IRC logs for #openttd on OFTC at 2019-03-07
⏴ go to previous day
00:18:21 <Eddi|zuHause> Beerbelott: so, that's a merge conflict, but what is "lost" there?
00:24:05 <Beerbelott> well it inherited a f*** up situation
00:25:56 *** Smedles has joined #openttd
00:29:57 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7165: [core] Implement SmallVector using std::vector https://git.io/fhpuF
00:35:59 *** supermop_work____ has quit IRC
00:36:35 *** michi_cc has joined #openttd
00:36:35 *** ChanServ sets mode: +v michi_cc
00:48:17 *** drac_boy has joined #openttd
01:02:33 <DorpsGek_II> [OpenTTD/OpenTTD] M3Henry updated pull request #7165: [core] Implement SmallVector using std::vector https://git.io/fhSz0
01:08:49 <m3henry> I guess that means it's nearly over, if you're reached the penultimate commit
01:35:24 <drac_boy> mm quiet night for coding .. anyway I'm going off for today still :)
01:37:51 *** Thedarkb-T60 has joined #openttd
01:44:24 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
02:28:22 <Eddi|zuHause> oh man this last portal chamber was hidden :/
02:58:48 <Samu> I see gamescripts can create subsidies, that's interesting
02:59:08 <Samu> but it could have a bit more control
03:05:03 <peter1138> Give it more control then.
03:11:14 <Samu> how many months it lasts
03:11:37 <Samu> and... how many months it lasts when completed
03:12:07 <Samu> but still, it's a game script slot occupied just for this :(
03:18:03 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on issue #7333: Libtimidity could be trivially enabled to play music through mixer.cpp https://git.io/fhpgX
03:46:13 *** Smedles has joined #openttd
04:29:58 *** Beerbelott has left #openttd
04:52:45 *** ChanServ sets mode: +v tokai
08:03:38 *** andythenorth has joined #openttd
08:26:02 <peter1138> At least 1 approving review is required by reviewers with write access.
09:48:50 *** m3henry has joined #openttd
10:07:22 *** Beerbelott has joined #openttd
11:27:04 <Eddi|zuHause> is that just me or is there a spam attack on the forum?
11:34:13 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7338: Fix: spelling for a few real town names https://git.io/fhpKm
11:49:02 *** Thedarkb-T60 has joined #openttd
11:53:32 *** Thedarkb-T60 has joined #openttd
11:53:32 *** Eddi|zuHause has joined #openttd
12:07:52 <LordAro> peter1138: there was a reason for the double quotes, something to do with the escape codes
12:08:01 <LordAro> let me see if i can find it...
12:10:21 <LordAro> oh wait, your comment has disappeared
12:10:32 <LordAro> guess you found out about "warning: hex escape sequence out of range" :)
12:12:25 <peter1138> I realised as I posted, didn't actually try it :)
12:17:58 <peter1138> Okay, here's one...
12:18:38 <peter1138> Why is road vehicles "Drive on left" / "Drive on right" so prominent in the Game Options window?
12:19:01 <peter1138> Town Names should also be a world-gen option.
12:21:47 <LordAro> because TTD, i imagine
12:26:42 <peter1138> watching trees grow and die on ffwd is mesmerising :p
12:33:00 <andythenorth> game options is so weird :D
12:33:08 <andythenorth> also can I switch drive side during game?
12:33:12 <andythenorth> make it possible
12:33:30 <peter1138> Yes, if there are no road vehicles.
12:36:39 <andythenorth> send em all to depot
13:29:02 <peter1138> Make road-side NewGRF configurable in NRT ;p
13:29:47 <Eddi|zuHause> anyone ever study the halflife of a "good feature"?
13:57:55 <peter1138> Mmm, reheated pizza
14:05:38 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhpPQ
14:17:40 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhpXv
14:21:25 <Eddi|zuHause> this time not from the bin?
14:31:19 <peter1138> No this one was handed to me yesterday in its box, still hot.
15:04:41 *** octernion has joined #openttd
15:14:45 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7272: Change: [NPF] Add path cache for ships. https://git.io/fhp1b
15:45:03 <Beerbelott> peter1138: Updated my comment
15:46:36 <peter1138> Oh.. Yes, I misread it.
