IRC logs for #openttd on OFTC at 2024-01-12
⏴ go to previous day
00:08:18 <_glx_> I only know `int i = 0; while (node->m_parent != nullptr) { i++; node = node->m_parent; }
00:17:12 <wensimehrp> peter1138[d]: I meant that in most languages, words are separated by spaces, and others don't have a separator. How would the game process the characters in those language without separators, say, *CJK*? I'm asking this question because I saw that in issue #11485 only the word "faster" got copied, but in #11752 only one characters was copied.
00:18:32 <wensimehrp> Also resizing the window (in #11752) doesn't help getting rid of the duplicated characters.
00:22:11 *** ufo-piloot_ has joined #openttd
00:22:20 *** ufo-piloot has quit IRC (Ping timeout: 480 seconds)
00:29:33 <peter1138[d]> OpenTTD just does what the system layouter tells it to do.
01:07:45 *** ufo-piloot_ has quit IRC (Ping timeout: 480 seconds)
01:17:08 *** ufo-piloot has joined #openttd
02:42:36 <wensimehrp> Is `advanced signal` in `{BLACK}Toggle showing advanced signal types` a proper noun?
02:47:23 <emperorjake> No, it's an adjective and common noun
02:50:30 *** Wormnest has quit IRC (Quit: Leaving)
03:00:09 *** herms has quit IRC (Quit: bye)
03:16:46 <wensimehrp> Is "闭塞信号", which means "block signals" in English, a good translation for that?
03:18:46 <wensimehrp> The current translation means "advanced signals", but block signals are not *advanced*, imo path signals are.
03:22:03 <emperorjake> Path signals are the basic signals that are available by default. Block signals (including presignals) are hidden away because they're only needed for certain niche playstyles. Hence why they're labelled as "advanced"
03:24:39 <wensimehrp> So they are labelled as "advanced" because of their users, but not because of how they worked.
03:35:10 <wensimehrp> Personally I don't think that labelling something as "advanced" or "worse" (ah I lack adjectives) based on their users is appropriate. Their labels should be based on their complexity and the techniques used. I think that pbs is "advanced" because it works better than block signals. What I think is that changing "advanced signals" to "block signals" is more accurate and improves readability,
03:35:10 <wensimehrp> because when the tooltip tells you "advanced signals" you won't really know how it is advanced.
03:42:28 <belajalilija> i dont think people generally count the pre signals as path signals
03:42:39 <belajalilija> despite the fact that they're basically funny path signals
03:43:27 <belajalilija> i think calling them advanced (or worse) is a good way to maybe dissuade people who dont need them from using them
03:44:35 *** Flygon has quit IRC (Read error: Connection reset by peer)
03:45:16 <emperorjake> They're not though, presignals are block signals
03:45:47 <emperorjake> so it wouldn't be wrong to call advanced signals block signals instead
03:46:34 <emperorjake> but they could also be called "obsolete" or "deprecated"
03:49:59 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:50:32 <reldred> Pre signals are significantly more complex to use than path signals, and cause a lot of confusion and chaos amongst new players. They were hidden for a reason and I would say are an ‘advanced’ feature (also obsolete lol)
04:09:13 <belajalilija> emperorjake: typo
04:09:18 <belajalilija> will edit, is 4 here
04:11:25 <belajalilija> may i also table "junk signals"
04:26:54 *** sjiveru has joined #openttd
04:26:54 <sjiveru> I think the point is not 'these are bad' nor 'these are block signals' but rather 'only touch these if you know what you're getting yourself into'
04:35:41 <lamarr> yeah exactly, PBS signals pretty much "just work". other signal types are more advance in their usage, even if they aren't as technically advanced
04:38:20 <talltyler> I chose “advanced” to strike a balance between giving info without context for a new user “show block signals” and being overly opinionated “show obsolete signals”
04:39:22 <talltyler> In any case, this is a fine debate to have but note that any change needs to happen in English first and then propagate down to translations — translations should follow English and not try to improve upon it. 🙂
04:46:47 <wensimehrp> Or a combination, "show obsolete advanced block signals".
04:48:22 <wensimehrp> or just "show obsolete block signals"
04:49:18 <wensimehrp> wensimehrp: As what I said, I don't really think they are *advanced*
04:55:51 <wensimehrp> Another problem of using the word "advanced" is that new players might think that advanced=good and try to use block signals even if they don't know how to use them.
04:56:22 <wensimehrp> At least that's what I think when playing a new game, advanced=good.
04:56:48 <reldred> Then that’s a misunderstanding of the English word ‘Advanced’ advanced does not mean good
04:57:17 <reldred> They describe two different qualities
04:58:08 <wensimehrp> In most of the times advanced is translated to "高级" in Chinese, and "高级" basically equals to "good quality" and "high level"
04:58:36 <reldred> Then perhaps to preserve meaning you should look for an analogue to ‘complex’
04:59:25 <wensimehrp> I need to discuss with some Chinese speakers about this one
05:04:13 <limyx826> How do you install vcpkg?
05:22:42 <reldred> limyx826: vcpkg itself isn’t too hard to install just follow the instructions on their GitHub (and make sure you install vs2022 first, vcpkg needs it regardless if you want to use vs code), the tricky one is getting vcpkg integrated with either vs2022 or vs code so you can compile.
05:22:59 <reldred> wensimehrp: that’s not productive
05:23:50 <reldred> Oh you need Git for windows as well
05:24:05 <limyx826> reldred: Yeah that's the part I confused, first I clone their repo into openttd project folder then install the dependencies that way.
05:24:58 <limyx826> Then I realised git might think it is a new folder, so I deleted the cloned folder and reclone it somewhere else.
05:25:12 <limyx826> Then right now I am lost.
05:26:30 <reldred> Nah keep vcpkg seperate
05:27:28 <limyx826> reldred: Yeah what I thought as much but then the command line won't work since the vcpkg is in other location.
