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSnS
13:58:28 <DorpsGek> [OpenTTD/BaNaNaS] TrueBrain merged pull request #94: Fix: [Actions] prevent multiple reloads at the same time https://git.io/JOS3a
13:59:25 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOSnx
13:59:27 <TrueBrain> that should be the end of the annoying emails ..
14:00:07 <glx> ok, now let's redo the grfcodec fix
14:00:10 <frosch123> hmm, how to translate "DAU" :p
14:00:35 <michi_cc> If I knew a proper English acronym, I'd have used it :)
14:00:53 <DorpsGek> [OpenTTD/website] glx22 approved pull request #209: Fix: [Actions] prevent multiple deployments at the same time https://git.io/JOScL
14:01:02 <DorpsGek> [OpenTTD/website] TrueBrain merged pull request #209: Fix: [Actions] prevent multiple deployments at the same time https://git.io/JOSYR
14:01:24 <frosch123> how about MCD - maximum credible douchbag
14:02:18 <andythenorth> how do I mark a PR draft?
14:02:26 <Rubidium> "stupidest imaginable user"?
14:02:41 <TrueBrain> andythenorth: dropdown on the create pullrequest button
14:02:45 <DorpsGek> [OpenTTD/OpenTTD] andythenorth opened pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOScW
14:02:53 <DorpsGek> [OpenTTD/OpenTTD] frosch123 approved pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOScB
14:03:41 <andythenorth> meh, I can't computer properly :P
14:04:07 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOScr
14:04:44 * andythenorth finds the subtle 'make draft' button
14:05:18 <michi_cc> The create pull request button has a small dropdown to directly create a draft as well.
14:05:38 <andythenorth> frosch123 vehicle_sl.cpp?
14:06:27 <Rubidium> I guess BFU (brain free user) comes close
14:08:25 *** jottyfan has joined #openttd
14:08:47 <andythenorth> dates seem to be SLE_INT32?
14:09:45 <peter1138> What are you doing with date of last station departure?
14:10:52 <glx> hmm current CreateGRF generates new grf in source, then copy it to bin dir, meaning it deletes source file on cleaning, maybe it should generate in bin and copy to source instead
14:11:20 <frosch123> TrueBrain: funny, there is no longer a "speak to all" button
14:11:27 *** tokai|noir has joined #openttd
14:11:27 *** ChanServ sets mode: +v tokai|noir
14:11:30 <andythenorth> oof I need current SLV version?
14:11:39 <frosch123> though everyone uses the hotkey probably
14:11:39 <peter1138> andythenorth, for what?
14:11:54 <andythenorth> not sure, cargo culting saveload code here
14:12:15 <peter1138> You add a new version, and use that.
14:12:23 <TrueBrain> frosch123: lol, indeed ..
14:12:30 <TrueBrain> I assume everyone knows "enter"
14:12:33 <TrueBrain> but .. that is an assumption
14:13:27 * andythenorth looking for actual current SLV number to add 1 to
14:13:50 <peter1138> There is no number.
14:13:59 <Rubidium> so the window also needs a megaphone icon for speak to all?
14:14:13 <glx> what are you trying to do ? Just add a new SLV_ to the list if needed
14:14:20 <peter1138> You just add your SLV_ identifier to the end of the enum (before SL_MAX_VERISON) and that's it.
14:15:37 <TrueBrain> Rubidium: strictly seen, I would say so ... but I am so done with this GUI, I don't want to :P
14:15:56 <andythenorth> and the identifier I need would be...?
14:17:38 <Rubidium> FLHerne: about the company names being the same. What happens is that companies are made unnamed, and then the first thing clients do is rename the company to "<client name> Transport". That rename prevents duplicate names, and since TB was testing creating companies it was the same player making companies and each and every company getting renamed to the same (and failing to rename). So practically
14:17:44 <Rubidium> it's not that buggy, just a side effect of the same client making multiple companies
14:18:23 *** tokai has quit IRC (Ping timeout: 480 seconds)
14:18:25 <FLHerne> I confess to having no idea what the point is
14:19:31 <FLHerne> Rubidium: Well, I'd think a server admin's concern would be clients using duplicate client/company names to cause problems
14:19:58 <frosch123> uh, what, tb's branch does not compile for me
14:20:03 <FLHerne> "Wossname Transport griefed me!" <ban> "no, the *other* Wossname Transport"
14:20:06 <DorpsGek> [OpenTTD/OpenTTD] andythenorth updated pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOScW
14:20:28 <TrueBrain> frosch123: huh? How?! :D
14:20:39 <Rubidium> FLHerne: client names cannot be duplicate, and neither can company names except for the company "Unnamed"
14:20:41 * andythenorth encounters "FixupTrainLengths()'
14:20:48 * andythenorth steps away from that
14:21:14 <FLHerne> <_dp_> there can even be clients with the same name ;)
14:21:28 <FLHerne> ^ did you fix that, or is _dp_ mistaken, or am I misreading stuff?
14:21:38 <TrueBrain> game should add #1, #2 etc behind duplicated names
14:21:41 <TrueBrain> it does that for me at least :D
14:22:20 <TrueBrain> owh boy, I broke something with rebasing .. start multiplayer button doesn't work anymore :D
14:22:22 <TrueBrain> frosch123: let me see
14:22:28 <peter1138> andythenorth, pretty sure you didn't add that variable back in SLV_31.
14:22:40 <andythenorth> no I just picked a number
14:22:46 <TrueBrain> frosch123: what kind of compiler is this? :)
14:23:00 <frosch123> g++ (Debian 8.3.0-6) 8.3.0
14:23:15 <peter1138> andythenorth, saveload.h:328 ... you add you own identifier to the end.
14:23:21 <TrueBrain> so the template doesn't work with 8.3 but does with 9.3 ?
14:23:26 <frosch123> TrueBrain: looks like it does not have enough CTAD support
14:23:30 <TrueBrain> odd .. any idea why it breaks?
14:23:57 <frosch123> class-template-argument-deduction
14:24:12 <frosch123> allow you to skip explicit template parameters in constructors
14:24:12 <Rubidium> FLHerne: _dp_ is probably mistaken. In the join packet the client name is made unique, and if it cannot the client is actually rejected. In the set name packet the name is not changed when the new client name is duplicate. The only thing that could be the case is that the logic to determine a duplicate name is flawed
14:24:14 <frosch123> which is what you do
14:24:32 <TrueBrain> frosch123: what do I need to do to solve it? :D
14:24:47 <glx> oh dependency are wrong for grf commands
14:25:39 <TrueBrain> Rubidium: I shouldn't have forced your change to the company_colur_indexes.html .. it contains trailing whitespaces :P
14:26:30 <_dp_> well, they can definitely be same with a patched server
14:26:42 <_dp_> but can probably be reproduced without patch as well in some cases
14:27:16 <FLHerne> It's not silliness with visually-identical Unicode chars or so?
