IRC logs for #openttd on OFTC at 2018-04-27
⏴ go to previous day
00:00:15 <peter1138> My brain is frazzled already from trying to get jenkins/warnings/gendarme to work.
00:02:54 <TrueBrain> peter1138: and I am not against openttd-useful; it is only a very old solution for an old problem :) Migrating to a modern solution does have my preference
00:03:01 <TrueBrain> (as I have been doing with the linux and OSX targets too)
00:03:06 <peter1138> TrueBrain, I agree.
00:03:22 <peter1138> I have no idea who compiled it, what versions are in there, etc...
00:03:25 <TrueBrain> it just means I am grumpy while figuring out what the heck it is doing :P
00:03:37 <TrueBrain> there is also a -source package, which has the 'what versions' question
00:03:46 <TrueBrain> and Rb compiled it :) No clue which VS :P
00:03:55 <TrueBrain> I did try to keep the loop as close as possible :)
00:05:54 <TrueBrain> disabled GL, disabled other optimizations, MinSpace, etc .. no change in binary size
00:05:58 <TrueBrain> so I guess it is really the libraries
00:06:32 <peter1138> Well, that's installed cmake...
00:06:50 <TrueBrain> most of vcpkg is done via cmake
00:06:56 <TrueBrain> so much easier than a weird project file
00:09:48 <Eddi|zuHause> why does it need to be static?
00:14:18 <TrueBrain> no; compiling against DLLs is even more of a pain
00:14:30 <TrueBrain> both in distribution as in telling people about them .....
00:14:39 <TrueBrain> we used to do it with SDL.dll .. ugh ..
00:18:19 <TrueBrain> okay, I ran out of ideas :(
00:18:45 <TrueBrain> tomorrow I will build x86, see what that does
00:18:57 <TrueBrain> and I want to figure out how to run regressions again (totally forgot)
00:19:09 <peter1138> "All installed libraries are immediately ready to be #include'd and used in your project without additional configuration"
00:19:12 <TrueBrain> means I at least can fix the CI ..
00:19:13 <Eddi|zuHause> i don't think i have something useful to contribute here
00:19:29 <TrueBrain> peter1138: ghehe, I had the same :D
00:19:35 <TrueBrain> did you install the static version?
00:19:52 <TrueBrain> vcpkg install zlib:x64-windows-static
00:19:56 <TrueBrain> installs the static version
00:20:06 <TrueBrain> either way, you can control if vcpkg is loaded in your projects by two settings
00:20:13 <TrueBrain> VcpkgEnabled (enables it, no shit)
00:20:25 <TrueBrain> VcpkgTriplet, this sets what version to use .. x64-windows, x64-windows-static, etc
00:20:29 <TrueBrain> this part is a bit poorly documented
00:20:39 <TrueBrain> no clue how it works with an UI :D
00:21:01 <peter1138> what's the magic to install icu 57?
00:21:20 <TrueBrain> open the ports file, change 61 to 57, (a few times .. 6 or so? I just used sed)
00:21:33 <TrueBrain> install icu, it yells that the sha no longer matches .. copy the sha in the ports file
00:21:58 <TrueBrain> and this will become more and more a problem
00:22:04 <TrueBrain> and ICU 57 is the last one with Layout
00:22:07 <TrueBrain> and it is deprecated
00:22:12 <peter1138> so nothing actually useful like "-version 57" :(
00:22:18 <peter1138> hmm, isn't Layout all we use it for?
00:22:25 <TrueBrain> no, we use ICU for more
00:22:28 <TrueBrain> but the Layout is the issue
00:22:32 <TrueBrain> ICU stopped doing that
00:22:46 <TrueBrain> we really really have to replace it soon
00:22:52 <TrueBrain> the next Debian cannot be compiled
00:23:05 <peter1138> pango is easy to install with this
00:23:12 <peter1138> but shit, that's a ton of dependencies.
00:23:18 <TrueBrain> Ubuntu 18.04 won't work either
00:23:20 <peter1138> oh well, if it's static no matter
00:23:33 <TrueBrain> so the preasure to get pango done is getting higher and higher :)
00:23:53 <peter1138> Has anyone started it?
00:24:19 <TrueBrain> but 18.04 is ICU 60 ..
00:24:27 <TrueBrain> so yeah, we are in a bit of shit when 18.04 hits the market :)
00:24:37 <TrueBrain> as I think many Linux users use that :)
00:25:04 <peter1138> distribute openttd as a docker image lol
00:25:35 <peter1138> hmm, i don't actually know what we use icu for
00:25:42 <peter1138> so converting it to pango won't be easy for me
00:26:41 <TrueBrain> something to do with fonts, I have been told :P
00:26:44 <peter1138> if it's even possible
00:26:47 <TrueBrain> but freetype and fontconfig? Dunno
00:26:56 <TrueBrain> I have been told there is a drop-in replacement
00:27:07 <peter1138> a quick look at pango docs and it seems tightly coupled to gdk, which ... we don't use.
00:27:50 <TrueBrain> ICU suggests using HarfBuzz .. no clue what that is :D
00:28:57 <peter1138> interesting link at the bottom
00:29:05 <TrueBrain> they have icu-le-hb
00:29:09 <TrueBrain> that is a drop-in replacement
00:30:25 <TrueBrain> Google as copyright holder
00:30:41 <peter1138> Damn, no port for that
00:31:17 <TrueBrain> there is for HarfBuzz?
00:31:33 <TrueBrain> that is something I guess
00:31:48 <peter1138> It's a dependency of pango
00:31:55 <TrueBrain> I am guessing you want to vendor icu-le-hb anyway
00:32:29 <TrueBrain> mainly as it should be a temporary thing :)
00:32:45 <TrueBrain> either way, from what I understood, we use ICU for other things too .. so it would still mean we need ICU :P
00:32:54 <TrueBrain> just no clue about any of that :)
00:33:08 <peter1138> Indeed. I just know all randomly appeared "one day"
00:33:22 <peter1138> Wasn't there when I implemented freetype :)
00:34:08 <TrueBrain> but okay ... tomorrow Windows binary for 32bit/64bit .. getting it linked to CI ..
00:34:19 <TrueBrain> than I guess I have to try mingw32 and mingw64
00:34:27 <TrueBrain> and link that to CI
00:34:32 <TrueBrain> than I have to fix a new supybot ..
00:34:38 <TrueBrain> as I want to say: bot, make release
00:34:42 <TrueBrain> and that it makes a releasefor a PR
00:34:47 <TrueBrain> as I want to test some of those .......
00:35:15 <peter1138> i guess compile farm is all separate to CI?
00:35:29 <TrueBrain> the CI dockers can also create release binaries
00:35:38 <TrueBrain> (in fact, they do, just not publish them)
00:36:05 <TrueBrain> peter1138: but if you have VS2017, possibly you can try all libraries except icu from vcpkg, and icu from openttd-useful
00:36:09 <TrueBrain> and see what your release binary size does
00:36:40 <TrueBrain> any additional information of VS2015 / VS2017 with any combination would seriously help
00:37:07 <TrueBrain> as currently I have no clue if this is optimizations against speed, or just something being dipshit crazy
00:37:10 <TrueBrain> 64bit doesnt need dmusic
00:37:19 <TrueBrain> so if you need dmusic, you are doing the wrong target :D
00:37:45 <TrueBrain> but I am off to bed :) Tnx both peter1138 and glx :)
00:37:45 <peter1138> vcpkg does 32 bit default and i couldn't be arsed to change it :p
00:37:59 <TrueBrain> vcpkg install libpng:x64-windows-static :)
00:39:06 <peter1138> wonder if you can put vcpkg install in the project :p
01:07:59 *** sim-al2 is now known as Guest1013
01:08:00 *** sim-al2 has joined #openttd
02:44:02 *** sim-al2 is now known as Guest1021
02:44:03 *** sim-al2 has joined #openttd
03:38:10 *** Sheogorath has joined #openttd
04:51:10 *** sim-al2 is now known as Guest1027
04:51:11 *** sim-al2 has joined #openttd
04:54:28 *** muffindrake1 has joined #openttd
05:01:59 *** ToffeeYogurtPots has quit IRC
05:02:44 *** ToffeeYogurtPots has joined #openttd
06:08:53 *** Wacko1976 has joined #openttd
06:40:26 *** ToBeFree has joined #openttd
07:21:16 *** Wacko1976 has joined #openttd
08:06:26 *** andythenorth has joined #openttd
08:19:24 <andythenorth> buy menu sprites eh
08:19:43 <andythenorth> maybe I can use sprite layers o_O
09:13:23 *** andythenorth has joined #openttd
09:23:50 *** andythenorth has joined #openttd
09:59:58 *** andythenorth has joined #openttd
10:03:54 <TrueBrain> but now the question is ..... for how long!
10:07:08 <andythenorth> until the next job I have to do
10:09:19 <peter1138> Damn, at some point I think I accidentally disabled password ssh logins on my VM at home.
10:09:52 <peter1138> That would explain why when I was trying to turn off password logins on a server the other day, the changes got lost. "lol"
10:15:38 <nielsm> re. commit messages, can someone elaborate on "change" vs "codechange" vs "update" vs "prepare" ?