05:28:19 <reldred> Yeah when you do vcpkg integrate install it whatever it tells you about putting it in your $PATH so it’s available elsewhere
05:29:47 <limyx826> I did vcpkg integrate install but yet it is not on my path
05:30:31 <reldred> Nah you fo that manually
05:30:46 <reldred> You’ll wanna google that bit lol
05:31:02 <reldred> I’m infront of a Mac atm so I can’t help too much lol
05:31:34 <limyx826> I did know how to place it in $PATH, it just that a bit tedious. Thank you mate.
05:32:57 <reldred> When you get to the vs code or vs2022 bit lemme know, I just went through this nonsense with my own setup, it’s a bit obtuse how you do it in vs code, in vs2022 it’s easier.
05:34:21 <limyx826> I use both actually. VS code is mainly for easy editing of code without building or loading much. VS2022 is for verify and building.
05:43:21 <reldred> Yeah fair, the config in vs2022 isn’t too bad, just a cmake options thing you paste the line in that came up during vcpkg integrate install
05:43:37 <reldred> Vs code you’ll want to edit some settings json file
05:49:27 <limyx826> For some reason cmake cannot find the libraries even I seem to install correctly
05:55:25 <reldred> Yeah so that’s the tricky bit you’re now up to: telling either vs code or vs2022 how to find that white
05:56:04 <reldred> So if you do another vcpkg integrate install pay attention to what it spits out where it says you need to add some DCMAKE tool chain thing
05:56:07 <limyx826> Right now is configuring vs2022
05:56:49 <reldred> You can copy paste that line that vcpkg integrate install spits out into a config option for cmake in vs2022
05:57:27 <reldred> You might need to google ‘vcpkg cmake vs2022’ and see what it says as I’m not infront of vs2022 atm
05:58:20 <reldred> But yeah you copy that line that pops up at the end of running vcpkg integrate install into cmake config options somewhwre and then it all just works (tm)
06:03:25 <limyx826> The thing is it's already there.
06:04:16 *** D-HUND is now known as debdog
06:21:05 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:23:34 <reldred> Yeah the path needs to look different wait one
06:25:07 <reldred> I always had to add this ``-DCMAKE_TOOLCHAIN_FILE="C:\Users\aegis\source\Repos\vcpkg\scripts\buildsystems\vcpkg.cmake" -DVCPKG_TARGET_TRIPLET="x64-windows-static"`` to a config option that was something like 'additional cmake arguments' or something?
06:25:46 <reldred> Yeah cmakeCommandArgs
06:25:56 <reldred> Obviously set your path to match
06:26:18 <reldred> But that’s the line I had saved in a text file to copy paste straight in, I know for a fact that works.
06:26:51 <reldred> I’ll be back at my desktop in an hour or two btw
07:25:09 *** kiothemodder has quit IRC (Quit: User went offline on Discord a while ago)
07:31:40 *** tokai|noir has joined #openttd
07:31:40 *** ChanServ sets mode: +v tokai|noir
07:32:08 <reldred> Oh, hang on, limyx826 did you put vcpkg in your $PATH? Both OpenTTD and JGRPP recent git clones can actually talk straight to vcpkg and pull those dependencies that they want and auto compile them ready for use. You can still use the DCMAKE_TOOLCHAIN_FILE and use the system-wide installed version of the dependencies (and you'll need to for older builds)
07:32:58 <reldred> This is my VS2022 btw
07:34:15 <limyx826> reldred: I did yet it can't be found
07:34:17 <reldred> But yeah couple ways of skinning the cat now it seems
07:34:28 <reldred> limyx826: might need to log out of windows and back in
07:34:39 <reldred> changes to $PATH don't take effect immediately.\
07:34:48 <reldred> they're loaded on login
07:35:05 <limyx826> reldred: To be precise I can access vcpkg via cmd or powershell
07:35:08 <reldred> new command prompt windows will work, but rest of the system not til logout and login
07:35:24 <reldred> and you're trying to compile a fresh git clone of openttd right?
07:36:05 <limyx826> reldred: I already did restarted my PC that's why I can access it via command
07:38:15 *** tokai has quit IRC (Ping timeout: 480 seconds)
07:39:01 <wensimehrp> talltyler: Do we need a seperate Translator Discussion channel for that?
08:00:06 <LordAro> wensimehrp: we have the translator groups on github for that
08:01:08 <reldred> peter1138[d]: I'm told you were playing with vehicle expiries recently? I'm seeing some weird issues with some sets that set vehicle expiry to 0, I'm guessing the original intention of expiry/lifetime 0 is 'lasts forever'. I can reproduce this with NARS Addon, all the NG wagons are gone by 1900 (1850 intro, model life 0, vehicle life 25)
08:03:10 <Eddi|zuHause> i don't think that was ever in the specs
08:03:24 <LordAro> wensimehrp: occasionally this channel does get busy, sometimes you just have to wait your turn :)
08:03:30 <wensimehrp> LordAro: Yes but basically nobody is using github discussions.
08:05:40 <wensimehrp> TrueBrainviaGitHub: Let see what would happen
08:06:09 <reldred> Eddi|zuHause: Yeah I don't know enough about vehicles in nfo/nml to be able to comment on that, but it's certainly a use of it. I mean, nothing stopping me from just rolling my own copy of NARS addon set but it's going to be a bit of an issue having a set on bananas that's functionally broken once 14.0 hits, I suspect some other sets would be hit with this, probably some of the 2ccset stuff.
