IRC logs for #openttd on OFTC at 2021-06-15
⏴ go to previous day
00:41:26 *** tokai|noir has joined #openttd
00:41:26 *** ChanServ sets mode: +v tokai|noir
00:48:19 *** tokai has quit IRC (Ping timeout: 480 seconds)
00:58:27 *** ChanServ sets mode: +v tokai
01:05:19 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
01:21:53 *** Wormnest_ has joined #openttd
01:49:10 *** Wormnest has quit IRC (Quit: Leaving)
01:51:29 *** tokai|noir has joined #openttd
01:51:29 *** ChanServ sets mode: +v tokai|noir
01:58:18 *** tokai has quit IRC (Ping timeout: 480 seconds)
04:02:23 *** outtatime has joined #openttd
04:08:40 *** bro has quit IRC (Ping timeout: 480 seconds)
04:13:04 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9371: Assorted C-string -> std::string conversions in network https://git.io/JZxHY
05:12:16 *** roadt__ has quit IRC (Ping timeout: 480 seconds)
05:41:03 *** snail_UES_ has quit IRC (Quit: snail_UES_)
05:47:36 *** ChanServ sets mode: +v tokai
05:54:11 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
05:57:41 *** nielsm has quit IRC (Ping timeout: 480 seconds)
06:16:28 *** sla_ro|master has joined #openttd
06:25:28 *** TinoDidriksen has joined #openttd
06:25:59 *** TinoDidriksen is now known as Guest2205
06:26:12 *** Guest2205 is now known as TinoDidriksen
06:46:00 *** andythenorth has joined #openttd
06:46:48 <andythenorth> where's the fun? :P
06:47:12 <andythenorth> if I swap out the npm for pip, and get it to run my makefile...
06:47:18 <andythenorth> oh and some crap about versioning
06:47:28 <andythenorth> and not over-writing previous versions
06:50:23 <andythenorth> the distributed size of the site is currently around 500MB, and will grow with every grf release
06:50:35 <andythenorth> I don't want to be re-publishing 500MB on every tag?
07:01:50 <andythenorth> I like that in the modern world, I give access to my S3 bucket to a GHA plug in made by unknown 3rd parties with possibly no accountability
07:26:59 *** HerzogDeXtEr has joined #openttd
07:37:07 <TrueBrain> I see you are having fun :P
07:50:09 <Xaroth> andythenorth: form that page: `In the next view, select Attach existing policies directly and check AdministratorAccess. Note that this will give the account access to EVERYTHING in our AWS account.`
07:50:36 <TrueBrain> lolz, really? Well .. that is a good way to introduce people to GitHub Actions with AWS :D
07:50:43 <Xaroth> They should have renamed that blog post "How to create a giant security hole in your AWS environment in one easy step"
07:51:02 <TrueBrain> and adding roles is really not difficult ..
07:51:15 <andythenorth> I will not be copy-pasting those instructions
07:51:29 * andythenorth must save up this fun for the weekend
07:51:47 <TrueBrain> andythenorth is a smart cookie
07:52:05 <Xaroth> I dread to guess how many people _do_ blindly copy those instructions
07:54:47 <LordAro> freenode real dead now
07:57:27 <reldred> Lel, I noticed I got banned from Freenode for the heinous crime of... *checks notes* using IRCCLoud?
08:00:26 <TrueBrain> LordAro: how nice of them they helped out with that
08:00:49 <TrueBrain> reldred: how DARE you! They have their own (FREE!) services! It is an FOSS community, you should be using THEIR service
08:00:57 <TrueBrain> wait .. something went wrong in that sentence .. not sure where ........
08:02:42 <TrueBrain> I cannot imagine I would ever do a "migration" like this
08:02:53 <TrueBrain> but it is a nice way for us to get real stats about how many people still use freenode :P
08:03:52 <SpComb> it seems so delusional
08:04:08 <TrueBrain> I wonder what is going on in his head
08:04:19 <TrueBrain> "everyone will be fine with this, they love me, and they will just come with me" ?
08:06:14 <reldred> The sort of people who like to smoke out of hollowed out lightbulbs I'm guessing.