14:27:29 <TrueBrain> Rubidium: okay, found another bug, and I don't have suggestions .. when I start the game via WSL without config, I get an error hardware acceleration is not supported
14:27:33 <TrueBrain> this is a critical error
14:27:39 <TrueBrain> I just always leave it on, as I see it all the time
14:27:43 <TrueBrain> now when I press Start server
14:27:46 <TrueBrain> Playername is empty
14:27:50 <TrueBrain> but that error is not being shown
14:27:54 <TrueBrain> as .. the critical error is still active
14:27:57 <TrueBrain> so I got really confused :P
14:28:01 <TrueBrain> NO clue if it needs solving
14:28:50 <andythenorth> adding a vehicle property will also need migration when loading old savegames? Or that's handled automatically?
14:29:15 * andythenorth loads a savegame, empiricism
14:30:04 <Rubidium> TrueBrain: the only thing to "solve" that is making that error WL_CRITICAL, but then it will block any other errors from building if you keep it open which might be considered a bug too
14:30:20 <TrueBrain> I was thinking putting the focus on the player name edit box
14:30:28 <TrueBrain> that might be already a hint enough
14:30:31 <TrueBrain> but I dunno, when I wrote it out
14:30:33 <TrueBrain> it felt like a me issue
14:30:36 <TrueBrain> not an OpenTTD issue :D
14:31:47 <TrueBrain> frosch123: anyway, back to your bug .. I am trying to understand .. what do you mean with "skip"?
14:31:52 <TrueBrain> I use T in the ctor, not?
14:32:08 <Rubidium> oh, even making it WL_CRITICAL does not help. The only solution is making the other issue non-critical, so probably remove the OpenGL video drivers then ;)
14:32:16 <TrueBrain> or does it need Button<CompanyID>(..) ?
14:35:32 <TrueBrain> yeah .. was hoping that was not needed, but if old compilers can't compile OpenTTD otherwise ... :D
14:35:34 <frosch123> the problem is that two parameters of the constructor (id and proc) depend on the template parameter
14:35:52 <frosch123> so it's too hard for my compiler to figure out what "T" is
14:36:00 <TrueBrain> "using" .. never knew that existed
14:36:07 <TrueBrain> I would have typedef'd it :P
14:36:20 <frosch123> TrueBrain: cppreference lists gcc7 as enough for CTAD :) you hit a complicated case
14:36:57 <DorpsGek> [OpenTTD/OpenTTD] andythenorth updated pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOScW
14:37:03 <TrueBrain> well, I am happy you took the time to compile it
14:37:27 <frosch123> iirc "preview" does not support multiplayer
14:37:32 <FLHerne> TrueBrain: Why is hwaccel not being supported a critical error?
14:37:39 <FLHerne> It falls back anyway, no?
14:37:51 <TrueBrain> FLHerne: why do you think I have those answers? :)
14:37:58 <TrueBrain> frosch123: only localhost :D
14:38:32 <FLHerne> TrueBrain: You have all the answers
14:38:42 <FLHerne> Even the ones you don't know about yet
14:38:48 <TrueBrain> hate to disappoint :D :D :)
14:38:57 <TrueBrain> either way, I think it should be a critical anyway
14:39:10 <TrueBrain> otherwise it might not be visible to the user
14:39:20 <FLHerne> Now, if you come over here so I can slice your brain very thinly so my new device can scan it...
14:40:34 <DorpsGek> [OpenTTD/nml] andythenorth opened pull request #211: Feature: support for vehicle var 4E, date_of_last_station_departure https://git.io/JOSBr
14:41:17 <TrueBrain> right .. that should fix all the feedback
14:42:02 <FLHerne> quick! more feedback!
14:42:04 <frosch123> hmm, i found a bug, but rb already pointed it out, the conversion is marked as resolved, now i have to find the conversation ... :p
14:42:31 <TrueBrain> frosch123: if it was a reason thing he pointed out .. rebase :D
14:42:51 <frosch123> why do you enable the "default" button for server and player name?
14:43:01 <TrueBrain> frosch123: rebase :)
14:43:14 <TrueBrain> dunno, pick a word :P
14:45:02 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOSRZ
14:48:48 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOSRH
14:49:05 <andythenorth> I wondered about desyncs
14:49:25 <andythenorth> in the same way as I wonder about gruffaloes
14:49:38 <peter1138> What do you intend to change based on this var?
14:50:57 <andythenorth> but what might other people change?
14:51:07 * andythenorth BIAB, chores to do
14:51:10 *** andythenorth has quit IRC (Quit: andythenorth)
14:51:11 <TrueBrain> why did I keep "My Server", lol
14:51:15 <TrueBrain> normally I add TODOs around those blobs
14:52:04 <frosch123> lol, if i do the same for players name, i get scolded with a big red window
14:52:06 <TrueBrain> ideally I don't allow empty strings ..
14:52:21 <TrueBrain> Rubidium: can't you copy what you did for client name? :D
14:53:14 <frosch123> aw, your buttons do not animate/lower :(
14:53:25 <frosch123> so much gui stuff :p
14:54:09 <TrueBrain> I also found out OpenTTD reacts on mouse-down
14:54:14 <TrueBrain> where modern shit reacts on mouse-up
14:54:17 <TrueBrain> which made me laugh a bit
14:54:32 <Rubidium> TrueBrain: I guess I could
14:54:54 <TrueBrain> Rubidium: I was kinda joking, tbh :P
14:54:59 <Rubidium> though I might have found something that would be more pressing
14:55:06 <TrueBrain> but of course I am not going to stop you either .. so I guess I was partly joking :P
14:55:16 <TrueBrain> SERVER_INFO to TCP? :D
14:56:06 <TrueBrain> hmm .. guess the error if you edit to an empty player-name is wrong now
14:56:10 <TrueBrain> that only works for that one specific window
14:56:40 <TrueBrain> but I really should stop adding more stuff to this PR
14:56:41 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOS0S
14:57:06 <TrueBrain> too much gold-plating is not going to help anyone :D
14:57:14 <frosch123> TrueBrain: missing "set server password" button intended?
14:57:40 <TrueBrain> "intended" in the sense I really tried not to do too much in this PR, yes
14:57:44 <TrueBrain> I will add it to the "future work"
14:58:06 <TrueBrain> no, nothing to be sorry about
14:58:14 <frosch123> i expected a "presence api retires all passwords" :p
14:58:16 <TrueBrain> this is what I am constantly battling .. tons of extra things we can do
14:58:33 <TrueBrain> nah, server passwords are always good to have
14:59:35 <TrueBrain> but it is so hard to limit myself not to add "everything" in this one PR :)
14:59:56 <TrueBrain> hopefully after this PR others feel inspired to also add stuff :)
15:01:32 <frosch123> i didn't know the host can join any company without password
15:02:11 <TrueBrain> I initially was looking at it as a bug
15:02:16 <TrueBrain> and was heavily confused
15:02:20 <TrueBrain> but .. it is intentional
15:02:29 <TrueBrain> I think it is really bad
15:02:46 <TrueBrain> oops, wrongly phrased: I really think it is bad
15:03:22 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOStT
15:09:56 <frosch123> haha, found a line that exposes you as a C programmer :) (no bug, just a stylistic choice, that no C++ programmer would choose)
15:10:53 <frosch123> uint DrawButtons(uint x, uint y, const std::vector<std::unique_ptr<ButtonCommon>> *buttons) const <- i would have passed "buttons" as reference
15:11:24 <frosch123> for (auto &button : *buttons) { <- i noticed, because that looked usual
15:11:42 <TrueBrain> well, it is by reference now :P
15:11:49 <TrueBrain> but yeah, I kinda had a hard time getting that the way I wanted ..