08:06:53 <reldred> Depending how this goes nek0master_disc0rd might need to revise model_life and push an update out before 14.0 hits and people start sooking at them 🙂
08:08:05 *** Eddi|zuHause has quit IRC (Read error: Connection reset by peer)
08:08:48 <reldred> Wiki states 0-254 or VEHICLE_NEVER_EXPIRES
08:10:17 *** Eddi|zuHause has joined #openttd
08:19:23 *** Eddi|zuHause2 has joined #openttd
08:19:30 <truebrain> _glx_: lol, so rust-cache released a new PATCH version where they bumped node16 to node20 .. yeah, that is a major breaking change, and should have been a MAJOR version change 🙂
08:21:47 *** Eddi|zuHause3 has joined #openttd
08:21:47 *** Eddi|zuHause2 has quit IRC (Read error: Connection reset by peer)
08:24:21 *** Eddi|zuHause has quit IRC (Ping timeout: 480 seconds)
08:31:18 *** Eddi|zuHause3 has quit IRC (Remote host closed the connection)
08:31:56 *** Eddi|zuHause3 has joined #openttd
08:32:59 <reldred> that's written far too nicely and politely, you should stamp your feet up and down and make an ass of yourself.
08:39:53 <LordAro> "what even is semver" was my immediate thought when i saw your PR too
08:41:30 <LordAro> andythenorth: YOUR FAULT
08:41:42 <andythenorth> actually is, just the wrong grf 😛
08:42:32 <truebrain> LordAro: honestly, I can understand that a node16->node20 feels like something minor, when you don't actually maintain workflows 🙂
08:43:20 <truebrain> in a more serious issue, I have no clue how we could get node20 to work in that workflow ... so let's hope it survives 14.0, so we can use the survey to say: we can upgrade glibc to something newer 🙂
08:43:49 <truebrain> using a 10 year old glibc might be a bit .. old 😛
09:04:54 *** nek0master_disc0rd has joined #openttd
09:04:55 <nek0master_disc0rd> reldred: whats this about model life?
09:06:03 <peter1138[d]> reldred: Yeah... the spec is 255 for never expires. Hmm.
09:09:00 <reldred> peter1138[d]: Says 0-254 on the wiki
09:09:41 <reldred> nek0master_disc0rd: Dunno yet, model life 0 is broken on openttd nightlies atm after some changes peter made, waiting to see what he thinks
09:10:11 <reldred> peter1138[d]: in NML there's VEHICLE_NEVER_EXPIRES but I figure that maps to 255
09:10:22 <nek0master_disc0rd> I dont understand whats going on
09:10:40 <nek0master_disc0rd> I mean my stuff has a model_life, is that the issue? Like thats going to break shit in the future?
09:11:13 <reldred> Nah a model_life of 0 so that it never expires
09:11:30 <reldred> I doubt your set is the only one that has it
09:11:45 <reldred> But your set is the only one I play with that does 🙂
09:11:49 <nek0master_disc0rd> Why is this even a problem?
09:12:06 <reldred> Peter changed some calculations and it's broken it
09:12:36 <reldred> so I'm trying to find out what needs changing, the calculations to honor lifetime of zero, or everyones .grf's
09:13:56 <nek0master_disc0rd> I just dont really feel like doing anything with coding at the moment
09:14:03 <reldred> nek0master_disc0rd: do you still have access to your bananas? worst case I can update using the sources you sent me in one big batch job and then you can just upload it.
09:14:04 <nek0master_disc0rd> even a tiny thing, because I'm already doing stuff with other games
09:14:23 <reldred> I can do the code update if it comes to that, I just can't upload it to bananas for you.
09:14:55 <nek0master_disc0rd> Nope, I dont have access to Bananas, I never migrated and forgot my password
09:15:29 <nek0master_disc0rd> oh wow last updates I made was Sept 2022
09:15:47 <reldred> nek0master_disc0rd: yeah we were on github auth by then pretty sure
09:16:08 <reldred> anyway, let's see what peter says before we get too carried away
09:17:06 <nek0master_disc0rd> Thing is since I cna't be sure that all the model life lines inmy code are the same "format" with the spacing or tabbing, it wont be simple to just do one big "find and replace all in files" kinda thing to make it a big simple fix
09:17:12 <reldred> but you've got the upload update button, if I bounce out an updated version and send it to you should only take a minute or two to update it.
09:17:26 <nek0master_disc0rd> and I have over 200 files to change and right now I'm already frustrated doing shit with STALKER ANomaly XD
09:17:26 <reldred> nek0master_disc0rd: don't worry about it, I'll do it line by line if I have to.
09:17:40 <reldred> Let's see what Peter says anyway
09:17:41 <nek0master_disc0rd> I mean I can do it its just I'm not in the mood at this moment
09:18:27 <reldred> I cycle back and forth on TTD
09:18:41 <nek0master_disc0rd> I have not played OpenTTD seriously for a while now
09:18:51 <nek0master_disc0rd> infact I havent really been around the openttd discord much either
09:19:05 <truebrain> nek0master_disc0rd: means nothing; could also mean you are just a developer 😛
09:19:18 <nek0master_disc0rd> Well that includes all things OpenTTD
09:20:01 <nek0master_disc0rd> Once in a while I try booting it up after updating my verison of JGRPP and then like 5 minutes later after updating the JGRPP version of my scenarios i end up just quitting and going back to minecraft or stalker
09:20:24 <wensimehrp> reldred: I asked a few people, and one suggestion I got is to use "专用信号" which means "dedicated signals" in English.
09:20:24 <wensimehrp> Others also suggested "遗产信号", which means legacy signals; "复杂信号", "complexed signals"; and "早期信号", "Early signals".
09:20:24 <wensimehrp> But yeah, the differences between languages makes translating hard. As Tyler said, translations should stick to the original text, but on this one I think that it might break the rules.
09:20:26 <nek0master_disc0rd> Doesnt help that I wanna start really early with stuff but that also means nearly 2 centuries of progress and hcanges I'll have to do
09:20:27 <reldred> Yeah, I was like that for a while, then over christmas I got really stuck in again.
09:21:23 <nek0master_disc0rd> in the last couple years I've been doing more stuff with LTX/LaTeX than NML
09:21:34 <nek0master_disc0rd> (because stalker uses LTX for its configs and stuff like item configs)
09:21:56 <reldred> Huh, I thought LaTeX was mostly for document authoring?