08:07:01 <TrueBrain> well, guess we will find out how many people "migrate" with him :P
08:12:54 <andythenorth> crack is bad for you
08:13:00 <andythenorth> quite moreish though
08:13:25 <andythenorth> about the same yes
08:22:33 <TrueBrain> @calc 3146381 - 3145956
08:22:40 <TrueBrain> @calc 3146352 - 3145983
08:23:05 <TrueBrain> @calc 3146381 - 3145983
08:23:10 <TrueBrain> @calc 3146352 - 3145956
08:23:28 <TrueBrain> for some reason the savegame doesn't pick up that I read 2 bytes too few from a chunk, and I do not know why :P
08:24:24 *** andythenorth has quit IRC (Quit: andythenorth)
08:26:59 <peter1138> I'm sure your DHCP settings would've told you that.
08:27:11 <TrueBrain> that joke really never gets old :D
08:28:55 <peter1138> How badly is joining Samba to a Windows DC going to go?
08:33:09 <SpComb> there's documenys from a 2016-2017 lawsuit going around now with the new freenode owner as the defendant, and it certainly paints a pretty poor sense of judgement. Hookers and crack quite literally, no mention of blackjack though
08:33:36 <TrueBrain> ah, lol .. we have a string in the settings-chunk ... and that is gladly accepting any misreads from before in that chunk
08:33:38 <TrueBrain> that is a bit evil :D
08:35:24 <TrueBrain> there is a length-byte and everything
08:36:35 <TrueBrain> so I read 2 bytes too few, so the length-byte is read 2 positions early .. and it seems it contains a value of the string length + 2
08:36:41 <TrueBrain> so by pure luck, it loads in fine
08:47:43 <TrueBrain> no clue what we can do about it, sadly
08:48:32 <peter1138> So it's a problem with the current savegame design?
08:49:31 <TrueBrain> basically .. and not sure there is a design that would prevent it
08:49:49 <TrueBrain> if the bytes roll in your favour ..
08:50:07 <peter1138> How does it "just work" right now then?
08:50:47 <TrueBrain> not sure I follow the question, but this only happens when the data is bad, unexpected
08:50:51 <TrueBrain> so a developer that has been zzzzzing
08:51:38 <TrueBrain> in this case, master writes 2 extra bytes, and my PR wasn't updated for it
08:51:58 <TrueBrain> and by luck the fields aligned that no integrity issue was detected
08:52:35 <TrueBrain> well, after my PR the system should be a lot more resilient here, so I guess that counts for something
08:52:40 <peter1138> Oh right so it's specifically because you missed a savegame update?
08:53:11 <peter1138> Could always add CRCs...
08:53:27 <TrueBrain> well, a CRC wouldn't pick it up either, not a normal data-CRC anyway
08:53:29 <peter1138> Maybe not, the raw bytes would be the same..
08:53:36 <TrueBrain> it is a fun little problem
08:53:41 <TrueBrain> that can also happen by bitflips
08:53:58 <TrueBrain> but yeah, this was 100% my issue
08:54:04 <TrueBrain> I was just surprised it didn't refuse the savegame
08:54:14 <peter1138> If the new format solves the problem going forward, I don't think it's problem.
08:54:23 <peter1138> Er, phrase that to make sense :)
08:54:59 <Rubidium> you can't really fix those cases, as there's always a chance one thing can be validly be interpreted by another thing
08:55:10 <peter1138> With the new format, you can't miss a field any more, I guess. You'd have a field you don't know what to do with.
08:55:19 <TrueBrain> peter1138: yeah, exactly
08:55:28 <TrueBrain> so all savegames after this are somewhat protected against it
08:55:32 <TrueBrain> and indeed Rubidium
08:55:36 <TrueBrain> bits have no meaning till given one
08:55:49 <TrueBrain> just took a while to figure out if it wasn't a bug in my code :D
08:55:49 <peter1138> You are worrying about a non-issue I think :)(
08:56:01 <peter1138> Yeah, would've been nice for it to have been caught, but...
08:56:05 <TrueBrain> yeah, sorry, could have been a bit more clear, but I mostly wanted to validate it wasn't a bug :P
08:56:16 <peter1138> Alright, how much will having a corner desk annoy me?
08:56:19 <TrueBrain> but exactly on the one string we have .....
08:56:33 <TrueBrain> anyway, you are right peter1138 , non-issue :)
08:59:27 <TrueBrain> right, now for the fun part of this PR ... making the "this field is not here, skip it" part
08:59:33 <TrueBrain> as that "field" can be a struct
08:59:37 <TrueBrain> and in that struct can be ... another struct
09:02:46 <peter1138> Maybe I should look into kubernetes.
