IRC logs for #openttd on OFTC at 2024-12-25
⏴ go to previous day
01:16:56 <michi_cc> Xitter/Reddit announcement done BTW
01:24:59 <belajalilija> johnfranklin: Some celebrate it on the 24th/8th
02:44:58 *** Wormnest has quit IRC (Quit: Leaving)
03:25:03 *** gnu_jj_ has joined #openttd
03:28:26 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:35:41 *** debdog has quit IRC (Ping timeout: 480 seconds)
07:39:03 *** _zephyris has joined #openttd
08:00:15 <wensimehrp> Happy holidays guys!
08:32:58 <truebrain> and published Steam News
08:35:30 <wensimehrp> is there a way to use custom soundfonts?
08:35:43 <wensimehrp> and apply e.g. reverb effects?
08:54:18 *** gelignite has joined #openttd
08:57:20 <LordAro> wensimehrp: i believe fluidsynth has an env var you can set
08:57:31 <LordAro> the audio driver might have a parameter too
08:57:50 <wensimehrp> ah thanks, I'll check
09:22:43 <truebrain> well, that was a very tough nut to crack 😛
09:23:40 <truebrain> turns out the general rule still holds: if you have a problem you are trying to solve with regex, you have 101 problems 🙂
09:25:28 <truebrain> with 15.0-beta1. I do have the latest OpenGFX. Should I be worried? 🙂
09:32:32 <truebrain> at least let's make sure we don't link to 404 pages in news posts 😄
09:33:06 <truebrain> the aftermath of a simple file rename 😛 😛 😛
09:33:23 <truebrain> (it is okay, really; shows how fragile some parts of our software is 😄 )
09:42:40 <peter1138> Yeah, sorry about that, I really fucked it up.
09:43:13 <truebrain> Nothing to be sorry about, and you didn't fuck it up 🙂
09:43:33 <truebrain> It just turns out that having undocumented shell-scripts that use magic regex, is a terrible idea 😄
09:44:48 <truebrain> I am however dissapointed that browsers still don't render markdown files in any fancy way
09:46:31 <truebrain> Hmm, while at it, should we fix this to be less ... like that?
09:46:52 <truebrain> maybe `git log --oneline`?
09:51:08 <truebrain> okay, enough fiddling with these things 😄
09:53:54 <truebrain> sometimes you just look back on those things are you are like: why on earth did we do it like this in the first place 😛
09:55:22 <LordAro> iirc nightly changelogs weren't implemented for ages after the move, so don't think you can blame that :p
09:55:54 <truebrain> No, without joking, the subversion nightlies also had this very verbose changelog
09:56:53 <truebrain> But now the changelog is a markdown file ... we have better opportunities to do this right 😄
10:00:35 <merni> My 12am PR description was not clear? How shocking 😆
10:00:59 <truebrain> It is really appreciated you tackling this problem 🙂
10:03:53 <truebrain> I am trying to find who actually wrote `changelog.sh`
10:04:14 <truebrain> LordAro approved it!
10:04:24 <truebrain> but I really can't remember ever writing such an ingenious piece of `awk` 😛
10:04:31 <merni> Impressively concise is what I can say about it 😆
10:04:53 <truebrain> it is horrible, is what I think about it 😛
10:07:11 <truebrain> owh, right, before that time we used our own compile-farm
10:07:17 <truebrain> so that script most likely always lived somewhere on a VM
10:07:25 <truebrain> which means it is more likely that someone like Rb wrote it 😛 😛 😛
10:07:30 <truebrain> (always make sure you have someone else to blame!)
10:08:16 <truebrain> there is no git history of those scripts used in the compile-farm
10:09:44 <truebrain> 2018 we migrated to Azure Pipelines, 2019 we improved that drastically .. 2020 we migrated to GitHub workflows 🙂
10:10:05 <kuhnovic> Time to move to zip files on dropbox
10:10:22 <truebrain> at least better than maintaining your own compile farm! 😛
10:10:48 <truebrain> or before that, around 2007 or so, it was even better: when there was a release, several different developers had to manually build that release into a release binary for their OS
10:10:55 <truebrain> which could take weeks 😄
10:12:02 <kuhnovic> At least that is somewhat of a self-maintaining compile farm 😛
10:12:41 <kuhnovic> Just the "X is on holiday so we can't release because nobody else can compile for MacOS" 😆
10:12:51 <truebrain> owh, but we did release!