15:46:48 <peter1138> Your baseset is in ~/.openttd/content_download/baseset.
15:47:07 <peter1138> Afaik, changing the config path means it doesn't use ~/.openttd, that's by design.
15:47:37 <peter1138> (Not necessarily a good design!)
15:57:56 *** supermop_work has joined #openttd
15:59:08 *** supermop_work_ has joined #openttd
16:01:06 *** supermop_work__ has joined #openttd
16:03:52 *** supermop_work___ has joined #openttd
16:08:13 *** supermop_work_ has quit IRC
16:11:09 *** supermop_work__ has quit IRC
16:20:14 <Eddi|zuHause> i think there's room for improvement :p
16:20:38 *** Thedarkb-T60 has joined #openttd
16:24:15 <Beerbelott> It might be seen as logical to scan the dir (maybe the arborescence of known dirs) from where the config file is started, but that should not replace, in any way, the 'standard' dirs from locations the game files are said to be looking into
16:25:36 <Beerbelott> I guess the proper thing to do would be to add the dir of the config file (not the arborescence)
16:25:41 <Beerbelott> to the list of scanned files
16:26:14 <Beerbelott> But... well... that'd suppose it's logical we put other files there too
16:26:22 <Beerbelott> I dunno, quite a big design question actually
16:27:01 <Beerbelott> To keep things simple, I wouldn't add dirs to scan
16:27:25 <Beerbelott> I'd keep the standard dirs, merely allowing to temporarily overload the config file by another one
16:27:58 <Beerbelott> That would ensure consistency in what is loaded. The config file is then there to make sure what is configurable can be changed
16:28:47 <Beerbelott> This ensured consisency would save us from this kind of nightmare
16:30:01 <Thedarkb-T60> I've gotten an idea.
16:30:15 <Thedarkb-T60> Why not statically link libicu52.
16:30:47 <Thedarkb-T60> It would make OpenTTD a lot easier to install on Ubuntu and Debian.
16:31:22 <Beerbelott> openttd package manager compiled openttd against libicu57, starting in testing
16:31:27 *** andythenorth has left #openttd
16:31:54 <peter1138> ^ are any of those packages hard to install?
16:31:58 <Beerbelott> and it seems tests were done against libicu60
16:32:24 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhpy2
16:32:24 <Thedarkb-T60> I'm trying that testing release right now.
16:32:30 <Beerbelott> Moreover, you can compile on stretch w/o trouble
16:32:40 <Beerbelott> picking up libicu-dev
16:33:02 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhpyw
16:33:11 <Thedarkb-T60> That's good to know.
16:34:12 <Thedarkb-T60> It installed painlessly!
16:35:28 <Beerbelott> I'd recommend using official packages on the website though :)
16:35:43 <Beerbelott> a simple script would do the trick
16:35:57 <Thedarkb-T60> I used to make packages for TinyCore Linux.
16:36:06 <Thedarkb-T60> I might well package up OpenTTD when I get around to it.
16:36:07 <peter1138> I made an Android build.
16:36:46 <peter1138> But only as a dedicated server build... Need to sort out libsdl :)
16:36:47 <Beerbelott> Thedarkb-T60: The compiling instructions might be a lil bit off-sync w/ reality coming up to requirements list, but the spirit is there
16:37:30 <Beerbelott> Can't think of anything easier
16:38:53 <Beerbelott> To build on stretch, I used that list: zlib1g-dev liblzo2-dev liblzma-dev libfontconfig-dev libicu-dev libsdl-dev libpng-dev li bfreetype6-dev libfontconfig-dev
16:39:16 <Beerbelott> sry: zlib1g-dev liblzo2-dev liblzma-dev libfontconfig-dev libicu-dev libsdl-dev libpng-dev libfreetype6-dev libfontconfig-dev
16:39:29 <Beerbelott> (on top of wget xz-utils build-essential pkg-config debhelper)
16:39:56 <Beerbelott> then u got all the tools to make it roll
16:40:30 <Beerbelott> ofc pulling code from a git repo, scratch wget for git
16:40:39 <peter1138> apt-get build-dep openttd :)
16:41:22 <Beerbelott> Well, that too, but I'm a bit afraid to rely on the dependencies of a package not manage by your team :p
16:41:52 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhpSL
16:42:09 <peter1138> Beerbelott, check the maintainer ;)
16:42:45 *** Wormnest has joined #openttd
16:42:49 <Beerbelott> peter1138: I'll make a tiny PR for issue #7311 where the fix will be replacing that odd line with using the dirs array. At least, a proposal will be on the table to be debated :)
16:43:16 <peter1138> It is maintained by one of us, however obviously it follows the normal Debian package rules, hence why it's kinda old.