09:03:00 <TrueBrain> because you have a corner desk?! :P
09:03:11 <peter1138> Nah, thinking about work again.
09:03:19 <peter1138> Although I guess the desk is work too, as I work at it? Hmnm.
09:03:45 <TrueBrain> kubernetes is like regex
09:03:46 <peter1138> I used to have 3 monitors but they were 22" 16:10. I can't have 3 modern 16:9 monitors unless I have something really small.
09:03:47 <LordAro> do you have a corner to put it into?
09:03:48 <TrueBrain> if you use it to solve a problem
09:03:51 <TrueBrain> you just gained a few more
09:03:57 <peter1138> OR! I can switch to a corner desk which would give me wider space...
09:04:24 <TrueBrain> LordAro: you expect he is working in a circular room?
09:04:33 <peter1138> LordAro, yes, but I also have rudder pedals/yoke, or driving pedals/steering wheel, which can be a pain if the mounting isn't right.
09:04:47 <LordAro> might not necessarily be up a ggainst a wall
09:05:39 <peter1138> Okay, it's been 1 hour 5 minutes into my morning meeting, and my line manager hasn't turned up. I think it's okay to sack that one off..
09:05:59 <peter1138> (Teams video call, so no real hardship)
09:34:56 * peter1138 considers going back to the window-decorations-are-text system...
09:35:20 <TrueBrain> what do you consider decorations? :D
09:40:00 <peter1138> Standard window icons.
09:40:43 <TrueBrain> honestly, I would expected them to already be considered text, in terms of scaling etc :P
09:41:26 <peter1138> It was originally a mix, I standardized it. Possibly the wrong way...
09:41:36 <peter1138> (This was well over a decade ago)
09:42:02 <TrueBrain> past-us biting us in the buttttttt :P
09:42:43 <peter1138> But yes, something like fontawesome (perhaps custom) is where I was thinking.
09:44:22 <peter1138> I already drew some vectors of the transport-type icons for stations, though the sizing might be out of whack. They are actually text anyway though.
09:44:53 <peter1138> That's the other thing. Switching fonts mid-flow might be interesting.
09:45:23 <TrueBrain> the whole font system could use a ... fresh-up
09:47:32 <peter1138> I should revisit my scrollbar-iterator patch.
09:47:48 <TrueBrain> having issues focusing? :D
09:47:56 <peter1138> Why do you say that? :D
09:48:57 <peter1138> Anyway, window decoration was coming from my window scaling work, which also involved changing how padding is handled, and also scrollbars and then made me remember my iterator stuff.
09:49:05 <peter1138> So a link. Tenuous but a link.
09:49:14 <TrueBrain> :D And down down we go :P
09:53:56 <Rubidium> does that also make the scrollbar size_t ?
09:54:22 <TrueBrain> didn't Aro make a PR for that?
09:55:27 <LordAro> it did originally, i tried to scale it back
09:56:30 <Rubidium> so we still need to cast to (uint) everywhere ;)
09:56:58 <TrueBrain> happy my memory is not that shitty that I misremembered size_t :P
09:57:00 <peter1138> It allows something like like `for (auto it : sb->Iterate(list))`, so you don't need to case about indices.
09:58:27 <peter1138> I got stuck at one of the windows that combines two different lists :D
09:58:44 <TrueBrain> always this one exception .....
09:58:56 <TrueBrain> can we nuke the window? :P
10:03:54 <TrueBrain> ugh, one of the downsides of working in saveload .. any change to saveload.h recompiles a good portion of the game
10:09:14 <TrueBrain> looking at your "git stash" is funny sometimes
10:09:19 <TrueBrain> "how did I end up with 10 stashes?!"
10:15:00 <LordAro> and you never know what they are
10:16:50 <TrueBrain> I think most remain behind when I do a "git stash pop" and there is a conflict
10:16:55 <TrueBrain> as it keeps the entry also in the stash
10:17:04 <TrueBrain> but yeah ... code of which you are like: huh?!