10:13:00 <truebrain> just platform Y had to wait N days before they could enjoy it
10:13:29 <truebrain> MacOS, BeOS, OS2/Warp, Debian ... they were often delayed with days, if not longer 😛
10:13:40 <truebrain> and hopefully of the same revision as the other platforms
10:14:52 <kuhnovic> I bet it worked fine most of the time, but it is a pain in the ass haha
10:16:44 <truebrain> The reason around 2007 or so we started our own CF; most notable, cross building for MacOS, was was considered "impossible" in 2007 🙂 releases improved a lot by it 😄
10:16:59 <truebrain> But happy we now run on GitHub 😄
10:18:16 * Rubidium finds it awkward to get blamed for such scripts ;)
10:20:46 <Rubidium> though I would have blamed LA instead, after all... Co-authored-by: ... :)
10:21:15 <truebrain> Yeah, but the script existed on our old CF already, I think
10:21:30 <Rubidium> because as far as I am aware I've not been involved with the migration to git
10:21:44 <truebrain> No, this script existed in subversion time already
10:21:57 <truebrain> Well, these two lines 😛
10:22:45 <truebrain> So either me, you or frosch wrote it 😄 or a combination of those 3 😄
10:26:38 <peter1138> "Blame" vs "You did a good job writing a script that we only needed to look at a decade later..."
10:27:23 <truebrain> A very glass half full approach, and you are absolutely right 😄
11:21:42 <michi_cc> GOG (and Windows Store), BTW, or don't they get testing releases?
11:21:48 <michi_cc> And merry christmas, too 🙂
11:38:59 <truebrain> Owh, GOG .. I forgot
11:39:19 <truebrain> I have to manually press buttons for GOG, and it isn't in my system for some reason 🙂
11:52:48 <xarick> codeql is very strick about unused variables
11:59:52 <truebrain> GOG fixed; 15.0-beta1 now published under Testing
11:59:57 <truebrain> tnx michi_cc , I really forgot GOG existed 😦
12:01:11 <peter1138> Did we even release 14.x on GOG? 😄
12:01:29 <truebrain> (as I wasn't sure 😛 )
12:03:40 <truebrain> okay, that script is 20 times more readable 😄
12:03:44 <merni> Where does `shell: bash` go?
12:04:09 <truebrain> between `id:` and `run:`
12:07:16 <truebrain> and no regex .. 😄 😄
12:12:03 <merni> Changing the format of non-tag changelogs... should that be in another PR or do you want it in this one?
12:12:51 <truebrain> In theory it should be in another PR, but there is always this balance of practicality; so I would just do it in this one, especially as it is only a single line of code 😛
12:13:42 <merni> But I don't know how to test it :p
12:13:51 <truebrain> execute the lines of code on your console 😄
12:14:12 <truebrain> `git log --after="${last_week}" --oneline` will work fine
12:15:46 <merni> Hmm, it gives weird stuff like `e78cd7d6e9 Update: Translations from eints swedish: 1 change by joeax910 chinese (simplified): 13 changes by tinygrox polish: 2 changes by pAter-exe`
12:15:53 <merni> Is all that part of the title?
12:16:29 <merni> I guess for a nightly changelog it's fine anyway :)
12:16:47 <truebrain> haha, I think there is a thing or two wrong with how eints makes commit messages in that case 🙂
12:16:59 <truebrain> yeah, missing newline
12:17:06 <truebrain> that nobody noticed that before is amazing 😄
12:17:13 <peter1138> The randomness strikes again
12:17:39 <truebrain> peter1138: even WinAmp already learnt that "randomize tracks" should not be actually a full random 😛
12:18:52 <Rubidium> it should be fully random, if you don't want fully random call it shuffle and actually shuffle...
12:22:11 <xarick> can you make CodeQL compiling faster?