16:43:17 <Beerbelott> Matthijs Kooijman <matthijs@stdin.nl>
16:43:48 <Eddi|zuHause> haven't heard that name in ages
16:44:43 <Beerbelott> THere's a Jordi too
16:45:12 <Beerbelott> Well then Thedarkb-T60 I guess those debian packages might be to be trusted after all ;)
16:45:49 <peter1138> 7 grfcodec: flagged for removal in 26.9 days
16:45:49 <peter1138> openttd-opengfx: buggy deps grfcodec, flagged for removal in 26.9 days
16:45:49 <peter1138> openttd: buggy deps grfcodec, flagged for removal in 26.9 days
16:46:17 <peter1138> Ah, website changes
16:49:40 *** supermop_work has joined #openttd
16:50:19 <peter1138> Hmm, it's not in github.,
16:50:38 *** supermop_work___ has quit IRC
16:50:48 <LordAro> someone should fix that
16:52:33 *** supermop_work_ has joined #openttd
16:54:06 <peter1138> Detecting your project’s version control system…
16:54:17 <DorpsGek_II> [OpenTTD/OpenTTD] Berbe opened pull request #7339: Fix #7311: External configuration file does not change arborescence https://git.io/fhpS1
16:55:09 <peter1138> I'm not sure that word is ever used there...
16:55:22 <peter1138> "directories" would make sense.
16:55:37 <Beerbelott> Here goes the 1st force-push
16:55:48 <Beerbelott> Breaking a record on how fast after 1st push :D
16:56:10 *** sla_ro|master has joined #openttd
16:56:40 <DorpsGek_II> [OpenTTD/OpenTTD] Berbe updated pull request #7339: Fix #7311: External configuration file does not change arborescence https://git.io/fhpS1
16:59:50 <Beerbelott> It seems this IRC notification was made w/ the old title
17:00:09 <Beerbelott> Isn't this a bug? cc TrueBrain
17:00:14 <milek7> i don't see that attached bad.patch
17:03:56 <DorpsGek_II> [OpenTTD/OpenTTD] Berbe commented on pull request #7339: Fix #7311: External configuration file does not change directories https://git.io/fhpSj
17:21:20 <Eddi|zuHause> <Beerbelott> Breaking a record on how fast after 1st push :D <-- i don't think you quite made that record :p
17:22:19 <Beerbelott> I feel I pushed waaaay too much on my 1st PR, but that's another record I guess
17:22:31 <Beerbelott> There got to be a medal to earn! :D
17:23:56 *** Progman has joined #openttd
17:25:34 <LordAro> i'm still not a fan of "external" files being picked up during a regression test run
17:25:53 <LordAro> i'm of the opinion that if you specify -c for a config file, it should only use that file (and files relative to it)
17:27:00 <Eddi|zuHause> LordAro: but then, the regression should come with a regression baseset.
17:28:34 <planetmaker> regression baseset... sounds somewhat like overkill
17:29:20 <Eddi|zuHause> Beerbelott: what you should test now: 1) files relative to the config dir are still preferred over personal dir if they exist, 2) savegames etc. are stored relative to the config dir and not the personal dir
17:29:43 <peter1138> LordAro, is your bike back yet?
17:30:15 <planetmaker> though I recall that I could create a NoGFX baseset with OpenGFX source... iirc. The minimal baseset for a server
17:30:16 <Eddi|zuHause> planetmaker: but since the regression is run without UI, it doesn't need to contain actual sprites, just silence the "missing baseset" check
17:30:37 <planetmaker> 1px black pixels for all sprites except terragen
17:31:02 <peter1138> Random pixels where it's not transparent.