10:18:39 <DorpsGek> [OpenTTD/OpenTTD] Phyrik commented on issue #8252: Possible duplication of functionality in `screenshot.cpp` https://git.io/JJUWw
10:22:40 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8252: Possible duplication of functionality in `screenshot.cpp` https://git.io/JJUWw
10:25:13 <DorpsGek> [OpenTTD/OpenTTD] Phyrik commented on issue #8252: Possible duplication of functionality in `screenshot.cpp` https://git.io/JJUWw
10:30:12 <TrueBrain> hmm ... if there is a struct in the savegame not known to OpenTTD, it should skip reading it
10:30:21 <TrueBrain> but this is a lot more difficult than I expected :P
10:47:32 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #9369: Fix: don't show join button in online players window when the company is an AI https://git.io/JZNiJ
10:48:34 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9369: Fix: don't show join button in online players window when the company is an AI https://git.io/JnqPO
11:03:07 <TrueBrain> ha, that works :D Okay, this is cool
11:03:11 <TrueBrain> I now only have a memory leak
11:15:19 <TrueBrain> so what do you do when you need to create objects on the fly but don't want to bother with deleting them?
11:15:24 <TrueBrain> you make it an unique_ptr!
11:15:31 <TrueBrain> but .... this object is copied around a lot
11:16:17 <TrueBrain> shared_ptr it is! :P
11:17:54 <peter1138> TrueBrain, don't you store the size of the struct to just skip it?
11:18:16 <TrueBrain> we store the header of the struct
11:18:20 <TrueBrain> so we can calculate the length
11:18:25 <TrueBrain> there can be arrays in there
11:18:33 <TrueBrain> so we cannot just skip it based on the size
11:18:37 <TrueBrain> we actually have to read through it
11:19:09 <peter1138> Right, and the header is only the header, not each... record.
11:19:20 <TrueBrain> and array-length is per record ofc
11:19:38 <TrueBrain> so I need a "Skip" handler for the struct that is missing
11:19:41 <TrueBrain> load the data in there
11:19:44 <TrueBrain> and it works as expected
11:21:25 <peter1138> You could perhaps add a size to the struct record, however what you are doing seems to be better as self-describing is the whole purpose. Even if it is more fiddly...
11:21:37 <peter1138> You could just abort and say it's corrupted :D
11:21:48 <TrueBrain> well, mostly it was a matter of memory management I had to figure out
11:21:56 <TrueBrain> those dynamic objects need to be removed at some point :P
11:22:07 <TrueBrain> but shared_ptr seems to be doing fine
11:22:54 <peter1138> exit() will free them
11:23:24 <TrueBrain> just so awkward if you load a savegame 10 times, your memory keeps increasing :P
11:23:39 <peter1138> Unused memory is wasted memory!
11:23:41 <TrueBrain> (we are talking bytes here, but whatever)
11:24:42 <TrueBrain> okay, valgrind doesn't complain either
11:24:49 <TrueBrain> ==8279== total heap usage: 921,892 allocs, 910,779 frees, 9,191,432,701 bytes allocated
11:24:55 <TrueBrain> I ... how many bytes do we allocate? Lol
11:25:21 <TrueBrain> I do have a lot of NewGRFs, yes :D
11:25:31 <peter1138> Plus we seem to do several loads when starting up the game.
11:26:23 <peter1138> Ooh, just found my newgrf cache patch, that stores the cached data and avoids all newgrf-based desyncs.
11:26:29 <peter1138> Except it doesn't work at all.
11:27:55 <peter1138> It correctly detects the changes, puts the saved data back, and then... still desyncs!
11:30:22 <TrueBrain> and ... SLE_NULL / SLE_CONDNULL is gone :D
11:30:30 <TrueBrain> all this build-up for being able to do just that :P
11:33:20 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9375: Change: make as many savegame chunks a CH_ARRAY as possible https://git.io/JnUze
11:33:29 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Add: store table header for each chunk in savegame https://git.io/JGlmd
11:36:43 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/Jnq7m
11:37:01 <TrueBrain> just for niceness, lets ensure the test-suite still works fine
11:39:07 <peter1138> Perfect. Justification for "breaking" the "guaranteed" immutable savegame format that changes every update anyway... remove SLE_*NULL.