09:22:14 <nek0master_disc0rd> Nah, seems like it can also be used for doing configuration stuff
09:22:22 <nek0master_disc0rd> like how some games use XML for its content
09:22:33 <reldred> Yeah, or JSON ot TOML or whatever
09:22:42 <truebrain> seems equally painful as NML, using LaTeX for configuration
09:23:05 <reldred> computers in general are painful
09:23:13 <reldred> we just move the pain around in different places
09:23:36 <nek0master_disc0rd> enh, its pretty simple really, I think the LTX setup that STALKER uses is more human readable than NML
09:24:20 <truebrain> that is just INI format 🙂
09:24:22 <peter1138[d]> That is nothing to do with LaTeX
09:24:48 <truebrain> but so a declerative configuration system; nice
09:24:52 <reldred> That reminds me, I was going to have a play with YAGL at some point
09:25:00 <nek0master_disc0rd> Well it uses LaTeX syntax and the .ltx file extension
09:26:01 <peter1138[d]> That is LaTeX syntax.
09:26:06 <truebrain> peter1138[d]: THE NIGHTMARES! Take it away!
09:26:19 <nek0master_disc0rd> ok then, well I guess I'm terribly wrong
09:26:22 <reldred> truebrain: yeah, so that's very similar to what I cut my teeth on with Command and Conquer
09:27:18 <peter1138[d]> So should we change the NewGRF specs because some NewGRFs rely on a bug?
09:27:21 <reldred> nek0master_disc0rd: probably just a case where two groups picked the same file extension and came up with their own lingo independently of each other
09:27:45 <nek0master_disc0rd> TBH before NML I kinda wished t hat making new vehicles for Openttd was as simple as it is to add stuff to Stalker
09:27:47 <reldred> peter1138[d]: well what else would a model_life of zero infer?
09:27:56 <nek0master_disc0rd> aside from having to work with 3d model and animations and textures
09:27:57 <reldred> I get that it's not to spec
09:28:14 <peter1138[d]> Specifically shortest life possible.
09:28:19 <truebrain> nek0master_disc0rd: even I assumed it would be easier then it is 😄
09:28:46 <peter1138[d]> Yeah we went along with NewGRF because TTDPatch started it.
09:28:56 <nek0master_disc0rd> I just wanted to be able to "code" something in notepad and not have to figure out comilers and hex stuff or what ever NFO uses
09:29:19 <truebrain> nek0master_disc0rd: You are not wrong btw. Stalker actually calls it LaTeX. now I am curious why 😄
09:29:44 <nek0master_disc0rd> it'd be kinda neat if in the future OpenTTD could maybe load loose files to make it easier for new creators to get stuff going
09:29:50 <truebrain> nek0master_disc0rd: I guess I should continue work on TrueGRF 🙂 Which had that as aim 🙂 Only do the "fun" stuff 🙂
09:30:06 <reldred> Content creation for TTD is definitely miles harder than a lot of other games that's for sure
09:30:31 <reldred> THAT said, have y'all ever seen CD Projekt Red's RED engine? OH my lord that is a nightmare fueled game to try and mod.
09:30:46 <nek0master_disc0rd> though the issue with games that load up "mods" with lots of loose files is longer load times and moving files around takes for ever when you have thousands of small files
09:31:08 <nek0master_disc0rd> reldred: I know, I spent like 2 days trying to get shit working
09:31:13 <truebrain> nek0master_disc0rd: so many solutions in between .. you can "compile" those files at first load, for example
09:31:14 <nek0master_disc0rd> and I dred doing that shit again
09:31:33 <peter1138[d]> Ah Teams' random "Please sign in again" popup.
09:31:42 <reldred> nek0master_disc0rd: Yeah that was deffo the case with the bethsoft games, their BSA format really optimises the texture and model compression and loading/streaming.
09:31:46 <truebrain> peter1138[d]: don't get me started ...
09:31:56 <nek0master_disc0rd> reldred: As well as backing up and moving stuff aroudn
09:32:23 <nek0master_disc0rd> no matter how fast your PC and storage is, many many small files brings file operations to a crawl
09:34:39 <nek0master_disc0rd> OH I just thoguht of something, with my nars addon set pretty sure not everything is set to have a model_life of 25 years as I just used that as a sort of "place holder" since I dunno for sure how long each locomotive was produced and in service for
09:35:16 <nek0master_disc0rd> I mean I could just look up the build dates and then you'd have like a model life of like 3-5 years, but IRL I know that north american railroads will buy stuff in bulk when a new model is introduced and then they run around the tracks for decades
09:35:22 <nek0master_disc0rd> especially with stuff like the SD40 and SD70
09:35:40 <andythenorth> modding OpenTTD via python? 😛
09:35:46 <nek0master_disc0rd> Heck stuff built in the 80's, 90's, and 2000's are getting rebuilt today
09:36:44 <nek0master_disc0rd> For example you got Dash-8's, Dash-9s, AC's, and early GEVO's that are being overhauled/rebuilt these days to update, fix up, and bring them up to modern emissions standards.
09:37:13 <reldred> nek0master_disc0rd: Mostly was a problem with the wagons that needed to last forever, the engines were fine for the most part.
09:37:14 <nek0master_disc0rd> I'd like to think that for North American stuff at least, the model life should basically be how long is this locomotive used for insted of just its build years
09:37:50 <nek0master_disc0rd> reldred: Is that all that actually needs to be fixed?
09:38:01 <nek0master_disc0rd> because I only have a "handful" of railcars
09:38:07 <reldred> As I said, I'm happy to make the changes and send it to you.
09:38:13 *** thelounge345 has joined #openttd
09:38:19 <reldred> Yeah, narrow gauge rolling stock was the worst hit because you've got a bunch set to last forever.