17:31:16 <planetmaker> too complicated to make
17:31:19 <peter1138> Call it OpenOzTransGFX
17:31:26 <planetmaker> it's easy to replace all ... :D :D
17:34:34 <Beerbelott> Eddi|zuHause: Well... that I did not know of. The current proposal breaks that for sure
17:35:46 <Beerbelott> I'm more kin to LordAro stance
17:36:04 *** supermop_work_ has quit IRC
17:36:34 <Beerbelott> It's the usual behaviour you encounter for stuff allowing to specify external resources
17:37:03 <Beerbelott> They are specifie explicitly, no implicit derivation from any of them is usually made, sticking to known conf dirs
17:37:22 <Beerbelott> (as that could lead to impredictible situations)
17:38:26 <Beerbelott> But well... that means this little change has deep imapct on how OpenTTD is behaving, break backwards-compatibility
17:38:43 <Beerbelott> Oh dear, what have I put myself into again... :p
17:39:25 <Eddi|zuHause> Beerbelott: it always starts with "this should be an easy one-line patch" :p
17:39:53 *** HerzogDeXtEr has joined #openttd
17:40:56 <Eddi|zuHause> now... in all my travels i found 2 pieces of processed scrap, yet i need 3 to get started with making a shredder to make more pieces of scrap...
17:42:34 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh updated pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhd4b
17:42:37 <Beerbelott> I did not expect it to be that easy... but not unleashing that kind of hell either :D
17:50:15 *** supermop_work has joined #openttd
17:51:04 *** supermop_work_ has joined #openttd
17:52:24 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7339: Fix #7311: External configuration file does not change directories https://git.io/fhpH6
17:58:45 <Beerbelott> OMG when loading an external configuration file, the game creates all sorts of fiels there...
17:58:53 <Beerbelott> That's terrible behaviour IMHO
17:59:35 <Beerbelott> You might be willing to overload the default conf file w/o making a working dir where it's loaded from :\
18:00:03 <glx> it's a nice thing when you host many dedicated servers :)
18:00:06 <Beerbelott> Again, the implicit file locations if not explicitely set otherwise would be a better approach to this...
18:00:32 <Beerbelott> glx: Isolation badly done
18:01:21 <Beerbelott> I see that more as a threat than a feature
18:01:46 <glx> only one exe needed, and then 1 config per server
18:01:51 <Beerbelott> properly isolated instances could never load different configuration files fro mthe same binary ;)
18:02:01 <Beerbelott> exactly what's the worst
18:06:50 <Beerbelott> However, what you could be doing is mounting custom resources (conf file, savegames, scenarios, heightmaps, whatever) on a RO volume, share it between instances and have Server #1/#2 etc be able to be inter-changeable
18:07:31 <Beerbelott> each instance being properly isolated
18:12:04 <Beerbelott> Well I guess the codebase on that has been inherited for quite a while, so it's only fair to assume design choices were made according to (bad) habits of the 90s (80s?) when virtualization/containers were a concept at best, if anything at all
18:12:30 <Beerbelott> The q° is: Does OpenTTD wanna change?
18:15:50 <Eddi|zuHause> <Beerbelott> OMG when loading an external configuration file, the game creates all sorts of fiels there... <-- one of the main uses behind that is you can have both openttd and all data files on an USB stick and carry that around, without spoiling the OS of the computer
18:16:26 <Eddi|zuHause> Beerbelott: you can't tell these people "use containers/virtualization"
18:18:52 <Beerbelott> Then the switch shall not be about loading a configuration file and implying a directory structure around
18:19:10 <Beerbelott> It shall be about changing the working directory and searching for a configuration file inside
18:19:36 <Beerbelott> I totally get your use case, but the code is prepped the wrong way around
18:20:38 <Beerbelott> the scope of the '-c' switch is implicitely widened, which might not be expected
18:21:09 <glx> it's indicated in the readme
18:22:20 <Beerbelott> glx: the '-h' message does not even mention '-c' :D
18:23:14 <Eddi|zuHause> Beerbelott: but even if it's not "expected", you can't simply break existing setups for no reason
18:23:34 <Beerbelott> Well I suppose '-c' definition might be changed to 'change working directory', supporting configuration file loading (and implying working directory from it) as a backwards-compitable fallback?