15:12:00 <TrueBrain> was before the unique_ptr, so maybe it works now
15:12:14 <TrueBrain> but I couldn't get out of the vector that what I needed for the compiler to shut the fuck up :D
15:12:30 <frosch123> anyway, your map/unique_ptr usage looks fine, so we managed to drag you somewhat onto the right side :)
15:12:34 <TrueBrain> please feel free to tell me how a C++ programmer would do it, as I kinda want to learn that a bit more :)
15:13:07 <TrueBrain> C has pointers which are a nightmare to learn for most people .. C++ stacked "&" on top of that
15:13:14 <TrueBrain> which is really confusing for a C programmer
15:13:24 <TrueBrain> most of the time I am completely unsure if it copies the object or not :D
15:13:43 <TrueBrain> I should read about it, but I am lazy :(
15:17:29 <frosch123> OnClick does not work for RTL languages
15:17:39 <frosch123> i have to click left of the button, to trigger them
15:17:44 <TrueBrain> ghehe, I only have checked the visual aspect :D
15:20:58 <frosch123> GHA wants you to add "override" to OnMouseOver
15:23:49 <frosch123> looks awesome though :)
15:24:51 <TrueBrain> src\network\network_gui.cpp(1905): warning C4146: unary minus operator applied to unsigned type, result still unsigned
15:24:53 <TrueBrain> only windows ... :P
15:24:54 <frosch123> just need to find someone to deal with the rest of the feature creep :)
15:25:19 <TrueBrain> I mean, wtf Windows .. why does -(uint)1 remain uint?
15:26:25 <frosch123> oh shit... my 2 year-old neighbour got a toy that makes noise
15:26:29 <frosch123> not annoying at all....
15:26:46 <TrueBrain> go next-door and take it from him/her
15:26:51 <TrueBrain> lets see what noise that makes :S
15:28:43 <TrueBrain> owh, RTL doesn't take the scrollbar into account
15:29:16 <TrueBrain> do you think I care it is well defined or not? :P
15:29:25 <TrueBrain> it is just really stupid :)
15:29:39 <TrueBrain> in what earth do I expect "int a = -(uint)2" to be "2" ?
15:30:22 <michi_cc> It's not 2, it's 4294967294.
15:30:39 <frosch123> it probably even works :) the 4294967294 is casted by the assignment back to int, so -2
15:30:47 <TrueBrain> fair enough, but even worse :P
15:31:03 <michi_cc> Unary unsigned minus is defined as 2^b
15:31:03 <michi_cc> -a with b == number of bits in result.
15:31:16 *** andythenorth has joined #openttd
15:31:21 <michi_cc> Where did that line break come from???
15:32:31 <michi_cc> To be more "interesting", -(ushort)2 is actually -2 (signed), as types shorted than int are promoted to int before the operator is applied.
15:32:49 <TrueBrain> ..... fuck this shit, I am out :P
15:32:57 <frosch123> we should switch to rust :)
15:35:01 <Rubidium> what defines do we need for that? ;)
15:35:01 <TrueBrain> switching is a bad idea :P
15:35:10 <TrueBrain> but I did new projects in Rust, and it is pretty nice :)
15:35:36 <Rubidium> I remember someone (jokingly) proposing #define BEGIN { and #define END }
15:35:52 <andythenorth> do the desyncs arise from the vars, or fundamentally because cb 36 is unwise?
15:36:19 <frosch123> btw, i actually do not know how rust does arithmetics with different integer types :)
15:36:19 <andythenorth> some vars vary across clients, and some don't?
15:37:22 <frosch123> andythenorth: desyncs occur from optimisation, which ottd does
15:37:43 <frosch123> ottd does not reevaluate all callbacks every tick, and does some assupotions when results may change
15:37:44 <michi_cc> The whole reason for all the signed/unsigned stuff is of course that C does not specify how signed integers are stored (i.e. two's complement is not required).
15:37:52 <frosch123> if the newgrf disagrees with that, it desyncs
15:38:21 <TrueBrain> being able to navigate in RTL because you opened all the UIs so often .. lol
15:38:34 <michi_cc> If we bump up to C++20, we'd be actually guaranteed to get two's complement numbers :)
15:38:42 <frosch123> some patchpacks fix the desync by disallowing the newgrf to change values, though i think that's more like "hiding the bug" than "fixing the bug"
15:39:13 <frosch123> michi_cc: that does not change the -(uint)2 thing
15:39:27 <frosch123> TrueBrain: yeah, you have to learn that :)
15:39:34 <andythenorth> "adventures in extending the API"
15:39:42 <frosch123> most importantly you have to learn to find the game options, to switch back :)
15:40:28 <michi_cc> It doesn't, but it is the historical reason for all the signed/unsigned differences.
15:40:47 <TrueBrain> many small things "wrong" there
15:41:07 <frosch123> there is just this big blob of sorting/fitlering at the top
15:41:16 <frosch123> every button has some different magic
15:41:16 <andythenorth> that gui is horrible
15:41:24 <andythenorth> but I don't know how to make it better
15:41:42 <andythenorth> the purchase menu has to remain small for small screens?
15:41:48 <TrueBrain> fixed RTL I think :)
15:42:03 <TrueBrain> frosch123: it also tries to do too much
15:42:11 <TrueBrain> like "hide" and "rename" .. next to "build vehicle"
15:42:35 <TrueBrain> for example, one could add an "admin" icon on each row, so you can do that per train
15:42:41 <frosch123> the is also a "build+refit" button sometimes
15:43:01 <andythenorth> so how do I prevent the desyncs?
15:43:07 <andythenorth> limit the scope of the var or something?
15:43:09 <frosch123> andythenorth: you don't
15:43:31 <TrueBrain> and on the top, yeah .. too much going on
15:43:50 <TrueBrain> but the text at the bottom too .. just random letters without any visual separation
15:43:56 <TrueBrain> making it into 2 columns would already help
15:44:00 <TrueBrain> (for Power and Life)
15:44:06 <TrueBrain> or just, you know, not do it
15:44:09 <TrueBrain> and put it below each other
15:44:32 <andythenorth> the main reason I gave up on vehicle variants / liveries etc is that there's no place in the UI to handle them
15:44:38 <andythenorth> except the liveries UI
15:45:05 <TrueBrain> why can you rename an engine to start with?
15:45:11 <TrueBrain> what is the point there?
15:45:34 <andythenorth> if we had telemetry
15:45:36 <TrueBrain> euuuhhh .. so we bother 100% of the users because 1% of the players wanted a niche, which they properly never use after the first 5 times? :D
15:45:43 <andythenorth> wonder how many savegames use that feature :P
15:45:46 <TrueBrain> renaming vehicles, sure
15:46:31 <TrueBrain> "hide" I get, but not in combination with the filter, sort, etc
15:46:35 <andythenorth> the search box we have is so ugly also
15:46:47 <andythenorth> that backspace icon disturbs me greatly
15:46:53 <frosch123> TrueBrain: iirc those renames were persistent in ttd, so you could rename ginzu a4 to class a4 or something
15:47:29 <andythenorth> is it just TTDP did it?