12:23:33 <Rubidium> yes. Instead of focussing on making OpenTTD faster, put all that effort in CodeQL performance
12:24:17 <merni> merni-nsviaGitHub: me silly
12:25:21 <xarick> maybe not 20, but still much longer than the other comiping tests
12:25:41 <merni> Yes because it has to first compile the whole thing and then run the analysis
12:26:17 <truebrain> I still can't imagine nobody ever noticed this issue 😄
12:26:53 <merni> It doesn't show up on the web interface tbh
12:27:11 <truebrain> which is also weird 😛
12:27:16 <merni> Github thinks one newline ends the title, evidently
12:27:53 <merni> Github also has some other weird handling there... if a title is too long it will truncate and overflow to the description
12:28:01 <truebrain> am I right to assume you tested your Python script in various of ways?
12:28:25 <merni> Well, I added some dummy versions to the changelog and ran it repeatedly
12:29:08 <merni> Anyway, if it's broken it still won't be a regression :p
12:29:53 <truebrain> I can't believe the CI just worked on eints
12:29:56 <truebrain> I expected trouble there 😛
12:30:01 <xarick> okay, it's more than 20 minutes
12:30:11 <xarick> 90% compiled at 20 minutes 😦
12:30:50 <xarick> the other tests take like 6-8 minutes compiling
12:31:08 <xarick> what's so different about CodeQL compiler?
12:31:16 <merni> Now to look at the website thing
12:36:40 <Rubidium> xarick: check what CodeQL actually does, then you might get an idea why it's slower
12:37:02 *** D-HUND is now known as debdog
12:38:12 <truebrain> merni: there is also `openttd-useful-releases.md`; although that has been deprecated for years now, so not likely an issue 🙂
12:38:23 <merni> Whoops, I added it to the wrong end of the if
12:38:41 <merni> truebrain: Well, also I assume any new openttd-useful-release would have a markdown changelog
12:38:56 <truebrain> it has nothing to do with OpenTTD itself 😛
12:39:02 <truebrain> it was a package for MSVC to get going 🙂 Contains libraries
12:39:05 <truebrain> we replaced it with vcpkg 😛
12:39:14 <merni> I thought it was a useful release of openttd 😆
12:39:34 <truebrain> anyway, I think it would be fine to use a contains
12:39:45 <truebrain> especially as Jekyll doesn't have a built-in "startswith" 🙂
12:40:20 <merni> Yes, I did look for that
12:41:18 <merni> Commit history is such a mess
12:41:27 <truebrain> we squash it anyway
12:43:09 <merni> Hmm, do we not do builds of catcodec?
12:43:31 <truebrain> not everything is migrated to the new CDN, mostly as they haven't seen new releases in years
12:44:31 <merni> That format was what one stackoverflow answer used, so I used it 😆
12:45:46 <truebrain> the YYYY-MM-DD will also work fine, Ruby's documentation says 🙂
12:48:26 <truebrain> closing it once, closing it twice
12:49:00 <truebrain> commit message was a bit out-of-date, but I didn't spot auto-merge was enabled (or that you had permissions 😛 )
12:49:27 <truebrain> tnx again for the help here! Much appreciated 🙂
13:03:33 <peter1138> truebrain: Oh, that'll be road waypoint graphics.
13:03:49 <truebrain> so do we need a new OpenGFX?
13:03:58 <truebrain> I always get lost in this part of OpenTTD, I have to admit 😄
13:05:27 <peter1138> Maybe not. Those sprites exist in openttd.grf
14:23:51 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
15:32:22 <michi_cc> truebrain: Since the change to always load openttd.grf before the actual baseset, missing sprites are technically no problem anymore. Doesn't mean the result looks good of course.
15:42:56 <merni> michi_cc: Should the warning be removed then, rather than giving a warning for something that isn't going to be fixed in OpenGFX?
15:46:32 <michi_cc> Well, it would basically guarantee that nobody will ever volunteer to update the basesets. But then this isn't really a change over now, so 🤷
15:51:20 <Rubidium> not to mention that people might not update OpenGFX, and then complain about weird sprites
16:05:06 <xarick> Controlled folder access feature messed up with OpenTTD install
16:07:31 <xarick> had to remove and re-add openttd.exe
16:07:48 <xarick> or it wasn't letting me even save the openttd.cfg
16:13:36 <xarick> could not send a folder to recycle bin because i lacked permissions, but I'm the sole administrator of my computer
16:14:47 <xarick> openttd install nearly broke windows... or it's just microsoft incompetence
16:15:03 <xarick> restarting windows fixed the issue, but this scared me
18:34:50 <kuhnovic> You probably just had a stubborn file lock
18:35:51 <kuhnovic> LockHunter is a great tool to get rid of those, or at least to check if your files are locked.