18:24:03 <glx> " -c config_file = Use 'config_file' instead of 'openttd.cfg'\n"
18:25:17 <Beerbelott> Yes I am... Sry about that
18:25:39 <Beerbelott> What do you think about the aforementioned way of dealing with that?
18:26:00 <Beerbelott> I don't wanna swim against current :)
18:27:08 <glx> hmm yes, using -c could be equivalent to an implicit working directory change
18:28:36 <glx> but definitely setting specified config_dir as personal_dir feels wrong :)
18:29:40 <Beerbelott> So here we go for transforming configuration file locading in config_dir change...
18:29:56 <Beerbelott> (I dunno what I'm doing :D )
18:35:26 <DorpsGek_II> [OpenTTD/OpenTTD] Berbe commented on pull request #7339: Fix #7311: External configuration file does not change directories https://git.io/fhpQQ
18:47:13 *** frosch123 has joined #openttd
18:53:29 <nielsm> that one should be backported too, shouldn't it?
18:59:24 <peter1138> Can be, not particularly important.
19:00:09 <nielsm> could be used as a test for backporting via PRs? :)
19:00:34 <peter1138> #7334 is more important
19:00:39 <Gabda> nielsm: I think the max in ForAllStationsRadius can go up to MapSizeX
19:01:22 <peter1138> Er, #7335 I mean :)
19:01:26 <nielsm> Gabda I'll look at it a bit later, busy moving virtual logs on virtual rails atm :)
19:02:27 <Gabda> and with this I am finished reading, won't bother you with new comments :)
19:16:47 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7327: Cleanup: Remove repeated network error message on connection timeout. https://git.io/fhp7y
19:16:55 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7327: Cleanup: Remove repeated network error message on connection timeout. https://git.io/fhxwc
19:23:02 *** Gustavo6046 has joined #openttd
19:44:12 *** gelignite has joined #openttd
19:48:28 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh updated pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhd4b
19:50:02 *** synchris has joined #openttd
19:54:21 <nielsm> hmm, how are translation changes backported? there might need to be some janitoring in those when strings change in master
19:55:05 <glx> there used to be a script somewhere
20:12:59 <frosch123> yes, there is a script
20:13:58 <frosch123> it's not versioned anywhere though
20:16:34 <Samu> what does backport mean?
20:17:02 <Wolf01> Mmmh, friend's wife wants to gift the lego millennium falcon to friend, he already purchased it with a good discound and buried it under the old lego boxes in the basement so she wouldn't know...
20:21:48 <nielsm> Samu: take changes made against the master version and copy them to the 1.9 release branch, without getting any unrelated changes along
20:21:50 <Eddi|zuHause> there's some he/she mismatch in that sentence
20:22:07 <peter1138> Eddi|zuHause, nope...
20:23:01 *** supermop_work_ has quit IRC
20:24:32 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhp5F
20:26:13 <peter1138> All these cool things we can get in once 1.9 is branched.
20:29:08 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7279: Fix #7062: Remove ship max order distance. https://git.io/fhpdf
20:30:31 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhpdk
20:31:25 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhpdI
20:37:31 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpd3
20:38:31 <Eddi|zuHause> we really need some proper test cases for #7000, but i'm not quite ready for that
20:40:54 <peter1138> Just needs a test to confirm working as expected.
20:41:01 <peter1138> I'm not after full on unit tests.
20:46:24 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpdl
20:46:58 *** supermop_work has joined #openttd
20:48:43 *** supermop_work_ has joined #openttd
21:05:21 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
21:07:14 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpd1
21:36:09 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpFI
21:47:05 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpFR
21:49:50 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpFu
21:51:40 <peter1138> Bah @ micro-usb cables that stop working.
21:54:06 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
21:56:26 <Beerbelott> Where is _config_dir declared?
21:56:49 <Beerbelott> Both references in openttd.cpp & fileio.cpp are tagged 'extern'
21:57:27 <peter1138> There is not _config_dir ?