10:16:19 <TrueBrain> haha, cool, I can command a Windows Docker from linux :) Okay, that makes shit easier :)
10:16:34 <TrueBrain> nielsm: easiest is to look through the history to get an idea for that
10:16:46 <TrueBrain> but Codechange are changes that only affect code; no function change what-so-ever
10:16:49 <TrueBrain> (refactoring, moving, etc)
10:17:00 <TrueBrain> Update is rarely for code, mostly for shit like languages, readmes, etc
10:17:06 <TrueBrain> Prepare I have never seen use
10:17:17 <TrueBrain> Change is for when there is a function change to your code change .. give or take
10:17:19 <TrueBrain> not that black/white
10:18:22 <nielsm> here I have a rev that basically just adds some parameters to a function prototype (in all the music drivers) to prepare for an actual feature add, which is only possible in the new win32 driver
10:18:35 <nielsm> does that even warrant a separate commit?
10:19:10 <TrueBrain> commits should be a logic separation between changes so the reviewer can easily and trivially understand the change
10:19:25 <TrueBrain> up to you to fill in that line
10:23:41 *** Eddi|zuHause has joined #openttd
10:47:06 <peter1138> Problem with that particular change is the added parameters are only relevant to the one driver.
10:47:39 <peter1138> I do wonder if we should somehow replace all those music drivers with a built-in synth.
10:47:52 <peter1138> It would be a consistent experience then.
10:48:54 <peter1138> Mind you, someone would then complain that they can't use their expensive hardware synth :p
10:50:39 <nielsm> I got a roland sc-55, ie. the synth the music was written for, and hell do I want to use it
10:51:41 <nielsm> (in fact a large part of the reason I even started doing these changes was that I had trouble getting either win32 or dmusic driver to use it)
11:06:23 <peter1138> On Linux MIDI is especially awkward.
11:06:36 <peter1138> Windows at least has the default soft-synth.
11:08:22 *** tycoondemon has joined #openttd
11:15:00 *** FLHerne has joined #openttd
11:26:16 <andythenorth> can I be arsed to script buy menu sprites?
11:26:20 <andythenorth> or shall I just draw them :P
11:26:49 <peter1138> Does it need anything special?
11:27:05 <andythenorth> well I generate the spritesheets
11:27:07 <peter1138> Doesn't the game already just use the correct sprite? :p
11:27:18 <andythenorth> oh this is just for articulated vehicles
11:27:20 <peter1138> You're going to say "HAHAHA SO NAIVE" aren't you
11:27:33 <andythenorth> articulated vehicles don't exist in buy menu is it
11:27:44 <peter1138> Script if you can, a repeatable automated build is nice.
11:27:56 <peter1138> Jenkinsfile it etc :p
11:27:59 <andythenorth> I tried it with sprite layers
11:28:06 <andythenorth> but I don't know if they work in purchase menu
11:28:09 <andythenorth> I could read src I guess
11:28:46 <andythenorth> I automate a bit, see what happens
11:28:53 <Sacro> TrueBrain: how'd you do windows docker in Linux?
11:30:27 <TrueBrain> with a Windows Docker host ofc :)
11:32:24 *** Wormnest has joined #openttd
11:36:50 *** tycoondemon has joined #openttd
11:50:24 *** Smedles has joined #openttd
11:55:57 <TrueBrain> just happy you didnt stash us
12:01:58 *** Wacko1976 has joined #openttd
12:13:26 <TrueBrain> I still wonder about the DrawWindows patch .. we now have 2 functions that do something with the real-time ..
12:13:38 <TrueBrain> guess I need to install MSVC after all :P
12:15:28 <peter1138> It makes my graphical glitching worse :(
12:16:00 <TrueBrain> yeah, we dont have stuff like vsync
12:16:08 <TrueBrain> and 66 Hz is a very odd frequency
12:16:23 <TrueBrain> so it is very likely it is an issue for a lot of people honestly
12:16:51 <TrueBrain> current code is also broken, just less likely to show :)
12:17:03 <peter1138> Yes, GDI can't do vsync.
12:17:33 <peter1138> SDL may be able to, certainly with OpenGL.
12:18:16 <TrueBrain> normally it is not that relevant .. double-buffer and let it up to the driver itself
12:18:31 <TrueBrain> but it is also weird that Window stuff is influenced by FF
12:18:49 <peter1138> Yes, would be nice to replace the stuff that uses game ticks incorrectly.
12:18:54 <peter1138> There's quite a lot that does it.
12:19:06 <peter1138> FF should just make the game faster, not all the windows :p
12:19:20 <peter1138> We don't enable any double-buffering.
12:19:59 <TrueBrain> so there is an isuse :D
12:20:16 <TrueBrain> dunno .. feels this needs another approach, but I cannot really put my finger on it
12:20:18 <peter1138> Well, our use of GDI should be but it seems not to be.
12:20:27 <TrueBrain> not having touched OpenTTD in years has something to do with that :D
12:20:33 <peter1138> My mouse cursor disappears if I move it too fast.
12:20:54 <TrueBrain> this patch just makes it much more obvious there is more broken
12:20:59 <peter1138> It's temporal of course, everything in the backing buffer is fine.
12:21:06 <TrueBrain> but it is hard to say what to fix, without looking into it myself
12:21:30 <TrueBrain> and I hate asking others: please look into this, without giving any clues :D
12:21:32 <peter1138> I wondered how about switching to DirectX/OpenGL, rendering to a surface and then swapping that.
12:22:02 <TrueBrain> you have been talking about OpenGL ever since I met you :D
12:22:32 <TrueBrain> having proper OpenGL/DirectX would solve a few issues left and right
12:22:41 <TrueBrain> also shit like fullscreen, borderless, resolutions, etc
12:22:51 <TrueBrain> I cannot believe that if I go fullscreen, it switches to 640x480
12:22:58 <TrueBrain> that is such a 1990 thing to do
12:23:13 <TrueBrain> all modern games switch to your native resoltion
12:23:19 <TrueBrain> meh .. first lunch time
12:23:25 <peter1138> Yeah but my initial OpenGL stuff was, well, learning about it.
12:23:34 <peter1138> And based on simple tutorials which did everything wrong as well.
12:27:18 <andythenorth> maybe we could get modern
12:28:44 <peter1138> Although OpenRCT2's OpenGL renderer also appears to use the renderer to draw individual sprites as well.
12:29:19 *** justiceBEAVER has joined #openttd
12:29:21 <peter1138> michi has a nearly working OpenGL patch
12:29:28 <peter1138> It did crash for me :p
12:33:32 <andythenorth> automate buy menu sprites more?
12:33:47 <andythenorth> or exercise and go shopping? :P
12:38:28 <andythenorth> hmm, liveries could be controlled by invisible wagons
12:38:40 <andythenorth> add them to the rear of the train
12:39:10 <andythenorth> this idea came from a stupid bug I just found in Horse :)
12:48:54 <nielsm> just make sure the invisible wagon is sufficiently tangible inside the depot, so you can actually remove it again
12:50:04 <andythenorth> it could flash animated colours :P
12:56:18 <peter1138> Why's my train taking ages to load?
12:59:25 <nielsm> you need to upgrade it with an SSD
13:04:28 *** ChanServ sets mode: +v tokai
13:18:45 <TrueBrain> the moment that the Antimalware starts to scan your docker over and over again :D
13:20:08 <TrueBrain> 15 GB of Docker image to run MSVC .... ugh
13:20:24 <TrueBrain> 10 GB is the base image
13:21:25 <peter1138> btw, I'd say changes to use _realtime_ticks for ui stuff should be a separate commit
13:21:54 <TrueBrain> I think the solution is to take out realtime_ticks stuff, and put them all in a single function or something
13:22:04 <TrueBrain> so it is clear what is based on the realtime
13:22:08 <TrueBrain> and what is based on the gametime
13:22:15 <TrueBrain> than think hard about what is currently where
13:22:19 <peter1138> Eh, the code is spread around in window functions.
13:22:34 <peter1138> Using window OnTick() events.
13:22:51 <TrueBrain> as if we can split out what is realtime and what is not, it becomes more clear what is going on
13:22:55 <peter1138> Simplest might be OnRealtimeTick()
13:23:26 <nielsm> yeah that could be useful for some things
13:24:15 <nielsm> when I was still trying to fix things in the dmusic driver I ran into not wanting to update the volume control value too often and kind of wanted to throttle it based on real time but didn't find a good method
13:24:22 <nielsm> (this is of course moot now)
13:26:57 <peter1138> And then midi volume on linux does nothing ;(
13:28:05 <nielsm> yeah what I had to do to control midi volume in dmusic was manually keep track of individual channel volumes and inject channel volume change messages to scale everything
13:28:21 <nielsm> same thing the new win32 driver does actually
13:28:37 <nielsm> (and same thing the driver in dos ttd did)
13:30:03 <peter1138> So should we start parsing and processing the MIDI ourselves?
13:30:16 <peter1138> That might end up making things simpler. extmidi is a bad concept.
13:30:18 <nielsm> of course if you inject too many volume change messages in a midi stream getting sent to a hardware synth over a classic ~38 kbps midi connection you risk running out of bandwidth
13:30:24 <nielsm> so I had to throttle it
13:30:50 <peter1138> nielsm, that's easy, only set the volume on mouse up, or every 0.25 seconds or so
13:30:59 <peter1138> er... so throttle it, exactly as you did :p
13:31:14 <peter1138> mouse up only is annoying, some games do that for sliders :(
13:31:25 <peter1138> You know know what you're going get!
13:34:11 <TrueBrain> right, guess it is time to install MSVC build tools ..
13:39:20 <LordAro> -rw-r----- 1 root adm 5270022379820 Apr 27 10:33 error.log.1
13:39:49 <TrueBrain> okay, the win32 video driver is a bit weird .. there is a for (;;) {} with no break, yet there is code below it
13:40:03 <TrueBrain> pretty sure that code is never going to be called ...