11:40:19 <TrueBrain> wow, that is a hard sentence to parse :D
11:56:45 <TrueBrain> orudge: GOG just send the nicest email :D Something for you to pickup, but that is super nice of them to communicate it like this :)
11:57:13 <TrueBrain> (they basically made it more clear that we are a Free game)
11:58:16 <TrueBrain> I like GOG .. they actual seem to care about the games on their store :)
12:00:53 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9369: Fix: disable join button in online players window when the company is an AI https://git.io/JZNiJ
12:12:31 *** tokai|noir has joined #openttd
12:12:31 *** ChanServ sets mode: +v tokai|noir
12:19:30 *** tokai has quit IRC (Ping timeout: 480 seconds)
12:24:52 <jez> ^ is this calculation intentional? 6 overlapping pieces of track on a tile is as much maintenance as 36 pieces of straight track? seems a bit goofy.
12:25:26 *** snail_UES_ has joined #openttd
12:28:03 <Rubidium> well, a normal slip switch is already quite complex and more expensive to maintain. I wouldn't want to think about how a slip switch with 6 paths would look, but that sounds super complicated and as such expensive to maintain
12:28:50 *** WormnestAndroid has quit IRC (Remote host closed the connection)
12:29:06 <Rubidium> I know that here two normal switches are preferred over a slip switch purely on the maintenance costs alone, as slip switches are more prone to issues and thus more costly to maintain (and fix)
12:29:41 *** WormnestAndroid has joined #openttd
12:29:43 <Rubidium> the only reason to use slip switches is when there is no other choice due to the restrictions on size of the switches
12:43:11 <peter1138> Rubidium, realism eh? :D
12:43:19 <TrueBrain> I like how he could explain it
12:43:26 <TrueBrain> I would never have come up with that .. reason? :D
12:45:28 <LordAro> it's like it's his job or something
12:45:43 <LordAro> emphasis on the or something
12:50:34 <glx> and my only chan on freenode switched to discord today, seems I don't need to register on the new freenode
12:54:34 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
12:54:43 <TrueBrain> glx: sounds like you are broken up because of it
12:55:09 <TrueBrain> I think I covered it all ..
12:55:54 <glx> I was expecting a switch at some point when I saw the 3 globals last night :)
12:57:12 <glx> after stealing channels they decided to do a full wipe
12:59:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
13:01:09 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
13:01:12 <TrueBrain> okay, enough bikeshedding, lets first wait for a review before I keep pushing over and over :P
13:01:35 <glx> TrueBrain: [4..5] is 2 bytes, not 1 ;)
13:01:53 <TrueBrain> ugh .. I did it ALMOST right :P
13:02:44 <TrueBrain> not sure how useful they are, honestly
13:02:45 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
13:02:47 <TrueBrain> as with gammas it becomes a bit weird
13:04:27 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
13:07:18 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
13:29:52 <glx> I'm running a "quick" test on CI (building x64 MacOS RelWithDebInfo)
13:31:51 <LordAro> "denied, too many translators"
13:33:15 <LordAro> i thought there was supposed to be a different dorpsgek message when there's more than 10 existing translators?
13:33:47 <LordAro> oh no, just a different issue template
13:33:52 <planetmaker> there was a notification that the translator group is full next to the 'apply' button. But I applied nonetheless :D
13:34:36 <planetmaker> must be one of the quickest accepted issues :P
13:34:45 <glx> and yes DorpsGek should also tell us when there's too many translators
13:36:39 <Rubidium> LordAro: how I wish slip and three way switches were not a thing. That would make my work a lot simpler ;) On the other hand, it being complicated makes it a lot harder to replace me by someone else... so it's a sort of job protection
13:37:48 <glx> so apple compiler is broken and we need to work around it
13:44:35 <glx> ok let's try with -v (as the message suggests)
13:50:01 <milek7> disabling lto would maybe help?
13:55:28 *** virtualrandomnumber has joined #openttd
13:55:39 *** virtualrandomnumber has quit IRC ()
13:56:44 *** WormnestAndroid has quit IRC (Remote host closed the connection)
13:56:57 *** WormnestAndroid has joined #openttd
14:07:22 <Samu> crap, i can't test my ai in 1.11.2
14:08:35 <glx> what prevents you from testing in 1.11.2 ?