18:46:51 *** tokai|noir has joined #openttd
18:46:51 *** ChanServ sets mode: +v tokai|noir
18:48:52 <peter1138> Understanding basic computer security concepts also helps.
18:53:41 *** tokai has quit IRC (Ping timeout: 480 seconds)
19:55:47 <xarick> oh, it's been 2 months already?
19:59:45 <xarick> it seemed to be going the right direction
20:09:12 <Rubidium> though... at what cost? Rewriting all kinds of functions deep down that work just fine, just to please an overly ambitious check. And probably you haven't found all the problematic cases, so it would introduce many bugs due conversions not being done and conversion that have been done incorrectly. Given I barely ever seen TileAdd asserting, it might be much better to just remove that assert
20:09:55 <michi_cc> Features nobody asked for part 45678:
20:10:00 <Rubidium> I didn't fancy doing whack-a-mole for weeks. If you fancy it, just continue with that branch
20:15:09 <talltyler> Does that exclude 64-bit Windows, or am I misunderstanding something?
20:15:27 <talltyler> (Aren’t most Windows installations 64-bit these days?)
20:16:01 <michi_cc> Why would it exclude 64 bit? (Win32 is the prefix used for anything Windows since like whenever 🙂
20:16:33 <talltyler> Just an uneducated guess at what the 32 means 😛
20:17:58 <michi_cc> Well, historically speaking it means it is for the Win
20:18:41 <michi_cc> 64-bit windows is still using the Win32 API, just with a few changed data types.
20:19:25 <xarick> @Rubidium, just copied fix-overflow branch. I feel like experimenting stuff with it
20:33:00 <talltyler> …I am not smart enough or experienced enough to review this PR. 🙂
20:33:30 <talltyler> I am not challenging you with any of my comments, btw, just trying to wrap my head around it
20:33:32 <michi_cc> sizeof(bool) == 1, sizeof(BOOL) == 4, for the interested 🙂
20:34:09 <michi_cc> Well, C didn't have any real bool type for a long time, so other conventions evolved.
20:37:08 <Rubidium> michi_cc: lies... sizeof(bool) is undefined
20:38:13 <talltyler> Fascinating, thanks for the history lesson 😄
20:38:15 <michi_cc> Yes, but in the context of MSVC compiler (which is the only one that can compile the WinRT stuff anyway), it is 😛
20:38:43 <talltyler> (Not trying to make you feel old, I’ve only been writing code since 2020-ish)
20:39:23 <michi_cc> talltyler: Don't worry, I manage to feel old all by myself 🤣
20:42:46 * Rubidium hopes to make michi_cc feel young again... but I need to verify something for that to be sure ;)
20:45:30 <michi_cc> @Rubidium I think you are older than me, but I am firmly in the 39+ years club now 🙂
20:46:37 <Rubidium> from MSVC 4.2's includes: typedef int bool;
20:50:13 <michi_cc> I don't think MSVC that old even knew what the ISO C++ standard meant 🙂
20:52:49 <Rubidium> yeah probably. They were aware of ISO 9660 though ;)
21:10:43 <michi_cc> `((5 * foreground.G) + (2 * foreground.R) + foreground.B) > (8 * 128)` or `5 * foreground.G + 2 * foreground.R + foreground.B > 8 * 128` (i.e. lots of extra () in the code I copied from MS)
21:14:02 <talltyler> I’d say keep it as copied, and I like the ()
21:56:03 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:27:16 <xarick> I've been thinking, instead of asserting, return INVALID_TILE instead
22:28:16 <xarick> most situations seem to do a IsValidTile of some sort afterwards
22:52:37 <_zephyris> truebrain: I'm planning to do it, just need to find the time to get them codes... Already got some sprites made.
23:34:18 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:35:23 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:43:43 *** gelignite has quit IRC (Quit: Stay safe!)
continue to next day ⏵