13:42:26 <peter1138> I... don't see that.
13:43:25 <TrueBrain> _draw_threaded if statement
13:44:36 <TrueBrain> line 1329 in win32_v.cpp
13:44:44 <peter1138> Ah right, I see what you mean now.
13:44:53 <TrueBrain> that is some sweet dead-code right there
13:44:59 <peter1138> There's a return which should be a break.
13:46:53 <TrueBrain> hmm .. CLion really wants a CMake before it wants to work
13:47:02 <TrueBrain> owh well, enough for today, going to enjoy the weather for a bit :)
13:47:48 * andythenorth tries to use tab key to speed up compile
13:48:04 <peter1138> I assume, like Minecraft, you all have the same weather...
13:48:35 <peter1138> Simutrans has a day/night cycle. There's a minute of game time when you can see fuck all.
13:48:40 <peter1138> Useful feature that.
13:49:16 <andythenorth> we could map the palette to 1?
13:50:01 <andythenorth> the upside of automation is that it saves loads of time
13:50:11 <andythenorth> the downside is that it automates adding bugs :P
13:50:22 <peter1138> At least the bugs are consistent.
14:05:10 *** synchris has joined #openttd
14:28:44 <nielsm> hmm I thought all my commit messages should be good now, what's it complaining about?
14:38:35 *** FLHerne has joined #openttd
14:39:20 <peter1138> Would be nice if the checker would display the bad message, and commit hash.
14:41:44 <nielsm> hacking on it locally to do just that atm
14:44:58 <nielsm> seems I accidentally a space before the :
14:54:23 <peter1138> Oh yes, I see it now :p
14:55:38 <nielsm> (I'm not good at shell scripting)
15:01:02 *** HerzogDeXtEr has joined #openttd
15:05:52 <peter1138> Erm, I dunno what you changed :p
15:06:16 <nielsm> echoes commit ids as they are checked
15:06:36 <nielsm> instead of failing on the first error, sets a flag and continues checking the rest, then returns error flag at the end
15:13:14 * nielsm performs a little dance
15:13:19 <nielsm> "The commit looks good"
15:15:36 <nielsm> errors in the non-windows midi drivers
15:16:25 <peter1138> Premature little dance :p
15:28:38 *** chomwitt has joined #openttd
15:32:29 <ZehMatt_> Its somewhat tempting to clean up all the video driver code
15:32:42 <ZehMatt_> most of it is somehow duplicate code
15:34:04 <ZehMatt_> well still having the one PR ongoing
15:38:47 <Eddi|zuHause> ZehMatt_: make an abstract video driver factory? :p
15:56:35 <TrueBrain> [14:39] <peter1138> Would be nice if the checker would display the bad message, and commit hash. <- feel free to make a PR for the git-hook repo :)
15:56:53 <TrueBrain> owh, someone already did :P
15:56:57 <TrueBrain> should read up before commenting :D
15:57:32 <TrueBrain> and no, I missed the rain twice; lovely ride in the sun for me :)
15:58:29 *** ToffeeYogurtPots_ has joined #openttd
15:59:10 *** ToffeeYogurtPots has quit IRC
16:00:12 <nielsm> "All checks have passed" wee
16:08:15 <Eddi|zuHause> what about the czechs then?
16:08:47 <Eddi|zuHause> (anyone got the feeling my jokes got worse over the last year or two?)
16:11:52 <TrueBrain> yippie, basics of CMake works :)
16:14:33 <TrueBrain> now to figure out how to get stuff like strgen running :)
16:17:23 <TrueBrain> now it auto-detects vcpkg stuff :D No more fiddling! :D
16:25:08 *** andythenorth has joined #openttd
16:30:48 <TrueBrain> bit weird what Pikka said: "if andythenorth is joining, I am going to leave; really sick of all those images about trains"; not sure what that was about
16:30:58 <andythenorth> yeah I read the logs
16:31:14 <andythenorth> I know, that pikka
16:31:35 <andythenorth> shall I close some issues?
16:31:45 <andythenorth> anything to avoid drawing \ / views for trains :P
16:32:01 <TrueBrain> really happy the new OSX builds seem to fix a few OSX issues :)
16:32:12 <andythenorth> I haven't tried it properly
16:32:29 <andythenorth> I have a cheap tester, I'll get him to play tomorrow
16:32:41 <andythenorth> he doesn't eat or drink when he's playing games, so it's a saving overall
16:32:48 <TrueBrain> the mouse weirdness seems gone
16:33:03 <andythenorth> I wonder if it's stopped locking the cursor to left screen edge too
16:33:09 <andythenorth> that's kind of annoying
16:33:14 <andythenorth> quite unplayable
16:33:18 <TrueBrain> lemme know if it does :)
16:33:24 <TrueBrain> if it doesn't, tell LordAro
16:33:37 <TrueBrain> I dont care about stuff that is not working :P
16:33:55 <TrueBrain> now how am I going to do strgen in CMake I wonder ..
16:34:18 <andythenorth> your problems are same as mine, just several levels lower :P
16:35:04 <LordAro> TrueBrain: custom commands, i imagine
16:35:21 <TrueBrain> yeah .. but that doesnt result in code now does it :D
16:36:05 <TrueBrain> for example, how do I get it to compile strgen before it runs the custom command .. hmm
16:36:21 <Rubidium> TrueBrain: somebody definitely has a patch for it
16:36:38 <TrueBrain> Rubidium: sadly, that always is the case in this community .. bit annoying tbfh :P
16:36:49 <TrueBrain> so LordAro, why isnt this in OpenTTD yet?
16:37:19 <LordAro> because i'm still not sure i like cmake :p
16:37:20 <Rubidium> because he closed the CMake build ticket for OpenTTD? ;)
16:37:55 <TrueBrain> LordAro: your personally preference is irrelevant to me :P
16:38:10 <TrueBrain> this custom pile of shit really needs replacing :)
16:38:11 <LordAro> configure scripts have the advantage of portability
16:38:15 <TrueBrain> (I can call it that; I wrote it)
16:38:24 <LordAro> even autotools generated configure scripts have that
16:38:27 <TrueBrain> all OSes we support, have CMake support
16:38:44 <LordAro> with cmake, you need cmake to build
16:39:01 <TrueBrain> so what is next? Are we going to write our own compiler, because it is more portable?
16:39:13 <TrueBrain> current "configure" system is 14 years old ....
16:39:19 <TrueBrain> it needs replacement
16:39:44 <Rubidium> *explitive* is time really going that fast?
16:39:46 <TrueBrain> and cmake won that race
16:40:37 <andythenorth> is how I got to be 40 years old
16:40:44 <andythenorth> I was less when I started ottding
16:40:53 <LordAro> Rubidium: hey, i've been here 9 years now :p
16:41:17 <LordAro> TrueBrain: well, go ahead then
16:41:32 <LordAro> autotools avoids an extra (hard) dependency, is all
16:41:39 <TrueBrain> autotools really is a big no-no
16:41:47 <TrueBrain> they really lost the race, for so many reasons
16:41:55 <TrueBrain> but having Linux and Windows closer together, is really a plus
16:42:04 <TrueBrain> means we can stop generating MSVC files for 2017+
16:42:20 <TrueBrain> and they don't avoid a dependency at all btw
16:42:29 <TrueBrain> I have more systems without autotools than with :)
16:42:30 <LordAro> i don't have any particular opposition to cmake, just a general unease :p
16:42:48 <TrueBrain> I am also trying to get more and more systems without gcc
16:42:56 <TrueBrain> (but with clang, for example)
16:43:38 <TrueBrain> Rubidium: you really want to play this game? :)
16:44:24 <TrueBrain> hmm, the import of course doesnt let me step into that branch
16:44:37 <TrueBrain> and I am not annoyed your name is on that commit :P
16:44:48 <TrueBrain> not even shared credits!
16:47:29 <andythenorth> train 50 is not my finest work
16:47:40 <andythenorth> should look more like train 52
16:50:12 <andythenorth> $someone should fix it :P
16:50:18 <TrueBrain> Rubidium: okay, 2006-07-30 was the first commit to that branch; guess we call that the starting date :P
16:53:21 <TrueBrain> okay, cmake now first compiles strgen .. right .. it fails horrible .. endian shit .. at least in concept that works :)
16:53:34 <Rubidium> still amazing that it managed to function well for so long
16:53:56 <TrueBrain> same .. as it was written with little bash knowledge, and a lot crash&burn moments
16:54:08 <TrueBrain> but once it got stable .. it seems nobody touched it enough for it to break :P
16:54:14 <TrueBrain> like this big tower build on sticks
16:55:05 <peter1138> What was that other build system... hmm...
16:55:37 <TrueBrain> nowedays you have autoconf and cmake
16:55:40 <peter1138> Oh I remember. waf.
16:55:43 <TrueBrain> where cmake can address a few more targets
16:55:48 <TrueBrain> like Makefile, ninja, and a few more
16:56:08 <TrueBrain> that is the other nice thing if cmake works .. means you can switch to ninja if you like :)
16:56:11 <peter1138> waf had the concept that you distributed waf with your project.
16:56:19 <peter1138> But it was bytecode.
16:56:39 <TrueBrain> sounds like the sound a dog makes tbfh
16:57:14 <peter1138> I gave up with it and used... guess what?