14:21:42 <TrueBrain> I tried so hard not to make a remark about it :P
14:24:37 *** ChanServ sets mode: +v tokai
14:27:16 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9374: Change: make lists in savegames consistent https://git.io/Jnm86
14:28:00 <TrueBrain> stupid coding style :P
14:31:20 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
14:31:29 <TrueBrain> at least I was consistent(ly wrong)
14:32:13 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 approved pull request #9374: Change: make lists in savegames consistent https://git.io/Jnm4S
14:40:36 <Samu> when compiling openttd, antivirus service takes like 1/3 of the cpu resources
14:42:08 <LordAro> gotta check those object files for viruses
14:42:27 <LordAro> you should be able to exclude certain folders from Defender's scanning
14:42:34 <TrueBrain> you know the old saying ... what if you infect xgcc with a virus, how will you ever know if you compile everything with gcc? :D
14:43:31 <LordAro> not so much an old saying as a famous paper by Ken Thompson
14:44:32 <TrueBrain> does one exclude the other? :D
14:45:32 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9375: Change: make as many savegame chunks a CH_ARRAY as possible https://git.io/JnUze
14:45:39 <TrueBrain> I keep hoping no commit drops between the rebases :P
14:55:17 <TrueBrain> " On master, 285 files have changed and there have been 12,501 additions and 3,219 deletions. "
14:55:21 <TrueBrain> ahum .... right ... so that is happening
14:55:41 <peter1138> I like to remove code :D
14:56:16 <LordAro> how much of that is fmt?
15:09:26 <peter1138> TrueBrain, damn, saving dequeues was my piece of evilness, sorry :/
15:09:54 <TrueBrain> owh, yeah, you were not alone there
15:10:02 <TrueBrain> I just made it consistent .. could have gone either way :P
15:10:05 <TrueBrain> it was a bit 50/50 :D
15:10:24 <TrueBrain> otherwise I could easily reuse it for vector :)
15:10:35 <TrueBrain> otherwise? other than that
15:14:40 *** tokai|noir has joined #openttd
15:14:40 *** ChanServ sets mode: +v tokai|noir
15:21:24 *** tokai has quit IRC (Ping timeout: 480 seconds)
15:30:17 <glx> outch map_sl is conflicting very hard
15:30:31 <TrueBrain> what did you do to it?
15:31:24 <glx> well I just changed indentation, and some function renaming when enclosing in subclasses
15:31:44 <TrueBrain> I didn't touch map_sl a lot either
15:31:50 <TrueBrain> as fuck map-chunks :P
15:32:53 <glx> enough to conflict everywhere
15:32:54 <TrueBrain> forgot about that :P
15:33:05 <TrueBrain> s/SlArray/SlCopy/ :)
15:33:17 <TrueBrain> I only really changed MAPS
15:33:23 <TrueBrain> from CH_RIFF to CH_ARRAY to CH_TABLE
15:36:24 *** Wormnest has joined #openttd
15:37:28 <peter1138> Does that last batch of changes leave an invalid savegame state in the middle? (Not an issue, just trying to understand) -- version is bumped in a146bcfe, and then a version test added two commits later.
15:39:25 <TrueBrain> it shouldn't, but I have not validated it
15:40:37 <TrueBrain> ah, no, that PR, sorry, yes, in between is broken
15:41:14 <TrueBrain> every "Change" commit breaks it in that PR
15:43:05 <TrueBrain> I could have squashed it .. I could have bumped it every commit .. but I think this is more clear. Just when bisecting it could bite you
15:48:55 <jez> i've discovered a bit of a bug with v
15:50:13 <jez> if you pass in track bits including multiple rail pieces to be added to one tile, it doesn't calculate the rail pieces count correctly.
15:50:42 <jez> because "case MP_RAILWAY" isn't hit, pieces is just incremented by 1, when it should be incremented by number of rail pieces squared
15:51:03 <jez> (if the tile being added to is currently not a rail tile, of course)
15:52:48 <jez> there's probably a similar issue with road pieces
15:56:09 <peter1138> How are you adding multiple rail pieces on a bare tile?
15:56:37 <peter1138> So where's the bug?
15:56:57 <peter1138> Correct answer is... the patch.
15:56:58 <jez> clipboard. but still, a developer might well expect that function to actually increment the infra count appropriately given the track bits
15:57:02 <glx> so bug is not in openttd code
15:57:16 <jez> depends whether you think surprising behaviour from a function is a bug
15:57:52 <jez> if it's only valid to set one different track bit in a call to that method it should take a single enum value, not a bitfield
15:59:42 <jez> is there a way to build multiple pieces simultaneously?