15:47:36 <andythenorth> or should I gravedig in forum threads?
15:47:48 <andythenorth> that has been productive recently
15:47:57 <TrueBrain> but the lack of padding at the top on the text-blob is annoying ..
15:48:11 <andythenorth> lol Max. Tractive Effort
15:48:14 <TrueBrain> just 3 pixels helps a lot
15:48:21 <andythenorth> frequent source of discussion max TE
15:48:46 <TrueBrain> "Life" .. lovely label
15:48:51 <TrueBrain> is it already 30 years old?
15:48:54 <andythenorth> such lovely myths around max TE
15:48:57 <TrueBrain> or what is 30 years?
15:48:59 <TrueBrain> what does it mean?! :D
15:49:42 <TrueBrain> guess it is life expectancy?
15:50:03 <andythenorth> relates to the reliability decay curve
15:50:03 <TrueBrain> reliability in % still annoys me :P
15:50:13 <andythenorth> life is meaningless with breakdowns off
15:50:22 <TrueBrain> right ... anything we didn't bash in this window .. let me see ...
15:50:38 <TrueBrain> what does the 3rd from right icon in the top bar do?
15:50:46 <frosch123> "<andythenorth> life is meaningless with breakdowns off" <- awesome out-of-context quote
15:51:01 <andythenorth> put that on the blog?
15:51:05 <TrueBrain> I forgot we had that
15:51:09 <TrueBrain> still don't know why we have that
15:51:32 <TrueBrain> does anyone ever use that for realz?
15:51:48 *** Wormnest has joined #openttd
15:51:56 <frosch123> widelands added that after us, it must be good
15:51:58 <TrueBrain> Sort by icon not centered ..
15:52:22 <TrueBrain> "classic sort" .. how is that useful information?
15:52:36 <frosch123> it's the "newgrf author intended" sorting
15:52:37 <TrueBrain> and if it is default, is it still "classic"?
15:52:49 <frosch123> but that's a very complicated term :)
15:52:55 <TrueBrain> I mean, I read 2 words that consume a lot of space, and I know nothing more than before I read them :P
15:52:58 <DorpsGek> [OpenTTD/OpenTTD] glx22 updated pull request #9085: Fix: [CMake] Validate grfcodec and nforenum executables https://git.io/JOMXf
15:53:21 <andythenorth> I can't be bothered to set really really really short ship names
15:53:35 <TrueBrain> make smaller ships :D
15:53:50 <frosch123> andythenorth: there is the "default window size" button :)
15:54:04 <andythenorth> all of this is survivable
15:54:05 <frosch123> no idea whether it appleis to all vehicle types, or whether it is per vehicle type
15:54:18 <frosch123> andythenorth: also, noone reads ship names
15:54:54 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9085: Fix: [CMake] Validate grfcodec and nforenum executables https://git.io/JOSr7
15:55:24 <andythenorth> if I do big enough sprites, will the names be hidden? :)
15:55:25 <DorpsGek> [OpenTTD/OpenTTD] DonaldDuck313 opened issue #9094: Electric train waits for free path when track is not electrified https://git.io/JOSrh
15:55:32 <andythenorth> or I could just do string ""
15:55:52 <andythenorth> what grf is TB using also
15:55:58 <andythenorth> and why are the trains going the wrong way?
15:56:41 <TrueBrain> frosch123: I think in general if we want to improve our GUI, we have some long long long hours ahead of us :D
15:57:28 <andythenorth> is it the UI, or the UX?
15:57:53 <TrueBrain> the UX -> often the goal of a window is a bit vague .. like "rename" on engines, etc
15:58:02 <TrueBrain> the UI -> often information is hard to read / process
15:58:07 <andythenorth> autoreplace is a good examplar for UX
15:58:09 <TrueBrain> like padding in this window, or "Life"
15:59:03 <andythenorth> pls explain the livery UI :P
15:59:06 <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #9085: Fix: [CMake] Check nforenum and grfcodec return value https://git.io/JOSoE
15:59:50 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9085: Fix: [CMake] Check nforenum and grfcodec return value https://git.io/JOSow
16:08:12 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JOS6O
16:08:27 <andythenorth> how does var 0x4B (date_of_last_service) handle date cheat?
16:08:32 <andythenorth> I can't see anything to copy from
16:08:47 <frosch123> nothing handles date cheat
16:09:29 <andythenorth> so that needs solved?
16:09:40 <andythenorth> not even sure what correct behaviour would be
16:09:48 <glx> cheats are expected to break things
16:09:56 <Rubidium> the behaviour is probably broken in year 5M too
16:10:16 <frosch123> andythenorth: btw. your "motivation" is just a "description" again
16:10:21 <Rubidium> potentially getting negative values
16:10:35 <andythenorth> there's no way I'm describing my intended use case
16:10:42 <andythenorth> that's where newgrf API goes off to die
16:10:56 <andythenorth> we confuse 'do we like grf author intention' with 'is this useful'
16:11:12 <andythenorth> I could maybe add some other motivation?
16:11:36 <frosch123> andythenorth: i wrote it in my comment :) Use cases are similar to date-of-service: stuff that is refreshed/refilled/refurbished in stations, instead of depots :)
16:11:57 <andythenorth> thanks I'll copy
16:18:37 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9085: Fix: [CMake] Check nforenum and grfcodec return value https://git.io/JOSXY
16:19:56 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #9088: Fix: [NewGRF] industry variables 65 and 66, object variable 46 https://git.io/JOSXR
16:21:06 <frosch123> damn, too slow. i just wanted to merge, and now there is a new comment :p
16:21:48 <glx> about nml#209 I think it's the usual callback result are 15 bits
16:22:34 <frosch123> the profit callback has a very weird return unit
16:23:21 <frosch123> it's a signed value with bit14=sign, and 13 fractional bits
16:24:18 <frosch123> nml handles fractional values in properties, but does not know about "units" in callback results
16:24:36 <frosch123> that's why I asked for an example :) there are so many things to do wrong :)
16:27:14 <frosch123> glx: i think it's an ottd bug again :)
16:27:42 <glx> seems he's good at finding ottd bugs and blaming nml ;)
16:27:45 <frosch123> if (HasBit(callback, 14)) result -= 0x4000; <- i think that should be a 0x8000
16:28:15 *** gelignite has joined #openttd
16:28:29 *** snail_UES_ has joined #openttd
16:28:58 <glx> but yeah usually with nml a compilable example is better, because we can check the NFO
16:30:10 <peter1138> How can I make MSVC spout these warnings... :/
16:32:49 <Rubidium> the alternative would be to just skip those characters in the validation and let it run up to "last", and only keeping the valid Utf8Encoded characters
16:35:27 <DorpsGek> [OpenTTD/OpenTTD] frosch123 opened pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSDc
16:37:59 <TrueBrain> You want us to learn French now? :D
16:38:21 <glx> it's displayed in english for me
16:38:23 <TrueBrain> Je ne parle pas francais
16:38:43 <glx> I always disable automatic translation
16:39:00 <andythenorth> TrueBrain est un grand mechant
16:39:07 <TrueBrain> It really is a French page on a French url
16:39:26 <glx> yeah, MS always move me to french urls
16:39:36 <TrueBrain> It's fine we van find the English page :)
16:39:42 <peter1138> A multibyte '\0' is invalid, no?