16:57:56 <TrueBrain> the more I use cmake, the more I like it; especially as more people start to use it
16:58:08 <TrueBrain> pkg-config was a good idea, just not enough people used it for a long time
16:58:17 <peter1138> Hmm, this was 8/9 years ago ;)
16:58:37 <TrueBrain> find_package in cmake replaces pkg-config .. but enough packages support it for it to be useful
16:58:48 <TrueBrain> I remember in 2006 that pkg-config was only supported by 1 or 2 libraries
16:58:52 <TrueBrain> and even than they tend to break
16:59:03 <TrueBrain> like one of them added -pthread, no matter what youc ompiler was
17:00:40 *** rocky1138 has joined #openttd
17:01:50 <TrueBrain> but now, even on Windows the find_package(png) worked :D
17:03:16 *** Thedarkb-X40 has joined #openttd
17:04:56 * Rubidium wonders to what extend endian_check is actually really needed; seems like there's a preprocessor flag for it
17:07:26 <TrueBrain> something to try :)
17:07:34 <TrueBrain> removes a lot of silly code :)
17:07:48 <TrueBrain> are there systems that dont have htons etc defined I wonder :P
17:08:58 * andythenorth fixes train 50 with 'return []'
17:09:05 <andythenorth> hmm but now nml complains
17:09:12 <andythenorth> one step forward, one step back
17:10:51 <Rubidium> TrueBrain: only ones not conforming to POSIX.1-2001
17:11:15 <TrueBrain> given we are switching to stdc++11 or higher, can I assume that wont be a problem? :D
17:11:20 <TrueBrain> so much code that can be removed :P
17:12:22 * andythenorth has to make more MOAR CODE :(
17:12:30 <andythenorth> too much conditional shit
17:12:37 <andythenorth> time to split the template
17:13:16 <TrueBrain> don't hype! PROGRAM!
17:13:26 <LordAro> def detect_missing_files_by_reading_error_messages():
17:14:33 <Rubidium> TrueBrain: I can't think of any reasonably recent OS with UI that would have support for C++ compiled binaries but no support htons
17:14:57 <TrueBrain> so we can drop the whole endian-check-shit
17:15:54 <Sacro> But I run TTD on Windows 98SE still :(
17:16:14 <nielsm> what compiler supports C++11 and also win9x targets?
17:16:22 <Rubidium> Sacro: that still falls under reasonably recent ;)
17:16:42 <nielsm> and yes htons is in winsock 1 which arrived in win95
17:16:45 <Rubidium> TrueBrain: gfx_type.hpp's going to be the largest obstacle
17:17:03 <TrueBrain> who has a patch for that?
17:17:36 <TrueBrain> Rubidium: we can fix that with __BYTE_ORDER__ for now I guess
17:18:06 <Sacro> I've not played OpenTTD in ages to be fair, busy working
17:18:13 <Sacro> Is PBIMTTD still a thing?
17:18:32 <TrueBrain> does he now imply we are not working? Hmmmmmmmm
17:18:56 <Sacro> You know I'm here to troll
17:19:00 <Sacro> Same as the last 2 decades
17:19:08 <TrueBrain> at least you now admit it :D
17:19:12 <FLHerne> I thought meson was supposed to be the new shiny vs CMake/autotools :P
17:19:18 <nielsm> also keep in mind that any recent compiler (worth using) should recognize common patterns for assembling/disassembling values from integers into specific byte orders
17:19:44 <nielsm> and translate it into a known-fast machine code equivalent for the platform
17:19:48 <Rubidium> TrueBrain: yep, though test for #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ (i.e. invert the lines), so if __BYTE_ORDER__ and __ORDER_LITTLE_ENDIAN__ are not defined, then it still assumes it being little endian (which would be the right choice most of the time anyway)
17:20:11 <TrueBrain> I expect a PR any minute now? :D
17:21:03 * Sacro considers writing a daylength patch
17:21:30 <FLHerne> Definitely aren't enough of those
17:22:10 <Sacro> FLHerne: mine was the first
17:22:59 <Rubidium> TrueBrain: you can have the expectation, but at the moment I don't have the environment to make one
17:23:20 *** supermop_work has joined #openttd
17:23:23 <TrueBrain> guess you have some work to do when you get home :P :P :P
17:25:23 <TrueBrain> goes for you too LordAro :P
17:25:54 <TrueBrain> nah, tomorrow I will fiddle with getting strgen to work via cmake without endian stuff, make a branch for it, see how that goes
17:25:56 <LordAro> got to deal with some Ada for another 2 hours first though
17:26:15 <TrueBrain> guess we need to make a simple cmake parser for sources.list :)
17:27:19 <LordAro> the hardest bit will be handling the #if
17:27:32 <LordAro> could run cpp on it :p
17:27:47 <TrueBrain> even if wehave to modify sources.list a bit, as long as the others understand it again
17:27:54 <LordAro> wouldn't be very cross platform though
17:28:17 <LordAro> or... would it? msvc must have a preprocessor somewhere
17:28:52 <TrueBrain> but in generate(.vbs) has a parser for sources.list .. so I am sure we can figure something out there :)
17:30:16 <LordAro> is the goal not to get rid of the generate scripts entirely?
17:30:28 <TrueBrain> but never go for the goal, always make a step towards it
17:31:22 <Sacro> openttd/bionic 1.7.1-1build1 amd64
17:31:23 <nielsm> okay theoretically I think you could write a module for msbuild that allows you to read an external source to generate items for the compile
17:31:26 <Sacro> Will I get away with it in the office
17:31:41 <nielsm> which would mean it could read source.list directly and produce something
17:31:56 <nielsm> except that wouldn't cover the .vcxproj.filters stuff which is handled differently iirc
17:31:59 <TrueBrain> nielsm: only we are talking about cmake; not msbuild :)
17:32:15 <nielsm> <LordAro> or... would it? msvc must have a preprocessor somewhere
17:32:55 <nielsm> but I think either vs2017 or maybe the next version might have direct support for cmake?
17:32:56 <TrueBrain> if you go that road, build a custom application to do the conversion for you
17:33:16 <TrueBrain> happy you caught on; VS2017 supports CMake, the whole reason I started this ;)
17:33:41 <TrueBrain> the whole of vcpkg uses it
17:33:55 <nielsm> but what would need a preprocessor then?
17:33:57 <TrueBrain> blathijs: ICU Layout needs replacing
17:34:06 <TrueBrain> blathijs: there are drop-in replacements for it
17:34:11 <TrueBrain> blathijs: but OpenTTD currently still needs it :)
17:34:21 <TrueBrain> blathijs: LordAro keeps telling me he is working on it .. he is just slacking :P
17:34:41 <Sacro> TrueBrain: are you trying to bring everyone else out the woodwork?
17:34:42 <TrueBrain> (and yes, 18.04 cannot compile OpenTTD :D)
17:34:54 <Sacro> Can it not? I've not tried
17:34:58 <LordAro> last i got to was deciding pango wasn't worth the effort, and going for harfbuzz instead
17:35:01 <TrueBrain> Sacro: no, blathijs asked a question in another channel; thought it was better to answer him here :)
17:35:16 <TrueBrain> LordAro: that should be an easy fix. Vendor icu-le-hb, and go go go :)
17:35:26 <TrueBrain> Sacro: 18.04 comes with ICU 60, which doesnt have the Layout engine
17:35:43 <Sacro> Ahh, I didn't know that, I don't do any icu dev anyway
17:36:04 <LordAro> TrueBrain: yeah, but that's the hacky fix
17:36:07 <LordAro> i want to do it properly
17:36:19 <TrueBrain> LordAro: but we dont have 3 years; 18.04 is due this week? :)
17:36:22 <blathijs> Anyone know how much we really need the ICU ParagraphLayout? Seems we use it for wordwrapping with right-to-left support at quick glance
17:36:25 <blathijs> Apparently support for it is a bit dodgy in ICU: the layout framework it uses was deprecated for buggyness, removed from icu and is now developed as a separate library apparently, which significantly complicates compiling the Debian packages (since the external library depends on ICU, whose ParagraphLayout part again depends on the external lib). Apparently we're the only ones in Debian that
17:36:26 <LordAro> TrueBrain: yesterday, actually :p
17:36:27 <blathijs> actually uses the ParagraphLayout, so things might get easier if we wouldn't :-p
17:36:56 <TrueBrain> LordAro: so if wecan fix this by vendoring some code and adding hb as dep, and be done in a weekend; later we can do it properly :)
17:37:09 <LordAro> blathijs: iirc, it's only "required" for RTL texts
17:37:12 <blathijs> Note that that post suggests asking the OpenTTD folks, but I've not been asked so far (I just stumbled upon this post by accident)
17:38:35 *** Eddi|zuHause has joined #openttd
17:38:36 <blathijs> LordAro: What would be alternatives? SKipping ParagraphLayout and using icu-le-hb directly?
17:38:46 <blathijs> Or completely implmenting it ourselves?