09:38:28 <reldred> but using a 'nonstandard' way
09:38:56 <reldred> Yeah model_life: 0 instead of model_life 255 or 'VEHICLE_NEVER_EXPIRES'
09:38:58 <nek0master_disc0rd> I think it set it to last like 65535 years
09:39:22 <nek0master_disc0rd> I already closed dropbox and I just wanna get back to stalker and youtube XD
09:39:47 <nek0master_disc0rd> I didn't even know at the time there was a proper way to do it
09:40:08 <nek0master_disc0rd> I dont think I ever even saw any other sets that used like say VEHICLE_NEVER_EXPIRES
09:40:12 <reldred> Yeah don't worry about it, once Peter gives his verdict I'll either send you an updated zip file like the one you sent me before, or I'll tell you to forget abouit it 🙂
09:41:03 <reldred> I've got someone who expects me to be playing grand theft horse right now anyway 😛
09:41:31 <nek0master_disc0rd> I just started a new run in STALKER ANomaly and put on the video about the Soviet Era Doomsday Machine
09:41:39 <nek0master_disc0rd> (aka Dead hand)
09:45:07 <peter1138[d]> Funnily enough, the default if you don't set it is 255 🙂
09:47:37 <reldred> Lmao, that's actually pretty funny
09:48:47 <reldred> anyways, it is time to grand theft horse, lemme know your thoughts peter1138[d] and I'll sort out neko with the nars addon if necessary
10:18:00 *** jinks has quit IRC (Read error: Connection reset by peer)
10:30:02 <alfagamma7> nek0master_disc0rd: Ah yes
11:29:23 <xarick> my current goal: pathfind to depot but limit the low_lvl pf to the restricted boundaries of corridor tiles
11:31:21 <xarick> goal is to aim at getting an exact cost
11:33:07 <xarick> if this turns out to be slow, then I have a backup idea
11:35:03 <kuhnovic> That's exactly what the ship pathfinder does (yapf_ship.cpp)
11:36:12 <xarick> i don't mean it like that
11:37:03 *** thelounge345 has quit IRC (Ping timeout: 480 seconds)
11:37:04 <xarick> i mean to have the low lvl pf exclude the entire map outside the water regions, but consider all tiles of the regions
11:38:02 <xarick> a region has 256 tiles
11:38:12 <xarick> a corridor has multiple 256 tiles
11:39:07 <xarick> the low lvl pf will be constrained to search within these tiles, instead of being free to use a tile outide the corridor area
11:40:12 <xarick> possibly even only just the tiles of the same patch label
11:44:16 <xarick> nevermind, maybe you're right
11:45:08 <xarick> i just can't use the restrict
11:46:29 *** thelounge345 has joined #openttd
11:47:18 <xarick> you thought of everything didn't you?
11:50:03 <kuhnovic> Hehe I spent a good amount of time working on this yeah 😛
11:51:05 <xarick> SetIntermediateDestination might need a small adjustment for the final region
11:57:39 <xarick> oh well, I'm getting used to them
11:59:36 <kuhnovic> Still better than using abbreviations
12:00:28 <_jgr_> These names are really not gigantic
12:04:28 <xarick> corridor doesn't keep track of label
12:06:22 <xarick> std::vector<WaterRegionDesc> m_water_region_corridor; should be WaterRegionPatchDesc
12:06:31 <xarick> let's see what happens
12:20:06 <xarick> are labels unique per region or unique per map?
12:42:29 <peter1138[d]> Damn, single-shot cyber-demoned 😦
13:04:36 <xarick> hmm, labels are unique per region 😦
13:09:44 <andythenorth> `GestaltGraphicsIntermodalContainerTransporters` is a proper long name
13:09:57 <andythenorth> and yet...it can't really be shorter
13:14:42 <kuhnovic> Unique per region. Because then you don't have to worry about the values in other regions, which makes the whole thing a lot simpler. And the biggest reason: there is no need
13:15:50 <xarick> that means more nodes for the pf 😦
13:18:51 <xarick> looks like I need to go to plan B
13:20:45 <xarick> give it several destinations
13:21:09 <xarick> in case of multiple depots found in the region
13:22:29 <xarick> but determining which one is in the right patch from which the ship enters from, hmm how
13:25:20 <xarick> maybe if there was a way to save from which label the region is entered from
13:29:47 <xarick> a kind of m_trackdir but for regions
13:49:33 <_glx_> we don't test GS API at all
13:51:14 <xarick> you mean the ScriptText
13:53:33 <limyx826> Err, I can build openttd but I can't start debugging?
14:29:58 <andythenorth> are flat docks just docks that don't need a slope tile?
14:30:03 <andythenorth> I never really had a concept
14:31:16 <limyx826> andythenorth: Pretty much and it is for canal and river, not water tile.
14:33:28 <limyx826> Just test in game that it is possible to build canal on top of water tile. ???
14:34:43 <emperorjake> why have 2 tile docks when you can have 1 tile docks?
14:35:28 <andythenorth> because there's existing code that assumes 2 tile (UI, placement, etc)
14:35:42 <andythenorth> but it's probably better to do them better
14:37:29 <limyx826> I think it is best to first use current pre-existing code for ship behaviours for this new dock.
14:37:57 <limyx826> Then new behaviour can be written if needed.
14:39:16 <_zephyris> I always imagined a 2 tile flat dock. One ground tile, one water tile, with a normal dock on the water tile and a slope up to it on the ground tile.
14:44:23 <limyx826> _zephyris: Funny I imagine it to 2 tile completely on river/canal more similar to drive thru roadstop. But your version works as well.
14:46:13 <peter1138[d]> Carefully finish level with 200 health/200 armour, all weapons and a bit of ammo.
14:49:00 <reldred> We should just bring back TTDPatch’s Buoys as docks feature and then do the rest with objects or stations 😛
14:54:44 <peter1138[d]> I have a patch for docks 😄
14:59:17 <reldred> Ya gonna merge that patch though? 😛
15:46:58 *** HerzogDeXtEr has joined #openttd
15:55:07 <xarick> GetDiagDirFromParent - how do I access this from a different ... thing
16:11:07 <xarick> `node.Set(&old_node, water_region_patch, node.GetDiagDirFromParent());`
16:11:07 <xarick> Exception thrown: read access violation.