15:59:58 <glx> and argument is not a bitfield, it's a Track
16:00:17 <glx> there's no way using the mouse
16:00:41 <peter1138> Yes, in master you cannot even call CmdBuildSingleRail with multiple tracks.
16:00:51 <peter1138> So it's definitely a bug in your patch.
16:00:57 <jez> that's changed then i guess
16:01:04 <jez> used to work (even if infra calculation didnt work)
16:02:00 <peter1138> If the patch changes the function to allow multiple tracks, then the patch should also change the function to take account of the calculation too.
16:02:37 <peter1138> TrueBrain, ok. Doesn't bother me, just understanding it : )
16:02:37 *** gelignite has joined #openttd
16:05:31 <TrueBrain> Anything else I can help with, let me know :)
16:15:50 *** Progman has joined #openttd
16:31:23 <_dp_> jez, are you making a new clipboard patch or updating some of the existing ones?
17:01:29 *** ChanServ sets mode: +v tokai
17:08:21 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
17:32:54 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 approved pull request #9375: Change: make as many savegame chunks a CH_ARRAY as possible https://git.io/JnYfh
17:33:55 <TrueBrain> wait, no comments? :o
17:36:11 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9375: Change: make as many savegame chunks a CH_ARRAY as possible https://git.io/JnYLN
17:36:18 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9375: Change: make as many savegame chunks a CH_ARRAY as possible https://git.io/JnUze
17:36:31 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9322: Feature: store table headers for each chunk in savegame https://git.io/JGlmd
17:37:00 <TrueBrain> compat-tables are lengthy :P
17:37:55 <glx> oh I fear another complicated rebase
17:38:31 <TrueBrain> and one more PR that will be complicated too
17:38:40 <TrueBrain> but I hope that doesn't come as a surprise? :D
17:39:10 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
17:52:49 *** andythenorth has joined #openttd
18:57:30 <DorpsGek> - Update: Translations from eints (by translators)
18:59:50 <andythenorth> game is surprisingly better without cdist
19:01:06 *** gelignite has quit IRC (Ping timeout: 480 seconds)
20:04:33 <glx> it's so fun to fix conflicts in one commit, to introduce conflicts in following commit
20:14:27 *** tokai|noir has joined #openttd
20:14:27 *** ChanServ sets mode: +v tokai|noir
20:21:26 *** tokai has quit IRC (Ping timeout: 480 seconds)
20:26:24 *** geli has quit IRC (Quit: Stay safe!)
20:55:50 *** Samu has quit IRC (Quit: Leaving)
20:56:43 <peter1138> Alright, three postgresql instances... I need to enable failover.
20:57:56 *** ChanServ sets mode: +v tokai
21:04:41 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
21:15:12 *** outtatime has quit IRC (Remote host closed the connection)
21:15:26 *** outtatime has joined #openttd
21:16:33 *** sla_ro|master has quit IRC ()
21:17:35 *** tokai|noir has joined #openttd
21:17:35 *** ChanServ sets mode: +v tokai|noir
21:18:42 *** outtatime has quit IRC (Remote host closed the connection)
21:18:50 *** outtatime has joined #openttd
21:22:44 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:24:13 *** tokai has quit IRC (Ping timeout: 480 seconds)
21:29:32 *** andythenorth has quit IRC (Quit: andythenorth)
21:36:44 *** Progman has quit IRC (Remote host closed the connection)
21:59:22 <TrueBrain> did they update software? :P
22:04:21 <glx> no, same image version (20210607.1)
22:07:32 <glx> hmm unless it's one of the deps from vcpkg
22:08:33 <glx> but everything should be the same as we use vcpkg version from the image
22:10:08 <glx> and indeed same versions, as it's in the cache
22:11:59 <peter1138> So this version was previously compiling okay but isn't now?
22:13:41 <glx> well there are 2 extra commits on top, but they should not affect the build in any way
22:21:33 *** ChanServ sets mode: +v tokai
22:28:21 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
22:50:01 *** Ttech has quit IRC (Quit: Este é o fim.)
23:01:50 *** Speeder_ has joined #openttd
23:09:05 *** Speeder has quit IRC (Ping timeout: 480 seconds)
23:18:28 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:50:07 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
23:51:10 *** WormnestAndroid has joined #openttd
23:54:31 *** nielsm has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