17:38:58 <Rubidium> the latter is bonkers
17:39:10 <LordAro> icu-le-hb is the shortterm make-it-work fix
17:39:20 <LordAro> longer term, porting to "native" harfbuzz
17:39:25 <Rubidium> as it's not only for the order of the characters, but also for the joining of glyphs
17:40:23 <blathijs> Oh, I thought icu-le-hb and harfbuzz were the same. I haven't really read up yet :-)
17:40:44 <TrueBrain> icu-le-hb is een drop-in replacement for the Layout stuff in ICU to work with HB
17:40:56 <TrueBrain> it basically implements the ICU interface on top of HB
17:41:10 <TrueBrain> its very clever tbh :)
17:41:58 <LordAro> incidentally, there's no debian package for it
17:42:05 <LordAro> which would complicate things
17:42:08 <TrueBrain> like I said, I would vendor ice-le-hb
17:42:15 <TrueBrain> it is very small, very simple
17:42:49 <TrueBrain> possibly we need to extend our license that we have a vendored package with another license
17:43:22 <andythenorth> karma will repay me for this next comment
17:43:30 <TrueBrain> we already have that solved because of squirrel
17:43:33 <TrueBrain> so license cannot be a problem :)
17:43:34 <andythenorth> my proxy, temp, placeholder sprite to find a size
17:43:43 <andythenorth> is better than some of the release sets I'm looking at :P
17:43:49 <andythenorth> I will suffer for my arrogance ;P
17:44:22 <TrueBrain> LordAro: and ice-le-hb license allows vendoring
17:44:41 <TrueBrain> so from what I can see, it is just a matter of dropping ice-le-hb in 3rdparty folder
17:44:57 <TrueBrain> add HB as dependency, upgrade ICU, and lalalaaa
17:45:05 <TrueBrain> I keep writing ice-le-hb .. it is icu-le-hb
17:45:10 <TrueBrain> dunno why my fingers do this ..
17:45:34 <Eddi|zuHause> you're secretly wanting some ice cream
17:48:47 <Eddi|zuHause> from Lippe to Bremen?
17:50:18 <Eddi|zuHause> actually, LE seems to stand for "Lemgo" (in Lippe)
17:50:57 <Eddi|zuHause> i somehow don't think they have an ICE connection :p
17:52:08 <peter1138> Well, it's Friday and I go home in 8 minutes.
17:52:28 <peter1138> Maybe I will check out this icu-le-hb library.
17:52:40 <peter1138> pango itself doesn't seem to be the right thing to use.
17:52:42 <TrueBrain> solves a lot of dependency issues :)
17:53:03 <peter1138> The thing with "joining of glyphs" is... do we need that?
17:53:24 <Eddi|zuHause> some languages are full of these things
17:54:00 <peter1138> Fortunately harfbuzz has a vcpkg port :)
18:12:58 *** Tharbakim has joined #openttd
18:17:18 <nielsm> what is the proper way to use media/baseset/translations.awk, is there a make target for it or something?
18:20:13 <blathijs> TrueBrain: LordAro: Sorry, was away for a bit
18:21:12 <blathijs> TrueBrain: Ah, I see about icu-le-hb. The bug I linked above suggests that there will be a Debian package for icu-le-hb, which is what they'll be using to get ParagraphLayout working on the upcoming ICU version
18:25:17 <blathijs> Apparently the layout stuff was removed from ICU 58, but Debian is still at 57, so this was not a problem before
18:29:18 <blathijs> TrueBrain: In any case, I think that icu-le-hb won't help directly (either in Debian or vendored), since we use ParagraphLayout, which is *not* in in icu-le-hb, but still in icu for some reason
18:32:29 *** Progman has joined #openttd
18:34:00 <andythenorth> should I play OpenTTD?
18:34:13 <peter1138> blathijs, ParagraphLayout is still in ICU.
18:34:15 <Eddi|zuHause> actually play the game? what blasphemy
18:34:49 <peter1138> Hmm, this Qwertee t-shirt is... very thin :S
18:35:42 <nielsm> I split my GUI changes for music into a separate PR, which I think should be easier to review ;)
18:36:18 <nielsm> and the rest is also much more cleaned now
18:36:39 <nielsm> I can't think of any more immediate improvements
18:37:16 <nielsm> apart from having the (untested) parts actually tested so those notes can be removed
18:37:29 <peter1138> Actually tested or just compiled?
18:37:50 <peter1138> Hmm, cold. Might put the heating on.
18:38:19 <nielsm> the big thing is whether you can get the dos version music playing on non-windows systems
18:38:32 <nielsm> the CI compiles do pass for it
18:38:57 <andythenorth> peter1138: get your big coat
18:40:25 <peter1138> Problem is it's too big.
18:40:54 <peter1138> Not quite that big :p
18:41:20 <peter1138> Losing weight is expensive ;(
18:43:23 *** sla_ro|master has joined #openttd
18:44:10 <andythenorth> let's try NoCarGoal again
18:44:22 <andythenorth> can't remember if I hacked it in place to fix the cargos :P
18:44:33 *** Wacko1976 has joined #openttd
18:48:39 <blathijs> peter1138: Yes, that's what I said, ParagraphLayout is in ICU, not icu-le-hb, which creates a weird circular dependency. Moving it to icu-le-hb would seem sensible, since that seems to be the only lib to offer the layout enging dependencies needed by ParagraphLayout.
18:49:52 <peter1138> I don't think there's any circular dependency, you can just link both.
18:55:32 <blathijs> peter1138: icu-le-hb (and/or Harfbuzz) depends on ICU, so to build ParagraphLayout you need to build ICU without ParagraphLayout first, then build icu-le-hb, then build again with ParagraphLayout AFAIU
18:56:07 <blathijs> Hm, some store apparently
19:01:00 *** Maraxus has joined #openttd
19:02:27 <TrueBrain> blathijs: people suggest that icu-le-hb resolves the LayoutEngine problem that is no longer in ICU; I suggest we just give it a try, and see what it does
19:02:38 <TrueBrain> from everything I read, HB really did their best to make transition as easy as possible
19:29:17 <Thedarkb-X40> I'm trying to configure the Company Value GS
19:32:17 <Thedarkb-X40> I'm confused as to what units it actually uses.
19:42:17 <blathijs> TrueBrain: I think we do not actually need to change anything to try that, as long as the Debian icu package does the right dance to compile ParagraphLayout against icu-le-hb
19:42:50 <blathijs> TrueBrain: But the icu maintainer seems to imply that it would be better if that dance was not required, which would be the case if OpenTTD stopped using ParagraphLayout
19:43:16 <TrueBrain> sorry to break it to you, but Debian is not our only customer :D
19:43:36 <TrueBrain> we need a fix that works on every OS :)
19:43:53 <TrueBrain> give peter1138 a bit of time to try it out, see what happens
19:44:40 <andythenorth> did anyone patch roadstops that don't need orientation?
19:44:49 <andythenorth> cos it's kind of silly having to choose a direction
19:45:15 <glx> feel free to make a PR andythenorth :)
19:45:17 <blathijs> TrueBrain: True, but it would indeed seem that no changes are required on the OpenTTD side in any case, only in building icu (but I guess for some platforms, building icu is done by us, not the OS vendor indeed)
19:45:48 <peter1138> andythenorth, you mean automatically picking the right orientation?
19:46:12 <glx> I agree it's a smart idea
19:46:14 <andythenorth> there are only a couple of valid bit patterns
19:46:21 <peter1138> TrueBrain, to be fair, it sounds like blathijs has done more reading on it than me now ;)
19:46:31 <andythenorth> if they're not valid, don't build
19:46:39 <TrueBrain> blathijs: I try to switch that we build not 3rd party what so ever .. not ICU, not anything else :P
19:47:44 <glx> TrueBrain: I think vs140 and vs141 vcprojx will need an upgrade for vcpkg
19:48:00 <TrueBrain> glx: that statement is very vague :)
19:48:03 <glx> else it's a pain to use static stuff
19:48:11 <TrueBrain> you mean the library names?
19:48:22 <TrueBrain> I just renamed the bunch :P
19:48:38 <TrueBrain> I think cmake is more promising, and make that work with vcpkg
19:48:42 <blathijs> LordAro: A while ago, you said you looked at pango but it was not worth it and going for harfbuzz directly would be better. But AFAUI Harfbuzz doesn't offer directionality detection and doesn't offer linebreaking, (which Pango and ICU with Paragraphlayout does), so would you suggeste we implement that ourselves on top of Harfbuzz?
19:48:43 <TrueBrain> than we can just deprecated vs141+ :)
19:49:00 <glx> not only the library names, by default vcpkg integration only support dynamic
19:49:25 <TrueBrain> yes; you have to tell vcpkg to use the static triplet
19:49:39 <TrueBrain> not sure if adding stuff to vcprojx is worth it
19:49:39 <Thedarkb-X40> I think there's something broken with Company Value GS
19:49:57 * blathijs is off for a bit. I'll see if I can ping the ICU maintainer, maybe he has an idea how this ICU-loop-mess is going to be fixed on the longer run, or if there are alternative libraries to consider porting to.
19:49:58 <glx> so x64-windows-static stuff is still searched in x64-windows and not found
19:50:19 <TrueBrain> VcpkgTriplet=x64-windows-static
19:50:22 <TrueBrain> is needed as option
19:50:27 <TrueBrain> euh, property they are called :P
19:50:33 <peter1138> blathijs, I looked briefly into Pango, it appears very tidied in with gdk rendering surfaces. I may have been looking at the wrong thing though.
19:50:39 <TrueBrain> vcpkg has a page about how to do static glx :)
19:50:45 <TrueBrain> well, not project perse .. just somewhere
19:50:57 <TrueBrain> but I think it is not worth adding to our project, and that adding cmake support is easier/more flexible
19:51:01 <Thedarkb-X40> It works but the units are wrong.
19:51:23 <TrueBrain> Thedarkb-X40: "units are wrong". This is very vague. A bit more info really helps :)
19:51:50 <Thedarkb-X40> It says it goes up in 1000 pound increments.
19:52:24 <glx> it says to put it in our vcprojx
19:52:58 <TrueBrain> lol, thought they said to put it in your personal thingy
19:53:16 <Thedarkb-X40> But in practice it goes up in £500 increments.
19:53:20 <TrueBrain> anyway, I dont really mind/care :) What everworks for you glx :)
19:53:28 <peter1138> Thedarkb-X40, did you check your currency settings?