16:40:27 <TrueBrain> I rather learn French honestly :p
16:40:46 <TrueBrain> Tried many times in my life
16:41:13 <glx> it's not easy, even as native language ;)
16:41:29 <DorpsGek> [OpenTTD/OpenTTD] LordAro requested changes for pull request #9067: Feature: rework in-game Network GUI https://git.io/JOSyq
16:41:35 <glx> but at least we start early
16:41:48 <frosch123> TrueBrain: start with simplified french (canadian) instead of traditional french?
16:42:02 <frosch123> or is it the other way around?
16:43:12 <michi_cc> I guess canadians will say the same about french? :D
16:43:36 <Rubidium> canadian french is arguably more difficult
16:43:49 <Rubidium> just look at the stop signs in Quebec vs Paris
16:44:26 <glx> isn't it a red octogone ?
16:45:13 <glx> ah yes they are forced to translate everything
16:46:29 <Rubidium> though why isn't Wallonia using arret then? They got the same "problem"
16:46:43 <frosch123> well, the majority of germans do not know how to spell "Stopp" in their own language :)
16:47:22 <LordAro> though there are fewer good pictures on that wiki page
16:47:39 <frosch123> LordAro: i expected a stop sign with 37 letters :p
16:47:45 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
16:48:27 <Rubidium> maybe, but the "stop" sign being "arret" in Canada is just to spite the English speakers as "stop" would have been an equally valid French word
16:48:40 <LordAro> Welsh isn't particularly long
16:48:50 <LordAro> just lacking in (recognisable) vowels
16:49:47 <LordAro> yeah, but that's a PR stunt :p
16:50:26 <glx> IIRC new zeland has some nice ones
16:51:10 <peter1138> I added /Wall to by C(XX)_FLAGS. No warnings at all, so not sure if it worked :p
16:52:04 <LordAro> ooh! i've got the "cursor drifts to the left" in FF on linux that nielsm has
16:53:08 <nielsm> LordAro: mine drifts down tho
16:53:30 <TrueBrain> so many comments on code I moved around from other places :P
16:53:38 <LordAro> makes clicking on things quite fun
16:53:44 <LordAro> TrueBrain: pfft, i don't look at the deletions
16:54:04 <TrueBrain> LordAro: no, it is okay
16:54:07 <glx> peter1138: check CompileFlags.cmake:51
16:54:09 <TrueBrain> but it is hilarious how bad we programmed :)
16:55:23 <TrueBrain> can you make callback functions inline? I don't even want to know, that sounds wrong :P
16:55:29 <peter1138> Oh, it's already included.
16:55:59 <glx> for gcc clang and osx yes
16:56:48 <LordAro> TrueBrain: probably does nothing, as inline is only a hint anyway
16:57:02 <glx> wow Wall for MSVC is noisy
17:03:09 <TrueBrain> "this line is too long and could be on multiple lines" - "no need to do this, you can just do it on one line" :D
17:04:28 *** jottyfan has quit IRC (Quit: jottyfan)
17:05:29 <TrueBrain> LordAro: you ask for newlines twice, and I have no clue where I would need to add one :)
17:05:47 <LordAro> TrueBrain: "} break;"
17:06:10 <TrueBrain> yes ... where do you want a newline?
17:06:16 <LordAro> break should be on its own
17:06:36 <peter1138> break before the } normally.
17:06:39 <TrueBrain> do I dare to ask: why? :)
17:07:46 <LordAro> well you wouldn't write "} some_function();"
17:07:57 <TrueBrain> but I will change it :P
17:08:12 <TrueBrain> "break" is not "some function()" :D
17:08:21 <LordAro> it's a statement though
17:09:12 <TrueBrain> I was fishing if there was a difference, but "because we do it that way", which is fine, that is what code styles are for :)
17:09:41 <LordAro> oh sure, there's no difference at all
17:09:48 <LordAro> but i've never seen it done like that before
17:10:20 <TrueBrain> at my last job this was the way to do it :)
17:10:31 <TrueBrain> it helps visually to see cases do have a break at the end, basically
17:10:51 <TrueBrain> but compilers these days warn about this too, I believe?
17:11:16 <TrueBrain> the accidental fallthroughs ..
17:11:35 <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOS7m
17:11:41 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOS7O
17:11:58 <TrueBrain> btw, I checked, we no longer show the user "advertised" now :)
17:12:08 <TrueBrain> wanted to be sure I really replaced them all :P
17:12:51 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9092: Add: [GitHub] use issue templates to make it more clear what we expect from users https://git.io/JOS7n
17:13:50 <TrueBrain> now to test everything again to make sure I didn't break rebasing :D :D
17:14:35 <LordAro> get it in soon and then rebase STUN PR, so we can play again :p
17:14:42 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain dismissed a review for pull request #9067: Feature: rework in-game Network GUI https://git.io/JOSoX
17:14:48 <TrueBrain> STUN PR is going to take some time to finish, sadly ...
17:14:56 <TrueBrain> owh, shit, I forgot doxygen still
17:14:59 <TrueBrain> I think I had them all
17:15:36 <TrueBrain> its so annoying we already miss many doxygen comments, that it is hard to see if you add to the problem or not :)
17:16:14 <peter1138> Is 9095 really correct?
17:16:43 <peter1138> hmm, cos it's unsigned.
17:19:09 <peter1138> value is 14 bits, 0-3FFF, so with -8000 it becomes -8000 to -4001?
17:20:13 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSdm
17:22:48 <didac> frosch123 wanna talk about sprites x-)
17:24:33 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 opened pull request #9096: Fix: String validation could leave invalid Utf8 encoded strings https://git.io/JOSdQ
17:24:39 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSd5
17:24:51 <frosch123> didac: i will have dinner now, so in 30 minutes or so :)
17:26:34 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSFt
17:26:55 <TrueBrain> LordAro: I think I got all the doxygen comments, please let me know if I forgot any still :D
17:26:59 <TrueBrain> hard to spot honestly :(
17:27:32 <TrueBrain> even made some more C++ stuff .. instead of uint void A(uint line) { line++; return line; }
17:27:38 <TrueBrain> uint A(uint line) { line++; return line; }
17:27:45 <TrueBrain> void A(uint &line) { line++; }
17:27:51 <TrueBrain> should make at least 1 person here happy? :P
17:28:56 <peter1138> I mean, wouldn't just line++ without a function be simpler? ;-)
17:31:59 <_dp_> uint &A(uint &line) {line++; return line; } :p
17:32:24 *** keenriser has joined #openttd
17:32:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9096: Fix: String validation could leave invalid Utf8 encoded strings https://git.io/JOSF5
17:33:20 <Rubidium> TrueBrain: that's not C++ enough. It should have been template<T> T A(T line) { line++; return line; } ;)
17:33:57 <Rubidium> that way you don't need to change it when lines change from one type to another
17:34:28 <TrueBrain> template<T> void A(T &line) { line++; }
17:34:50 <peter1138> Boost should be involved somewhere.