16:11:07 <xarick> this->**m_parent** was 0xFFFFFFFFFFFFFFFF.
16:15:13 <xarick> the node isn't set yet?
16:19:40 <xarick> okay I see now. Can't make omelettes without the eggs 😐
16:23:03 *** Wormnest has joined #openttd
16:23:12 <talltyler> I know, this is the lazy way, but checking out another copy of the OpenTTD repo just to push a single-line commit to someone else's repo is so time-consuming. And the description needed rewriting.
16:28:12 <truebrain> Wow, that PR is some new level of pedanticness 😄
16:29:34 <talltyler> Yeah, but "too pedantic" isn't worth rejecting
16:29:51 <talltyler> (in the case of the original PR, mine is just a helping hand)
16:30:05 *** michi_cc[d] has joined #openttd
16:30:05 <michi_cc[d]> Okay, I'm not gonna read that much backlog 😁 Anything more newsworthy than the ship PF from the last weeks?
16:30:19 <talltyler> This is his second PR that had major Git issues
16:30:45 <truebrain> michi_cc[d]: Social integration plugin almost done, daylength in testing phase .. think that covers it
16:31:06 <truebrain> Owh, and migrating issues to discussions is spammy as fuck 😛
16:31:41 <talltyler> Hmm, I forgot to make an announcement
16:32:59 <talltyler> Could be an `all` post, but I tend to avoid doing that unless it's really important 🙂
16:33:07 <michi_cc[d]> Is tomorrow really the 14th??
16:33:14 <truebrain> michi_cc[d]: Owh, and lot of positive responses to the idea of adding the OpenTTD font 🙂
16:33:32 <michi_cc[d]> truebrain: Well, I've updated that PR earlier today 🙂
16:34:25 *** palalet has joined #openttd
16:34:25 <palalet> truebrain: I may be bit late to the party, but I run the jgrpp version for multiplayers from my home computer. When I moved this to run on my linux server (huge one with plenty of ram and cpu and stuff), the server went frozen twice in two weeks (after having uptime of several years). It completly disconnected from network. I did not dare to run OpenTTD there again and it did not crash since then.
16:34:25 <palalet> The game was public, password protected.
16:36:51 <_jgr_> palalet: Better off raising an issue on the github for my branch, rather than here
16:37:03 <truebrain> That post of mine was in relation to OpenTTD infrastructure, not hosting a server 🙂 so those things are fully unrelated 🙂
16:37:40 <palalet> i am not sure if it was related or not, just now reading this thread it came to my mind 😦 i have no proof or replication scenario, so its not bug material 😦
16:39:24 <truebrain> Your issue sounds like a full disk, or a kernel spin lock, or a memory leak with a poorly responding OOM killer; but hard to debug without logs and further info 🙂
16:39:45 <truebrain> (Not OpenTTD logs; dmesg logs)
16:41:17 <palalet> I would like to know how this game coordinator stuff works, because currently I am having an issue with launching the game. It was running just fine till yesterday, when my pc (damn you, windows!!!) restarted and today when I want to launch the game, I can see it locally on local IP adresses, but cannot connect the public / internet version, it times out. In the game server log it says its
16:41:17 <palalet> connected to the Game coordinator, but in game the game is marked as offline. Have no clue what might happend, if there some new windows firewall settings or if there is some issue with the game coordinator itself?
16:43:15 <palalet> truebrain: yeah, that is the problem, there was really nothing in the logs. no dump, no panic, nothing ... so lets not worry about it, it might have been some other problem, just coincidence. Only if more people notice this will be worth of trying some investigation.
16:43:29 <truebrain> palalet: Some routers do not allow you to connect to your own server from your own local network via the Internet. Just use the local IP instead
16:44:14 <palalet> that did not change at all. Same router, same setting, same connection, same provider. Why it worked two days ago and not today. I hate this kind of issues 🙂
16:44:49 <truebrain> It is also not really useful to bounce via the Internet to your local server
16:45:12 <palalet> i am playing with others over internet, not locally
16:47:41 <_glx_> oh and they could use emplace_back too 🙂
16:49:39 <_glx_> ah no, my bad the font is correct
16:56:13 *** JacobD88 has joined #openttd
17:21:02 <michi_cc[d]> I wouldn't be surprised if I missed something with Uniscribe. Documentation is almost non-existent and there's only a single usable example from MS 😦
17:24:02 <_glx_> I think it might be an off by one somewhere
17:30:59 *** Eddi|zuHause3 is now known as Eddi|zuHause
17:39:01 <wensimehrp> wensimehrp: It's always frustrating that no one is reacting.☹️
17:45:17 <merni> wensimehrp: The issue is perhaps that nobody here knows Chinese
17:45:18 <wensimehrp> xarick: That translation is wrong
17:45:35 <wensimehrp> merni: You're right
17:45:50 <_zephyris> More constructively, I'd look at the translation of block and near-synonyms like section, region, chunk, domain, then pick the most consistent term.
17:46:34 <merni> Also, if you get no replies, just go ahead with what you think is right -- OpenTTD translations are not a life or death matter and any decision you make can be changed later if a different consensus evolves
17:48:05 <belajalilija> i think 2TallTyler suggested that the english translaiton should be changed first
17:48:14 <belajalilija> and that for non english it should trickle down
17:49:30 <merni> is there no way to say "complicated" (if "advanced" is too positive) that doesn't automatically imply "good"
17:51:10 <xarick> can you believe I lost 2 hours figuring why compiling was failing and it was because of a missing ";" that intellisense didn't point out to me 😦
17:51:11 <frosch123> "old people signals"
17:51:18 <belajalilija> i think complicated has negative connitations
17:51:41 <xarick> it was spewing all sort of errors that made no sense
17:52:08 <merni> belajalilija: in English perhaps
17:52:32 <_glx_> hehe rule number one when compilation fails is to jump to the first error and see if there's a missing `;`
17:53:08 <_glx_> first in compile log, not in error window because they can be reordered there
17:55:29 <xarick> at last! directions between segments,
17:55:40 <xarick> m_water_region_enter_dir
17:56:02 <xarick> the origin doesn't have a dir
17:57:29 <xarick> and now I forgot why I wanted to make this
18:04:32 <wensimehrp> belajalilija: Yeah that is the problem.