19:53:42 <TrueBrain> glx: the buildstuff just adds /p:VcpkgTriplet= to msbuild :D
19:54:32 <peter1138> This cmake stuff, does that somehow replace the vs project file?
19:54:38 <TrueBrain> but if you are going to make Vcpkg fixes, possibly think about a way to also fix the library names :)
19:54:39 <peter1138> Seems odd if it does.
19:54:44 <TrueBrain> it does, completely
19:54:50 <TrueBrain> as far as I understand, anyway
19:55:44 <glx> and I think cmake generates project files too
19:56:14 <TrueBrain> from cmake, you can generate anything basically :)
19:56:27 <TrueBrain> but it does it for you, is the idea: )
19:56:38 <nielsm> I think you have a stub vcxproj that loads the cmake and generates a full msbuild project
20:04:48 <Thedarkb-X40> I don't know why I'm spelling out £
20:05:02 <Thedarkb-X40> I have an Irish keyboard layout.
20:07:10 <Thedarkb-X40> The company value GS just pauses the game in singleplayer.
20:07:20 <Thedarkb-X40> Will it reset it in multiplayer?
20:07:34 <TrueBrain> who is the author of this GS?
20:08:55 <Thedarkb-X40> I think I should probably switch machines if I'm going to watch youtube videos and run openTTD at the same time.
20:08:57 <Thedarkb-X40> Client: HexChat 2.12.4 • OS: Ubuntu "artful" 17.10 • CPU: Intel(R) Pentium(R) M processor 1.40GHz (1.20GHz) • Memory: Physical: 1.2 GiB Total (434.3 MiB Free) Swap: 945.3 MiB Total (740.5 MiB Free) • Storage: 14.6 GB / 31.8 GB (17.2 GB Free) • VGA: Intel Corporation 82852/855GM Integrated Graphics Device @ Intel Corporation 82852/82855 GM/GME/PM/GMV Processor to I/O Contr
20:08:57 <Thedarkb-X40> oller • Uptime: 1d 22h 58m 7s
20:09:18 <TrueBrain> no need for that spam plz
20:09:50 <Thedarkb-X40> Oh, sorry, it's just /sysinfo in hexchat.
20:10:04 <TrueBrain> even if it is a crown on top of a monkey :)
20:10:05 <ST2> eh, he forgot the credit card number, pin, etc xD
20:11:00 *** frosch123 has joined #openttd
20:11:12 <peter1138> That's... a pretty basic machine :p
20:11:33 <peter1138> Hmm, I have more RAM than you have storage, nice.
20:11:47 <Thedarkb-X40> This only takes 1.8 inch hard drives.
20:11:52 <Thedarkb-X40> That's a dead format.
20:12:22 <Thedarkb-X40> So I put in an msata drive in an adapter with blue tack.
20:14:16 <LordAro> blathijs: i have to say i don't know much about this. I was initially turned on to pango by frosch123, but after looking at it some, and looking at its dependencies (glib, among others), decided to try harfbuzz instead. Documentation for either library is basically nonexistent, so i've not really got any idea what features either support
20:20:30 *** ChanServ sets mode: +v tokai
20:20:39 * andythenorth found the original soundtrack
20:20:43 <andythenorth> sounds ok on mac midi
20:23:41 <frosch123> LordAro: blathijs: initially i looked into harfbuzz and its icu wrapper, but concluded that it does not implement what we need. it only layouts a single line, no paragraphs. pango looked like the higher level one, with quite some documnetation, but it needs adjustment to out custom renderer
20:25:17 <TrueBrain> who thought it would be difficult, something as simple as letters :D
20:25:23 *** supermop_work has joined #openttd
20:28:27 *** tokai|noir has joined #openttd
20:28:27 *** ChanServ sets mode: +v tokai|noir
20:36:29 * andythenorth wants 'close dock'
20:36:33 <andythenorth> so I can relocate it
20:39:25 *** ChanServ sets mode: +v tokai
20:41:08 *** gelignite has joined #openttd
20:41:29 <blathijs> frosch123: Yeah, that's also my understanding. However, apparently the paragraph layout still exists in ICU, but it's a mess wrt dependencies. As for pango, I suspect it has a lot of dependencies indeed, which might not make it a good candidate just for wordwrapping...
20:42:03 <peter1138> Yeah, closing any sort of station tile would be useful sometimes.
20:42:05 <Eddi|zuHause> i feel we have had this discussion already
20:42:16 <peter1138> Eddi|zuHause, stations or icu? or both :p
20:42:28 <Eddi|zuHause> i meant the stations
20:42:41 <peter1138> There's probably a patch for it.
20:42:47 <peter1138> Isn't there airport closing?
20:42:56 <Eddi|zuHause> i think we included that
20:43:02 <Eddi|zuHause> but i'm not sure
20:43:09 <Eddi|zuHause> because the discussion was endless
20:43:27 <peter1138> I can't remember why I might've objected to it.
20:43:40 <peter1138> Can't have things that make life easier, eh?
20:43:42 <frosch123> blathijs: which system do you have in mind, which would run a openttd client and does not have pango?
20:43:53 <Eddi|zuHause> whether we should include it like it was, or extend it to some or all the things you listed above
20:44:04 <peter1138> Is OS/2 still around? hehe
20:44:30 <Eddi|zuHause> do we still have a morphos user?
20:44:31 <frosch123> we got some os/2 patches 2 years ago
20:44:53 *** tokai|noir has joined #openttd
20:44:53 *** ChanServ sets mode: +v tokai|noir
20:45:05 <peter1138> Hmm, I guess the thing is "Close airport" works on the whole thing, and is station-window based
20:45:08 <frosch123> anyway, with c++14 we likely drop all of them including 9x
20:45:17 <Eddi|zuHause> i think we dropped morphos when we dropped gcc 2.ancient, but there were talks that "soon" there might be an update to gcc 4
20:45:29 <andythenorth> airport closing is a thing
20:45:39 <blathijs> frosch123: Dunno, I don't have a clear view of the targets OpenTTD still supports :-)
20:45:47 <Thedarkb-X40> I don't know about dropping OS/2
20:45:55 <andythenorth> the problem with docks is that all the ships turn around when you demolish the dock to relocate it
20:45:58 <Thedarkb-X40> ArcaOS is a thing.
20:46:01 <andythenorth> and ship routes are often long
20:46:20 <peter1138> Yeah, but I've had issues relocating roadstops loads of times too
20:46:24 <peter1138> Platforms are a bit easier
20:46:28 <Eddi|zuHause> andythenorth: try build while paused?
20:47:09 <andythenorth> Eddi|zuHause: probably works, haven't tried
20:47:11 <andythenorth> good enough for today
20:47:25 <andythenorth> multi-docks would solve this too :P
20:47:31 <Eddi|zuHause> like <1% of all players
20:47:34 <peter1138> Yes. I had a patc.... oh
20:48:11 <Eddi|zuHause> think maybe we should drop MP? would solve loads of problems :p
20:49:12 <andythenorth> halloumi, rice and chips
20:49:17 <andythenorth> do I miss any major food groups?
20:49:42 <peter1138> "We dropped MP to let you use more NewGRFs"
20:49:55 <Eddi|zuHause> andythenorth: like anything involving meat?
20:49:58 <peter1138> "But then we dropped NewGRFs because modern games use scripting languages instead"
20:50:54 <andythenorth> why does site search 'tt-forums' search ebay in my chrome :P
20:50:56 <Eddi|zuHause> '); drop table trains; --?
20:51:02 <peter1138> Rice and chips is basically one food group though.
20:51:26 <andythenorth> might as well have both
20:51:43 <DorpsGek> andythenorth: zuu was last seen in #openttd 20 hours, 54 minutes, and 11 seconds ago: <Zuu> Oh and I think this whole OpenTTD moving to git is a good thing by the way.
20:51:49 <Eddi|zuHause> andythenorth: i'd argue "chips" is ambiguous
20:52:27 <andythenorth> NoCarGoal should have an option to delete my money
20:52:35 <andythenorth> I usually win it by mass-funding industries
20:52:58 <Eddi|zuHause> you can, like, disable funding?
20:53:14 <andythenorth> then it can't be won
20:53:20 <andythenorth> well, not at the goal levels I set
20:53:23 <andythenorth> there's some interim
20:53:40 * andythenorth wonders if GS can read fund costs
20:54:12 <Eddi|zuHause> probably, but what would it do with it?
20:54:25 *** ChanServ sets mode: +v tokai
20:54:53 <andythenorth> remove money so that you can only fund occasionally
20:55:00 <andythenorth> winning by funding is a valid tactic
20:55:14 <andythenorth> but capping money would incentivise transporting more than just the goal cargos
20:55:41 <andythenorth> I could just put industry prices up :P
21:07:15 <Eddi|zuHause> it's called "balancing", which this game is generally terrible at
21:08:42 <Thedarkb-X40> There's a reason why planes are disabled on my server.
21:09:39 <peter1138> Infrastructure costs?
21:14:48 *** Thedarkb has joined #openttd
21:20:26 <TrueBrain> frosch123: it is normal in CIs that errors are marked as "wrong". So a message always to the stderr seems wrong
21:20:31 <TrueBrain> to the stdout is not a problem
21:21:14 <TrueBrain> frosch123: "no output on success", what do you mean exactly?