17:35:32 <TrueBrain> okay, I do not understand #9096, and tempted to just YOLO approve it .. but maybe someone else has more luck understand the horror that is called UTF8 decoding :D
17:35:43 <TrueBrain> (don't understand the old nor the new .. "it looks fine" :P)
17:36:26 <LordAro> seems like the sort of thing that should be unit tested :p
17:37:08 <didac> OK be back in 20-30 min then ^_^
17:38:54 <LordAro> TrueBrain: all the static function ins NetworkClientListWindow
17:39:39 <TrueBrain> they all have doxygen, not?
17:39:47 <LordAro> missing parameters defs
17:39:54 *** EER has quit IRC (Ping timeout: 480 seconds)
17:39:57 <TrueBrain> haha, okay, being specific is sometimes important :P
17:40:03 <TrueBrain> steno is not my best capability :D
17:40:21 <TrueBrain> ugh, yes, the uselessness of documenting parameters that are pretty obvious ..
17:40:36 <LordAro> feel free to decide that it doesn't matter
17:40:43 <LordAro> but doxygen does warn about them :p
17:40:58 <TrueBrain> it is just pointless .. but either you do it everywhere or nowhere
17:41:04 <TrueBrain> for many functions it is useful
17:41:27 <TrueBrain> * @param w The instance of this window.
17:44:06 <TrueBrain> well, maybe I should word it differently: I wish you could include stuff in doxygen
17:44:11 <TrueBrain> just: @include <funcname>
17:46:35 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSNA
17:46:38 <DorpsGek> [OpenTTD/OpenTTD] frosch123 closed pull request #9095: Fix: [NewGRF] Negative results from cargo-income callback 39. https://git.io/JOSDc
17:49:36 <TrueBrain> lol, doxygen also wants to document ctors?
17:49:42 <TrueBrain> that sounds not so useful?
17:50:19 <LordAro> might be useful to document parameters
17:50:37 <LordAro> probably not so useful to "* Construct a Thing"
17:50:57 <TrueBrain> here too you miss how you can document stuff .. comments are not really made for this :)
17:51:07 <TrueBrain> as indeed, what the ctor does .. duh .. its about the parameters :P
17:53:13 <Rubidium> LordAro: constexpr-ing the hell out of it does not allow me to static assert it, so there's no way to do unit tests (yet?) But I agree it would be a good one for unit tests
17:54:04 <DorpsGek> - Update: Translations from eints (by translators)
17:54:07 <LordAro> Rubidium: i was thinking about #9038 :)
17:54:12 <LordAro> yeah, constexpr not a good fit here
17:57:04 <TrueBrain> did my best to doxygen the hell out of it :P
17:57:45 <TrueBrain> still didn't break it :D :D
18:06:37 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 updated pull request #9096: Fix: String validation could leave invalid Utf8 encoded strings https://git.io/JOSdQ
18:15:36 *** ChanServ sets mode: +v tokai
18:21:13 *** Progman_ has joined #openttd
18:22:49 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
18:26:31 *** Progman has quit IRC (Ping timeout: 480 seconds)
18:28:41 <peter1138> New Map Array will allouw smooth curves right?
18:29:00 <peter1138> Just watching someone play Voxel Tycoon. Placing tracks looks infuriating.
18:29:31 <peter1138> And they went with fucking pre-signals...
18:33:17 <andythenorth> the best pre-signals are pre-signals
18:34:20 <peter1138> Should doxygen comments be in the .h or the .cpp?
18:36:59 <DorpsGek> [OpenTTD/OpenTTD] TheDude-gh opened issue #9097: Bug Report - Cargo variable initial payment bad type or newgrf read wrong https://git.io/JO9Jw
18:51:13 <DorpsGek> [OpenTTD/OpenTTD] PeterN opened pull request #9098: Fix #9097: Upper 16 bits of cargo base payment rate were discarded. https://git.io/JO9Tz
18:52:09 <frosch123> peter1138: our style is to put doxygen closest to the implementation
18:52:20 <frosch123> so, in the .cpp if there is one. in the .h for templates/inline stuff
18:53:31 <peter1138> For me, in VS Code, it only shows the doxygen comment if it's in the header file.
18:54:22 <peter1138> Maybe that's a VS Code issue.
19:00:58 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
19:01:18 *** WormnestAndroid has joined #openttd
19:03:11 <TrueBrain> peter1138: works fine here for both
19:06:19 <peter1138> I'm on Windows, not using gcc.
19:06:28 <peter1138> So are you, openttd.exe.
19:06:31 <TrueBrain> I am on windows too
19:07:59 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 opened pull request #9099: "Feature": Bigger packets in network protocols https://git.io/JO9Ix
19:08:17 <peter1138> It works when the doxygen is in the .cpp, when I'm editing within that same .cpp file.
19:08:23 <peter1138> Otherwise, it must be in the .h file.
19:08:28 <Wolf01> WSL is probably the best thing they could have done
19:09:15 *** tokai|noir has joined #openttd
19:09:15 *** ChanServ sets mode: +v tokai|noir
19:10:34 <peter1138> Logically it makes sense that it only looks in the header, tbh.
19:11:58 <glx> intellisense looks everywhere for me, but some macros break it (like FOR_ALL_XXX macros)
19:12:33 <TrueBrain> peter1138: it does :)
19:12:38 <glx> but I use VS, not VS code
19:15:58 *** tokai has quit IRC (Ping timeout: 480 seconds)
19:16:10 <DorpsGek> [OpenTTD/OpenTTD] glx22 merged pull request #9085: Fix: [CMake] Check nforenum and grfcodec return value https://git.io/JOMXf
19:16:42 <peter1138> VS Code on Linux also seems to want it in the header file.
19:18:58 *** ChanServ sets mode: +v tokai
19:20:55 <peter1138> Hmm, maybe it's a cache thing?
19:21:22 <peter1138> Seems to be working for other doxygen comments in .cpp, just not the one I add. It works straight away for a doxygen comment in .h.
19:25:34 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
19:26:08 <peter1138> Hmm, not everything.
19:35:27 <peter1138> Hmm, 0xC064 is definitely wrong for a -100 result.
19:36:07 <peter1138> That would be the value if we did result = -result.
19:40:27 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9098: Fix #9097: Upper 16 bits of cargo base payment rate were discarded. https://git.io/JO9Op
19:43:47 <peter1138> "For example, the maximum returnable value of 16383 means that the player gets twice the price factor for every unit. "
19:44:02 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #9035: No obvious way to change player name during multiplayer session https://git.io/JOmiO
19:44:19 <TrueBrain> no more googling for you peter1138
19:44:39 <LordAro> missed the nightly tho :p
19:45:19 <TrueBrain> tomorrow there is another one :)
19:49:25 <LordAro> might be worth looking through the open issues for other things that are relevant
19:49:50 <frosch123> LordAro: this nightly you can play in frisian
19:50:03 <frosch123> we can't have two major features by a dutch in one nightly
19:54:19 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:04:18 <peter1138> Is "graphics { profit: testprofit; }" some NML short hand for "if it's callback 39, jump to the testprofit chain"?