18:09:27 <xarick> question, what is m_hash_next supposed to represent?
18:12:01 <xarick> i dont know what are hashes for to this day
18:18:28 <peter1138[d]> It points to the next items in the same hash bucket.
18:23:26 <xarick> oh, I see, so with directions, this becomes complicated
18:29:02 <peter1138[d]> xarick: Doesn't necessarily affect it at all.
18:36:35 <xarick> water regions could use some asserts
18:56:16 <_glx_> hmm should be possible to add support for filter to these 4 constructors, but it's not as easy
19:55:01 <wensimehrp> Strange... The characters should be in *one* line.
19:55:01 <wensimehrp> aXRobT1BV1M0LUhNQUMtU0hBMjU2JlgtQW16LUNyZWRlbnRpYWw9QUtJQVZDT0RZTFNBNTNQUUs0WkElMkYyMDI0MDExMiUyRnVzLWVhc3QtMSUyRnMzJTJGYXdzNF9yZXF1ZXN0JlgtQW16LURhdGU9MjAyNDAxMTJUMTk1MjUyWiZYLUFtei1FeHBpcmVzPTMwMCZYLUFtei1TaWduYXR1cmU9MWQxMmExOTNhY2MzOGQzODU3Mzk5YmE1ZmY0ZTVkMTdmNTViZTAwODAwZmJlYjA1YjBhYzE0NThiYTZjODJhMCZYLUFtei1TaWduZWRIZWFkZXJzPWhvc3QmYWN0b3JfaWQ9MCZrZXlfaWQ9MCZyZXBvX2]
19:56:38 <wensimehrp> That's still two separate lines
19:57:09 <wensimehrp> I'll test it on my computer later this day anyways
19:58:28 <_glx_> The fix was about the duplication (and there were 2 duplicates for this example string)
19:58:56 <wensimehrp> So it doesn't effect line changes
19:59:44 <Rubidium> the problem with where the line changes is that it isn't OpenTTD's code, but an external library that is used
20:00:15 <Rubidium> on my computer it does the right thing, but I'm not using the Windows uniscribe library for that
20:01:01 <_glx_> OpenTTD tries to find the best location to split
20:01:15 <Rubidium> having said that, replacing the single wide colon with the 'ASCII' colon and a space will probably solve the issue for Windows
20:04:05 <wensimehrp> The problem is that using ascii colons is not appropriate. ":" is only used for very specific purposes, like maths.
20:04:58 <wensimehrp> And using the wider version "：" breaks compatibility
20:08:43 <Rubidium> in the current zh_CN language file they are used interchangeably/inconsistently
20:10:52 <wensimehrp> *Latinize Chinese*
20:20:55 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:29:44 <xarick> how many number_of_patches can there be in a water region, realistically
20:30:14 <xarick> the maximum value that you can get if you try hard enough
20:30:33 <peter1138[d]> Presumably the wide "：" isn't triggering a word break, but ": " does.
20:31:31 <xarick> i know the theoretic max is 256, but I wanna see if i can get by using uint8_t
20:33:01 <xarick> uint8 is undefined nvm
20:37:17 <xarick> there's a lot of `this->` missing
20:50:56 <frosch123> would a zero-width-space work?
20:51:32 <frosch123> strgen could learn {ZWSP}
20:53:41 <_glx_> uniscribe doc says "All white space is breakable except nonbreaking space (NBSP) and zero-width nonbreaking space (ZWNBSP)."
20:57:29 <peter1138[d]> What's the unicode point for that? 😄
21:01:13 <peter1138[d]> Not sure if ICU behaves differently mind you.
21:04:03 <peter1138[d]> Yeah ICU breaks on the wide colon already. Hmm.
21:05:48 <Rubidium> xarick: I'd say 81 (assuming 16x16 tiles); if you make a grid, you'd have 7 by 7 whole patches in the center and at the border you'd have 9 patches along each edge, counting the corners double
21:06:29 <xarick> wasn't expecting an answer
21:10:54 <truebrain> _jgr_: how does that look in JSON?
21:11:46 <truebrain> (for NewGRFs, we make a new vector and fill it up like that)
21:12:01 <truebrain> owh, the other PR changes that
21:12:04 <peter1138[d]> Hmm, 0x3000 = IDEOGRAPHIC SPACE... what is this.
21:12:54 <_jgr_> truebrain: It's exactly the same, the temporary vector doesn't add anything
21:13:03 <truebrain> so "as a list", cool 🙂
21:13:13 <truebrain> did not know nlohmann also had conversions for spans 😄
21:13:40 <Rubidium> peter1138[d]: I guess the space variant of the wide colon, period and comma used in Chinese
21:14:19 <peter1138[d]> Ah, so the fixed-width space.
21:15:25 <peter1138[d]> Oh damn, I touched string_func.,h 😦
21:15:34 <peter1138[d]> 481 files to compile.
21:18:35 <LordAro> peter1138[d]: get you a threadripper
21:18:58 <peter1138[d]> Does that help me get tubeless tyres on the rim?
21:19:28 <LordAro> i imagine it could serve as a tyre lever
21:19:33 <truebrain> _jgr_: hmm ... why do you only assign it when there is something in there?
21:19:45 <truebrain> it might be easier to always set the key, just from a reading perspective
21:20:01 <truebrain> the same happens in NewGRFs; even an empty list is just set
21:20:20 <_jgr_> Mostly because that's how it's done in GraphicsSetSaveConfig
21:20:43 <_jgr_> I can change it to always write something
21:20:56 <truebrain> hmm .. if you can, I think I would prefer that
21:41:28 <LordAro> that is not helping your situation
21:42:29 <peter1138[d]> My zero-width space is not quite right 😄
21:42:59 <_glx_> missing handling in FormatString ?