21:22:18 <TrueBrain> you can go through life SP just fine
21:22:20 <frosch123> TrueBrain: "stdout" is never visible. So "no output on success" means "no message to stderr, if there is no problem"
21:22:59 <TrueBrain> frosch123: I would very much prefer that any message on stderr is because of an error
21:23:13 <TrueBrain> stdout is for everything else
21:23:14 <Eddi|zuHause> i don't think that kind of classification applies well :p
21:23:32 <TrueBrain> (if that answers your question :D)
21:23:48 <frosch123> i take it as agreement :p
21:23:56 <andythenorth> Lap_op: other MP modes are available
21:24:07 <Eddi|zuHause> that was actually my keyboard slipping away :p
21:24:09 <andythenorth> there are goal servers that have a patched version of the game
21:24:12 <andythenorth> somewhere on reddit
21:24:14 <frosch123> Lap_op: how do you classify minecraft?
21:24:14 <TrueBrain> frosch123: :D I am not fully sure I understand you, but yeah .. we seem to be saying the same :) The patch needs a fix, is my conclusion :)
21:24:22 <andythenorth> citybuilder and so on
21:24:46 <Lap_op> never played minecraft
21:24:59 <frosch123> Lap_op: in that case there are different servers. some do coop, some do ffa, some do ladder
21:25:44 <Lap_op> minecraft is MMORPG as far as i can gather
21:25:46 <frosch123> rpg in ottd is only singleplayer
21:26:02 <TrueBrain> andythenorth: just O? :)
21:26:15 <Eddi|zuHause> i've seen people around here acting like it's an RPG :p
21:27:14 <Pikka> openttd CTF sounds fun though
21:27:34 <TrueBrain> first to the industry wins .. guess it is CTF :P
21:27:43 <andythenorth> TrueBrain said Pikka had rage quit
21:28:02 <TrueBrain> andythenorth believed what TrueBrain said
21:28:11 <andythenorth> but the logs said it too
21:28:12 <Lap_op> if there was a network tycoon mod CTF would make plenty of sence
21:28:52 * andythenorth wonders about a CTF gamescript
21:28:58 <Eddi|zuHause> maybe something asymmetric like in Team Fortress where one company must connect two industries and delver some cargo, and the other company must do all in its power to prevent it?
21:29:19 <peter1138> Nah, it's too easy to prevent it.
21:29:20 <andythenorth> I was thinking of forcing a start on opposite sides of the map
21:29:20 <TrueBrain> some MPs feel like that :P
21:29:24 <andythenorth> and capture the centre
21:29:33 <andythenorth> but with high building costs or something
21:29:40 <andythenorth> can GS modify building costs for part of the map?
21:29:42 <Lap_op> were can i get the bots to play for me?
21:29:57 <andythenorth> Lap_op: there are AIs
21:30:02 <andythenorth> downloadable content
21:30:02 <Eddi|zuHause> how about OpenTTDota?
21:30:06 <andythenorth> not sure if any servers run AIs
21:30:52 <Eddi|zuHause> you can't have a bot in a player company
21:31:22 <Lap_op> you can if you run the server
21:31:37 <Eddi|zuHause> Pikka: there's like one line of code to prevent it :p
21:32:00 <Lap_op> since its opensource you could edit that line out
21:32:20 <Pikka> you can switch into ai companies in singleplayer
21:32:35 <Eddi|zuHause> Lap_op: well technically, even any client could comment that line out
21:32:55 <Pikka> but unless the ai is very simple or was written to be cooperative, it's unlikely to be very playable... :)
21:33:09 <Lap_op> so the AI is readily available
21:33:55 <Eddi|zuHause> there's no real way for the server to make sure the commands it receives are from a human
21:34:15 <peter1138> Wasn't the AI system conceived so that AIs could fight it out?
21:34:20 <Eddi|zuHause> possibly there are rate limits
21:34:42 <Eddi|zuHause> peter1138: i think that's where we got Rondje from :p
21:35:13 <Lap_op> you just need the AI to start easy and keeps growing to tackle the real player at the most infuriating point of his develpment plan
21:35:42 <Lap_op> you know the boiling frogs metodology
21:35:54 <LordAro> which is an urban myth
21:36:27 <LordAro> the boiling frogs thing
21:36:32 <Eddi|zuHause> there once was a patch that placed a sign like "hey i'm watching you", whereever a player did something
21:36:52 <Lap_op> you realize boiling frogs is a metaphore?
21:36:57 <LordAro> the scientist had to basically lobotomise the frogs to stop them from jumping out of the water
21:37:29 <Pikka> Eddi: I want AIs to have access to the chat (in SP too), so they can more easily trash-talk
21:37:33 <TrueBrain> Rubidium: nice :D Now make a PR out of it so I can review it properly :D
21:37:35 <Lap_op> it makes perfect sense
21:37:55 <Pikka> signs and the debug log just aren't enough :P
21:38:18 <Eddi|zuHause> Pikka: would that look anything like the chat in Goat Simulator RPG?
21:38:30 <Lap_op> must be smiling faces season over here
21:39:06 <Pikka> it's always a happy place in here, Lap_op
21:40:11 <Lap_op> who can you teall its always happy if you dont have comparison point?
21:41:18 <Eddi|zuHause> if the tree falls in the wood and makes no sound, how can you know nobody is watching?
21:41:19 <Lap_op> acording to einstatein everything is relative
21:41:34 <Pikka> there's plenty to compare it to. I think most of us have experience of life outside this irc channel...
21:41:52 <LordAro> outside? i don't understand
21:42:18 <Thedarkb-X40> Outside.... you mean the OpenTTD subreddit, right?
21:42:19 <Eddi|zuHause> maybe he means other channels?
21:42:22 <Lap_op> Eddi|zuHause: you probablly think you are deaf
21:42:35 <peter1138> Eddi|zuHause, Goat MMO Simulator :D
21:42:36 <Lap_op> Eddi|zuHause: you probablly think you gone deaf
21:42:45 <peter1138> The positioning of the word Simulator is important.
21:42:49 <Eddi|zuHause> peter1138: ah that's the name
21:42:53 <Eddi|zuHause> peter1138: i forgot
21:43:12 <peter1138> The trash talk was quite convincing.
21:43:32 <Eddi|zuHause> peter1138: i played it on a free weekend, was about all of Goat Simulator i could suffer :p
21:43:57 <Thedarkb-X40> That must be it.
21:44:12 <peter1138> Eddi|zuHause, oh yes, it's crap
21:44:33 <andythenorth> I made a goat game once
21:45:23 <Eddi|zuHause> anyone remember when there were half a dozen Moorhuhn games?
21:46:07 <Thedarkb-X40> LGR has a boxed copy of goat sim.
21:46:12 <Thedarkb-X40> God only knows where that's from.
21:46:39 <Lap_op> this channel is becoming sad all of a sudden
21:47:32 <Lap_op> whats about all this goat talk?
21:47:44 <TrueBrain> you do know which channel you joined, right?
21:49:03 <Lap_op> ive seen a movie about goats
21:50:39 <Lap_op> The Men Who Stare at Goats
21:52:02 <Lap_op> seems like this is the only live channel on this network
21:52:09 <peter1138> Hmm, if I add GUI sprites to the game, should I put them in openttdgui.png or a separate image?
21:52:18 <andythenorth> is the movie any good?
21:52:28 <peter1138> I guess in the same image is how we've always done it.
21:52:42 <frosch123> peter1138: if you add them to the ottdgui action5, also add them to openttdgui.png imo
21:53:18 <peter1138> Is there such a thing as an image diff? :D
21:53:20 <andythenorth> Lap_op: book is ok
21:53:29 <andythenorth> peter1138: in photoshop, yes :P
21:53:34 <andythenorth> it doesn't do what you want though
21:53:39 <andythenorth> you could write one in PIL
21:54:28 <peter1138> The image you sent me is nearly twice the size.
21:54:44 <andythenorth> possibly terrible compression
21:55:21 <andythenorth> photoshop has a habit of bloating pngs if not kept on a leash
21:56:00 <andythenorth> peter1138: how big is it for you in KB?
21:56:13 <peter1138> Ah, the layer modes don't work in indexed mode.
21:56:50 <peter1138> I reexported it in Gimp and it's now 26KB
21:57:00 <Thedarkb-X40> one of my neighbours has a goat.
21:57:02 <andythenorth> yeah I got the same in photoshop
21:57:06 <Thedarkb-X40> It looks angry all the time.
21:57:12 <andythenorth> let's hope it's supposed to be DOS palette eh :P
21:57:33 <Thedarkb-X40> Imagine OpenTTD in EGA
21:57:47 <peter1138> Hmm, which palette though?
21:57:58 <Thedarkb-X40> That awful one used by sopwith
21:58:08 <Thedarkb-X40> Some CGA games looked very good.
21:58:13 <peter1138> Hmm, actually it had a 64 colour mode. Hmm.
21:59:59 <glx> TrueBrain: using vcpkg static libs (except icu still from useful-6.0) openttd.exe x64 release is 13.4MB
22:00:07 <andythenorth> look at the refresh rate :P
22:00:25 <TrueBrain> glx: hmm .. so it is either ICU, or VS2017 ...
22:00:29 <TrueBrain> closing in on the issue :D
22:00:44 <TrueBrain> tomorrow I will see if I can locally build it on VS2017
22:01:33 <peter1138> andythenorth, my BBC Micro had a hardware flaw: in Mode 2, colour 6 defaulted to green instead of cyan.
22:01:49 <peter1138> It could be fixed by updating the palette, but of course, nothing ever did that.
22:01:55 <peter1138> It made jcbdigger a bit harder
22:01:58 <andythenorth> I think that game caused HEQS :P
22:05:14 <peter1138> Warning on sprite 1533 (level 1).