20:06:33 <andythenorth> goes it fix things?
20:06:56 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler updated pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JYNbo
20:14:35 <peter1138> So update more callback vars when a vehicle has finished loading?
20:17:04 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JO9nB
20:23:08 *** gelignite has quit IRC (Quit: Stay safe!)
20:33:57 <didac> frosch I've been playing around with subsprite but couldn't succeed yet. I noticed even using { -INF, -INF, INF, INF } ends up trimming it all
20:34:07 <peter1138> Well I don't have nmlc to compile #209
20:34:36 <peter1138> But stuffing the appropriate hex values into OpenTTD makes it do the right thing.
20:35:13 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
20:35:27 *** WormnestAndroid has joined #openttd
20:36:01 <glx> I can build the grf for you
20:37:03 <frosch123> didac: it shouldn't do that :)
20:38:34 <peter1138> Thanks, but le 404.
20:39:20 *** lobstarooo has joined #openttd
20:39:27 <glx> weird it fails locally too
20:39:34 <peter1138> I don't think it's a working example anyway.
20:39:34 <LordAro> peter1138: is that a french 404?
20:39:58 <peter1138> It's just a cargo, there isn't going to be anything to produce/consume this cargo.
20:40:33 <FLHerne> I built a grf, but my file-hosting burned down in that OVH mess and I still haven't got to fixing it...
20:45:29 <didac> frosch123 yep it seems to do it, but maybe I messed something up. looking at gfx.cpp GfxBlitter now...
20:45:45 *** lobster has quit IRC (Ping timeout: 480 seconds)
20:45:51 *** lobstarooo is now known as lobster
20:46:22 <didac> I can't believe I forgot how to mention someone on IRC. Maybe I should use mIRC like 20 years ago
20:47:01 <FLHerne> Oh, glx's works now too
20:47:03 <frosch123> the generated nfo looks fine. the original nml source has "-11000" in a comment, which would be out of range
20:47:05 <FLHerne> At least they're the same :p
20:48:29 <peter1138> Isn't it -16384 to 16383?
20:48:39 <peter1138> (after taking account of bit 14)
20:53:12 <peter1138> action 3_1 is the calculation in the switch?
20:53:45 <peter1138> the bit inside the [ ] I mean.
20:54:52 <peter1138> Not sure what this *329/256 is
20:56:15 <frosch123> so there is some nml return value magic
21:00:56 <frosch123> no idea what the intention is here
21:01:05 <frosch123> what unit does nml use?
21:01:18 <glx> btw [] calculation is sprite 6
21:01:51 <frosch123> the docs say, the result is divided by / 256 (instead of / 8192 as in ottd). so that would be a factor 32
21:02:03 <frosch123> but where does this 41 come from?
21:02:47 <peter1138> Is it munging the result of the callback value then?
21:02:51 <glx> oh of course var 1C is last_computed
21:03:45 <peter1138> I'm not sure why NML needs to have a different number range than NFO here, heh
21:04:53 <andythenorth> what's the unit conversion on price_factor?
21:05:04 <andythenorth> "price_factor": {"num": 0x12, "size": 4, "unit_conversion": (1 << 21, 10 * 20 * 255)},
21:05:16 <frosch123> ok, origin is the action 0 price factor
21:06:08 <andythenorth> nml makes it £pounds
21:07:52 <peter1138> So in between the return value from testprofit, and actually returning the callback result, it seems to be doing some sums? Possibly messing up signedness?
21:08:04 <frosch123> ok, nml does the same factor 2 error in the docs
21:08:24 <frosch123> actual range for the callback in nml is -6374..6374
21:08:53 <peter1138> Okay but here it's return a value much less than that.
21:09:00 <frosch123> but probably var 1C fails for negative values?
21:09:45 <glx> so has no issue with negative values (not like using direct return value for procedure call)
21:10:56 <frosch123> but 1C does no sign extension
21:11:15 <milek7> hmm, I think I found bug in opengl code
21:11:26 <frosch123> the procedure returns 0x70FF = -3841
21:11:35 <frosch123> what does that store in 1C?
21:12:17 <peter1138> This sort of conversion should be compile-time really, or just not happen :/
21:13:14 <peter1138> Bit late now I guess
21:13:19 <milek7> GetVideoBuffer is called from win32 AllocateBackingStore without matching ReleaseVideoBuffer
21:13:38 <milek7> and if !this->persistent_mapping_supported, then it will try to map same buffer twice
21:14:31 <milek7> I'm missing something?
21:16:03 <michi_cc> milek7: win32_v.cpp line 1430?
21:17:17 <frosch123> the multiplicaton would be fine, but the divison needs a correct negative number
21:18:49 <milek7> michi_cc: it releases previous, but LockVideoBuffer will call GetVideoPointer again
21:19:09 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JO9EZ
21:21:50 <glx> I guess the compile grf is not usable ingame
21:22:09 <peter1138> You can look at the payment graph
21:22:46 <peter1138> I just worked that one out myself :)
21:23:01 <michi_cc> milek7: If the buffer is currently mapped, it will be released. Then the buffers will potentially be re-created and the mapped again. Where is the again?
21:23:09 <glx> now to put the breakpoint in the right place :)
21:23:25 <peter1138> I got a value of 25
21:23:34 *** Samu has quit IRC (Quit: Leaving)
21:23:39 <peter1138> (which is 20 * 329/256)
21:24:50 <glx> and I forgot to put the grf in an accessible place
21:26:18 <peter1138> It definitely never goes negative in the graph.
21:26:30 <glx> I want to check 1C content :)
21:26:31 <michi_cc> milek7: AllocateBackingStore is called via ClientSizeChanged, that is called directly (or indirectly via Window Messages) from VideoDriver::Tick() while the buffer is locked (which is why it has to be mapped before exiting AllocateBackingStore).
21:26:44 <didac> my understanding is that SubSprite subsprite = { -INF , -INF , INF , INF } should be no-op and not cut the sprite
21:26:45 <frosch123> peter1138: i don't think the graph uses the callback
21:26:50 <peter1138> frosch123, it does.
21:27:26 <peter1138> There's 3 distinct values, as requested, but all positive.
21:27:51 <DorpsGek> [OpenTTD/OpenTTD] PeterN merged pull request #9098: Fix #9097: Upper 16 bits of cargo base payment rate were discarded. https://git.io/JO9Tz
21:27:54 <DorpsGek> [OpenTTD/OpenTTD] PeterN closed issue #9097: Bug Report - Cargo variable initial payment bad type or newgrf read wrong https://git.io/JO9Jw
21:29:22 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JO9ur
21:30:40 <glx> I see 20, 32511 and 28927
21:31:17 <glx> let's display them in hex
21:32:52 <glx> 1C is correct, nml calculation based on 1C is wrong
21:33:02 <frosch123> it's the division that fails
21:33:15 <frosch123> division needs a proper sign-extended negative value
21:33:34 <peter1138> It's not sign-extending. Can it?