21:43:40 <truebrain> peter1138[d]: wh?at?do?yo?um?ea?n??
21:43:44 <_glx_> hmm no it's a DrawString thing
21:43:45 <peter1138[d]> It still tries to draw it 🙂
21:43:57 <peter1138[d]> But it doesn't exist as a sprite character, so gets ?
21:44:18 <peter1138[d]> I added it to ZWSP to FormatHexBlahBlah for ease of testing 🙂
21:44:29 <_zephyris> Zero width space with a sprite font which has a minimum width of 1... Sounds like fun! 😉
21:45:12 <peter1138[d]> Just needs to know not to print it, but excluding it from IsPrintableChar() prevents breaking from working I think.
21:45:22 <peter1138[d]> (Or it gets removed earlier)
21:45:33 <peter1138[d]> It works with ICU though.
21:45:50 <_glx_> if it's not printable FormatString removes it
21:46:07 <_glx_> so layouter doesn't see it
21:46:07 <peter1138[d]> Figures. Then it doesn't exist to be treated as a break 🙂
21:47:04 <_glx_> it must be printable, but with an empty glyph if not found
21:48:55 <peter1138[d]> Well, gfx_layout_fallback should probably remove it before it gets to the sprite drawing part of DrawString.
21:51:38 <peter1138[d]> It's not. It's just a normal character that doesn't break... and happens to be a devoid of pixels.
21:52:28 <Rubidium> I guess a zero width sprite wouldn't work, would it?
21:53:01 <peter1138[d]> Not sure zero width sprites are possible.
21:53:22 <peter1138[d]> There is zero space in this tyre to get it over the rim 😦
21:59:09 <Rubidium> yeah, both nforenum and grfcodec dislike zero width sprites
22:00:14 <Rubidium> though... the pre-layouter that makes the runs for the different font colours could replace the zero width space with the creation of a new run
22:00:45 <peter1138[d]> It wouldn't be zero-width then
22:01:33 <Rubidium> does changing colour add spaces?
22:03:08 <peter1138[d]> I think I misrea d:)
22:12:06 <_glx_> hmm no changing font doesn't cause au split
22:13:45 <Rubidium> doesn't cause a split for the Windows implementation you mean? For the fallback it definitely ought to. Not sure about the others though
22:14:54 <_glx_> yeah uniscribe just put as much runs as possible then look for valid break location
22:15:15 <_glx_> and fallsback to last letter if nothing better
22:17:21 <peter1138[d]> And ICU just messes up 🙂
22:18:25 <Rubidium> my trick works with fallback and ICU
22:22:08 <xarick> i think i found a big bug in yapf
22:25:05 <xarick> oh, no bug, looks like int is big enough to store gigantic n.m_costs
22:25:31 <xarick> there's potential for overflow there
22:32:22 <xarick> i broke the pf somehow
22:33:00 <xarick> regions with diagdirection increase the number of nodes to search
22:35:46 <LordAro> xarick: potential for overflow was one of the reasons we were thinking of getting rid of the settings entirely
22:37:05 <xarick> someone should make the longest corridor possible on a 4k map
22:37:19 <xarick> so i can test extract the total cost
22:38:19 <xarick> maybe a GS that terraforms the map layout
22:38:26 <peter1138[d]> Rubidium: It disappears before it's treated as a space.
22:39:18 <peter1138[d]> Ah, no, it causes the broken text to be replicated, and broken at any character.
22:39:45 <peter1138[d]> Hmm, maybe I'm not up to date.
22:40:52 <andythenorth> goes it sleeping?
22:41:17 <peter1138[d]> Not until I get this tyre on.
22:41:26 <andythenorth> hmm this is annoying 😛
22:41:29 <peter1138[d]> Although I might give in and put the old tyre back.
22:41:31 <xarick> `return this->m_water_region_patch.label | (GetWaterRegionIndex(this->m_water_region_patch) << 8) | (this->m_water_region_enter_dir << 24);` is this a bad hash?
22:41:35 <andythenorth> all I did was reload the grf, having removed some engines
22:41:41 <andythenorth> goes it crash some trains?
22:42:10 <_glx_> don't change newgrf in a save
22:43:13 <xarick> labels go from 0-255, water region index go from 0-65535, enter_dir goes from 0-255
22:43:47 <andythenorth> don't change newgrf in a save, after deleting some engines, then reusing their IDs for wagons 😛
22:44:29 <andythenorth> oh this one is doing 1mph
22:44:40 <andythenorth> the other 'no power' train must be because railtypes 🙂
22:45:15 <peter1138[d]> Hehe, yeah, it needs the recent fix to fallback layouter 🙂
22:55:56 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:21:44 <wensimehrp> @glx could you please reopen issue #11752? The line problem is not solved yet.
23:22:55 <peter1138[d]> The bug as described in #11752 is not fixed?
23:23:19 <wensimehrp> This is the fixed version
23:23:29 <peter1138[d]> Yes. That is correct.
23:23:39 <peter1138[d]> The duplicate character, which was the subject of the bug report, is fixed.
23:24:14 <peter1138[d]> You can't really tack on another bug you found after the original bug was fixed, and then claim the first bug isn't fixed 🙂
23:24:25 <wensimehrp> peter1138[d]: the report contains two things actually
23:24:49 <wensimehrp> One is the duplicated character, another is the line problem
23:25:20 <wensimehrp> Both were mentioned in the issue
23:35:04 <_glx_> the other one is unfixable without changing the translated string
23:36:44 <_glx_> we ask windows where we can break the line
23:36:45 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:37:35 <_glx_> and the only valid locations it gives is between the ideograms
23:41:35 <_glx_> newgrf filename is the worst case for layouter 🙂
23:47:46 *** gelignite has joined #openttd
23:56:03 <xarick> i complexicated CYapfDestinationRegionWaterT
23:56:29 <xarick> added support for multiple destinations
continue to next day ⏵