22:05:19 <peter1138> Wow, that's useful.
22:07:07 <frosch123> yes, that warning is the reason i think we should drop renum
22:07:43 <frosch123> too much chicken/egg for just getting sprite 0 right
22:09:28 <peter1138> Hmm, sprite 0 is 0 as it should be.
22:10:50 <frosch123> we only need renum to put the "number of sprites in the grf" into sprite 0
22:11:25 <frosch123> we always get the renum warning "you added more ottd gui sprites, than renum thinks ottd expects"
22:11:40 <frosch123> which is a bit stupid :p
22:11:43 <peter1138> could we just disable the warning?
22:11:49 <peter1138> (which i've done for now)
22:12:05 <frosch123> if there is such option, it would make sense
22:12:17 <peter1138> It's kinda correct but yeah, makes adding things harder.
22:17:39 <peter1138> andythenorth, damn, those icons are right for the group window but the livery window ones are a bit larger ;(
22:17:41 <Eddi|zuHause> SimCity 1 was available as EGA game
22:17:58 <andythenorth> want me to do something?
22:18:10 <Thedarkb-X40> I can't understand why everyone used shitty palettes on EGA?
22:18:47 <peter1138> Thedarkb-X40, the palette could only be changed in 640x350 mode.
22:18:50 <andythenorth> looks like livery window icons are same size as toolbar
22:18:54 <Eddi|zuHause> "EGA" for me always meant 64 colours
22:19:02 <Eddi|zuHause> i don't know what paletes you mean
22:19:10 <Thedarkb-X40> I was thinking because a few CGA games look good.
22:19:23 <peter1138> EGA has 16 colours, but in 640x350 those colours could be chosen from a total of 64 colours.
22:19:28 <andythenorth> group window icons are some arbitrary amount smaller :P
22:19:56 <andythenorth> the proportions are bad on the group window icons anyway :P
22:20:04 <peter1138> 18x18 there, for some reason
22:20:05 <andythenorth> they're too narrow
22:20:12 <peter1138> Maybe we should just pad them to 20x20 like the rest?
22:20:29 <Eddi|zuHause> i'm fairly sure there was also a way to display 64 colours at the same time... but my memory might be off after all this time
22:20:30 <andythenorth> do it in trunk :P
22:20:40 <peter1138> Eddi|zuHause, I thought so too but it seems not.
22:20:58 <Thedarkb-X40> I didn't realise you could only change the palette in 640x350
22:21:13 <peter1138> Thedarkb-X40, it's not something that's relevant now :)
22:21:24 <Eddi|zuHause> there was also almost definitely also a way to change palette in text mode
22:21:35 <Thedarkb-X40> I can remember doing that in QBASIC.
22:21:42 <Eddi|zuHause> possibly with one fewer also
22:21:42 <Thedarkb-X40> On a CGA though.
22:21:49 <Thedarkb-X40> CGA with a monochrome screen.
22:21:55 <Thedarkb-X40> Which is basically just useless.
22:22:03 <Eddi|zuHause> CGA was weird, you had 3 palettes of 4 colours?
22:22:32 <peter1138> Fixed palettes. Wonderful.
22:22:38 <Eddi|zuHause> and most people used the one with the weird pink
22:22:50 <Thedarkb-X40> Except for Windmill software.
22:22:51 <peter1138> Yeah, "white", cyan and magenta was the default.
22:22:58 <Thedarkb-X40> They could make CGA look good.
22:23:12 <Eddi|zuHause> i don't think there was white in that
22:23:21 <peter1138> Rarely saw red, cyan and white.
22:23:30 <peter1138> And the other one was red green and yellow.
22:23:36 <Eddi|zuHause> there was also one with a brown?
22:23:56 <peter1138> Hmm, I think there may have been dark and light versions then.
22:24:08 <Eddi|zuHause> but VGA was also a bit silly
22:24:23 <Eddi|zuHause> because it somehow lost the colour flexibility from EGA?
22:24:42 <Eddi|zuHause> and 256 colours were only available with super low res
22:24:50 <peter1138> Yes, there were dark and light versions.
22:25:11 <peter1138> Yes, because there was not enough memory for 256 colours in high-res.
22:25:38 <glx> like the cpc and its 16 colours and big pixels
22:25:40 <peter1138> 8bpp in 640x480 is more than that.
22:25:51 <peter1138> Also 16 colour mode was some horrible planar thing.
22:25:55 <Eddi|zuHause> yes, but somehow the EGA-way of defining colours was also not available in VGA, so you had the fixed 16 colours
22:26:23 <peter1138> I think it may have been.
22:26:27 <peter1138> But nothing used it.
22:26:50 <Eddi|zuHause> well any VGA card you could run in EGA mode
22:26:53 <peter1138> "The 640×480 16-color and 320×200 256-color modes had fully redefinable palettes, with each entry selectable from within an 18-bit (262,144-color) RGB table"
22:28:57 <peter1138> And then there was MCGA.
22:29:02 <peter1138> Wow, this all totally relevant.
22:29:24 <Eddi|zuHause> well, i suppose some games did use that for animated colours
22:29:39 <peter1138> Hmm, Windows Store version of OpenTTD. I wonder what's involved with that.
22:30:03 <Eddi|zuHause> i think there was a forum topic a la "don't give money to this person"
22:30:11 <glx> I think orudge tried to talk to MS about it
22:31:19 <peter1138> Well, MCGA looked the same as the 256 colour VGA mode, but subtle differences behind the scenes.
22:31:57 <peter1138> Eddi|zuHause, erm, well, there's a forum post asking for OpenTTD on the store. I just wondered what's involved in making that.
22:32:10 <nielsm> biggest challenge in a windows store version will probably be file handling, both downloads, savegames, and perhaps screenshots and the like
22:32:27 <peter1138> I'm sure Windows Store titles can download things.
22:32:35 <nielsm> sure they can, it's where they store them
22:32:40 <peter1138> They probably have a sandboxes local file storage, surely?
22:32:41 <nielsm> and what parts of the FS they can access
22:32:55 <nielsm> you can't just fopen() random things
22:32:56 <peter1138> Doesn't need anything other than its own storage.
22:33:24 <glx> I think they have limited local storage space
22:33:33 <Eddi|zuHause> nielsm: openttd can easily be configured where to store files other than "Documents"
22:34:04 <nielsm> I'm not saying it will be a gigantic challenge, just that it will probably be the part that requires most care compared to everything else
22:34:35 <glx> anyway store apps have some limitations compared to desktop apps
22:35:04 <peter1138> So, let's get their source to see what's changed. And then if possible make an official free version.
22:35:14 <peter1138> I dunno about licensing, mind you.
22:39:26 <glx> oh the "disclaimer" on top of the description was not there last time I checked
22:48:50 <Thedarkb-X40> What's the point of the Windows store?
22:49:12 *** sim-al2 has joined #openttd
22:52:59 <nielsm> from a user's point of view, installing things from the windows store should be safer, since everything from there runs sandboxed and (barring bugs in the sandbox) can't affect the OS as a whole
22:53:16 <nielsm> which also means uninstalling an app installed from it really does remove everything
22:54:31 <Lap_op> the point of windows store is to mimic apple store and android store
22:55:35 <peter1138> But there are some Windows editions that can only use the store. Some cut down tablet thing?
22:55:49 <Eddi|zuHause> i'm fairly sure lots of "less literate" computer users benefit from a single "store"
22:56:25 <Eddi|zuHause> peter1138: yeah, some school version of windows was (meant to be?) store only
22:56:26 <Lap_op> linux works the same way
22:57:40 <Lap_op> linux store = distro repository
22:58:35 <Lap_op> albeit repositories can be mirrored
22:58:52 <nielsm> not really, no, the way package managers traditionally work on linux distros is closer to MSI (windows installer) in that the package manager has a database of files installed on the system, but files from all packages are installed mixed together in the directory structure
22:59:18 <nielsm> sure there's then some automated downloading and dependency resolution stuff on top of that
22:59:33 <nielsm> but packages are not isolated from each other
23:00:04 <nielsm> however if distros switch over to packaging everything in docker containers/similar things, then you have a comparison
23:01:58 <Lap_op> there is a distro that already installs every program to its own dir
23:02:13 <Lap_op> like windows \programs\
23:03:35 <andythenorth> specific engines per cargo class would be dumb, right?
23:03:37 <Eddi|zuHause> "separate directory" is something very different from "isolated virtual machine"
23:03:46 <Eddi|zuHause> andythenorth: yes
23:04:01 <andythenorth> the 'one obvious choice' rule for Iron Horse engines works
23:04:19 <andythenorth> there are a lot of freight cargos, and it gets monotonous choosing the same engine over and again
23:04:37 <Lap_op> i dont think android apps run in isolated virtual machines
23:04:37 <andythenorth> pax has more variety, depending on the route
23:05:25 <nielsm> they do have some level of sandboxing
23:05:42 <nielsm> seLinux based I think?
23:08:36 <nielsm> (if you have bugs in a sandbox, allowing apps to leak out/data to leak in, are they then called termites?)
23:08:47 <Lap_op> google took linux and turned its free philosophy around
23:09:26 <Lap_op> running rootkits in a VM hmm
23:09:46 <Lap_op> should avoid detection
23:10:48 <Lap_op> arguably even more difficult to remove
23:11:22 <Lap_op> the VM is the horse and trojan is inside it
23:13:29 <Lap_op> and since most if not all of them apps are always going online...
23:28:47 *** Mahjong has joined #openttd
continue to next day ⏵