21:57:51 <peter1138> Do you mean _config_file?
21:57:58 <Beerbelott> I rename all the occurrences to sth else
21:58:13 <Beerbelott> now the linker complains about undefined reference
21:58:14 <peter1138> _config_file is in settings.cpp
21:58:29 <Beerbelott> but the definition for the new name did not change place :\
21:58:33 <peter1138> It's config_dir, not _config_dir.
21:58:45 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7272: Change: [NPF] Add path cache for ships. https://git.io/fhpFX
21:58:59 <peter1138> config_dir isn't a global variable.
21:59:15 <peter1138> It only appears in src/fileio.cpp
21:59:43 <Beerbelott> sry I meant _config_file
21:59:59 <Beerbelott> but then I renamed all the occurrences and now... i's a mess
22:00:23 <TrueBrain> solution: don't! :D
22:00:23 <Beerbelott> What could have done wrong (many things, I know :p )
22:00:26 <TrueBrain> I am so good in these things :D
22:04:14 <peter1138> Right, SDL 2.0. Any objections?
22:04:19 <peter1138> I'm going with no :)
22:04:29 <peter1138> I'm also going for a pee.
22:05:21 <TrueBrain> we prepared everything to accept SDL 2.0
22:05:24 <TrueBrain> so no going back now :D
22:06:44 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpFx
22:07:51 <TrueBrain> we can pick either looking into that after merge, or before ;)
22:07:57 <TrueBrain> up to you peter1138 :)
22:08:12 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN dismissed a review for pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhpFx
22:08:46 <nnyby> looks like a case where all the video was initialized correctly, but somehow the updates just weren't triggered. really not sure how it's possible on the current branch.. dirty rectangles just working as usual. anyways, i haven't been able to reproduce this problem
22:09:49 <peter1138> Also why the heck did my pi kick off my ssh connection :(
22:10:18 <TrueBrain> you cannot even edit those comments
22:10:20 <TrueBrain> you can edit all, but not those
22:10:51 <peter1138> Well, just tried with a totally fresh config, OpenGFX was bootstrapped and it works fine.
22:12:03 <nnyby> anyways, once we push sdl2 to master and everyone is forced to use it, i'm sure any problems like this will be uncovered very quickly :P so that's one debugging strategy
22:12:31 <TrueBrain> it is the one I was hinting at ;)
22:15:40 <peter1138> Hmm, seeing some issues
22:17:26 <peter1138> Ah, that's something else.
22:17:45 <peter1138> Switching from window to fullscreen with alt-enter fails if the mouse is not in the window.
22:18:40 <peter1138> It flicks over and then back again.
22:19:36 <nnyby> hmm, i didnt know about that shortcut. i'm not seeing that behavior right now.. it seems to work for me, using xfce, even when cursor is out of the window
22:19:49 <peter1138> Hmm, and now it's not. What.
22:19:54 <nnyby> what window manager are you using?
22:20:35 <nnyby> i was only able to uncover some problems like this with a "normal" window manager.. i use a tiling wm - xmonad - on my laptop and so i guess i gotta be testing in both
22:21:51 <peter1138> Seems to be intermittent.
22:28:34 <peter1138> My issue, not #7296.
22:35:20 <Beerbelott> I'm fucking STUPID
22:36:06 <Beerbelott> I am amending my last commit every way possible, splitting it as hell, only to realize I am compiling another damn branch.......
22:45:52 <peter1138> Hmm, managed to make it crash.
23:00:28 <peter1138> So close and yet so far ;(
23:26:00 <Beerbelott> Has anyone anaged to load OpenGFX when calling trunk version w/ -c ?
23:38:10 <Beerbelott> These directories precedence are...
23:47:45 <peter1138> Hmm, trying with Mate desktop instead of Gnome.
23:57:28 <nnyby> hmmm. link to crashlog / steps to reproduce? i'll see if i can crash this thing... i have played significant amounts of time on this branch, but very likely there's some weird edge cases to be uncovered
23:57:54 <peter1138> No, it's a bug introduced by my changed. I've got it sorted.
continue to next day ⏵