21:34:15 <frosch123> we can add a shift left 17, and an arithmetic shift right 17 :)
21:34:31 <glx> it's not even a negative 16bit value
21:34:35 <frosch123> but i am still confused about the weird units nml uses
21:34:43 <peter1138> It's a negative 15bit value, that's the point.
21:34:46 <frosch123> glx: it's a negative 15bit value
21:34:59 <peter1138> Yeah, I have no idea why it would be using different units
21:35:17 <glx> nml is full of weird stuff
21:35:37 <frosch123> nml does that in various places, where nfo units are stupid. for example it does km/h for all vehicles, instead of different units for each vehicle type
21:35:43 <peter1138> How many cargo NMLs use this callback...? :)
21:35:55 <peter1138> In this case the units are a factor though.
21:36:18 <peter1138> I don't see how it makes sense to convert to something else
21:36:25 <peter1138> I don't know what this 199 or 41 thing is.
21:37:47 <andythenorth> how many grfs ever used this callback? :P
21:37:55 <andythenorth> decompile all bananas grfs :P
21:38:14 <frosch123> the the action0 property nml defines "1 unit = Payment for delivering 10 units of cargo across a distance of 20 squares (in British Pounds). "
21:38:39 <frosch123> i think the callback uses the same unit but / 256 for 8 fractional bits
21:38:43 <peter1138> Yes, but that's for the base cargo rate.
21:39:27 <peter1138> The cargo base rate is already included in the sum after the callback, so shouldn't be a factor in the callback result.
21:39:41 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9093: Feature: [NewGRF] vehicle var 0x4E, 'date of last station departure' https://git.io/JO9gk
21:41:20 <peter1138> andythenorth, I wish I hadn't mentioned desyncs now. The vars are never an issue, it's how they are used...
21:41:33 <Timberwolf> I think this may require tooling in order to get the correct 3D shape.
21:41:55 <andythenorth> Timberwolf hulls are just difficult :)
21:42:03 <peter1138> What voxel tool are you using?
21:42:07 <andythenorth> very hard to understand the actual shapes without projection drawings
21:42:14 <andythenorth> and even drawings, you need sections
21:42:33 <Timberwolf> MagicaVoxel, although I think I'll need to write something to generate plausible hull shapes.
21:44:01 <peter1138> frosch123, I am going mad right? It is -16384 to 16383... as it's 14 bit + sign bit. Not -8192 to 8191.
21:45:40 <frosch123> -32k..32k is 16bit signed int, -16k..16k is 15bit signed int (=14bit + sign)
21:46:00 <peter1138> Anyway, as a factor, this callback should return 0x2000 to use base payment rate
21:46:05 <frosch123> so, yeah, probably wrong every time :p
21:46:24 <peter1138> 0x3FFF is (approx) double
21:46:43 <peter1138> I don't see what the point of the scaling is
21:46:49 <frosch123> and if the callback result is multiplied by the base_price, then i also have no idea why nml would pick another unit
21:46:59 <frosch123> so maybe cargo_profit_value should just be deleted
21:47:24 <andythenorth> Timberwolf I'm now wondering if hulls are _worse_ in voxels. You actually have to place the blocks, not just fake it with shading like I do :P
21:48:14 <glx> lol stupid me (and NFO, there's only 3 valid shift operators in action 2)
21:49:09 <milek7> michi_cc: place breakpoint at GetVideoBuffer and ReleaseVideoBuffer
21:49:23 <milek7> observe that GetVideoBuffer get called twice without Release inbetween
21:50:41 <peter1138> Timberwolf, native vox support when? :D
21:51:22 <Timberwolf> andythenorth: Yup. But I think I could write something that will generate the hulls for me, then I can decorate them afterward.
21:51:43 <andythenorth> I have standardised hulls
21:51:49 <andythenorth> then I place the deck stuff
21:52:30 <andythenorth> I drew a ship once in Bryce 3D
21:52:48 <andythenorth> using a few splines for the hull cross sections at different points
21:53:02 <andythenorth> then letting the mesh be interpolated between them
21:53:08 * andythenorth hasn't done 3D for years
21:54:04 <peter1138> I loaded up Blender once. Ran away scared.
21:55:11 <peter1138> 1) how many people use this callback
21:55:25 <andythenorth> it's of almost no use
21:55:29 <peter1138> 2) how many realise the range is different
21:55:39 <andythenorth> I guess someone found the bug though
21:55:50 <peter1138> 3) how many noticed/cared
21:56:01 <peter1138> You'll only see the bug if you try to use negative values.
21:56:22 <peter1138> Otherwise, it works, but... why?
21:56:58 <michi_cc> milek7: Okay, that shouldn't happen this way. It seems something was broken during the game loop threading stuff.
22:00:04 <peter1138> Can you just get rid of it and pretend it doesn't exist?
22:06:48 <andythenorth> withdraw it, see who complains
22:15:04 *** andythenorth has quit IRC (Quit: andythenorth)
22:17:58 <DorpsGek> [OpenTTD/nml] glx22 opened pull request #212: Fix #209: Sign-extend value before cargo_profit callback result calculation https://git.io/JO9VQ
22:19:47 <peter1138> It still doesn't make sense.
22:21:01 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
22:21:10 <peter1138> Not that. There is no logical purpose for this scaling.
22:21:39 <glx> ah yeah I don't see why it was added
22:21:49 <DorpsGek> [OpenTTD/OpenTTD] michicc opened pull request #9100: Fix: [OpenGL] Main loop expects to start with the video buffer unmapped. https://git.io/JO9rv
22:21:51 <glx> but well at least now it works
22:25:32 <milek7> well I found it because my d3d11 driver crashed for same reason
22:25:35 <milek7> but now it might not be necessary anyway :D
22:26:59 <DorpsGek> [OpenTTD/nml] PeterN commented on pull request #212: Fix #209: Sign-extend value before cargo_profit callback result calculation https://git.io/JO9rg
22:32:20 *** erle- has quit IRC (Quit: Leaving)
22:35:28 <peter1138> Hmm, anyone remember that website that lets you compile code and show the output in assembly form?
22:36:29 *** sla_ro|master has quit IRC ()
22:41:27 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9100: Fix: [OpenGL] Main loop expects to start with the video buffer unmapped. https://git.io/JO9Ks
22:41:54 <peter1138> Hmm, / 2 vs >> 1 is... quite different.
22:42:48 <peter1138> signed vs unsigned. Hmm.
22:43:41 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #9100: Fix: [OpenGL] Main loop expects to start with the video buffer unmapped. https://git.io/JO9rv
22:52:44 *** Progman_ has quit IRC (Remote host closed the connection)
23:00:04 *** tokai|noir has joined #openttd
23:00:04 *** ChanServ sets mode: +v tokai|noir
23:06:43 *** tokai has quit IRC (Ping timeout: 480 seconds)
23:08:09 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:20:33 *** Zuu has quit IRC (Quit: Leaving)
23:56:35 *** didac has quit IRC (Remote host closed the connection)
continue to next day ⏵
