Config
Log for #openttd on 20th January 2019:
Times are UTC Toggle Colours
00:02:01  <glx> on build with untouched source, it just runs the settings vbs
00:02:20  <Samu> https://imgur.com/xnPVn6k
00:02:44  <glx> hmm no it's the version vbs
00:03:08  <Samu> ehm, i need paint
00:03:10  <Samu> brb
00:05:46  <Samu> paint won't do it
00:07:40  <nielsm> now the vs2019 preview installer wants to reboot
00:07:48  <glx> hehe
00:07:49  <nielsm> just going to shut down then
00:07:57  <nielsm> should be sleeping anyway
00:08:08  <glx> for windows update shutdown is not a reboot
00:08:29  <nielsm> yea...
00:08:32  <glx> cost me hours when I did it for 1803 upgrade
00:08:52  <glx> had to redo the upgrade from start
00:09:09  <nielsm> yea you can tell it to "update and shut down", but that's only the preparation phase it does then
00:09:15  <nielsm> still has to do more work on next bot
00:09:18  <nielsm> +o
00:09:31  <nielsm> anway, gn
00:09:34  *** nielsm has quit IRC
00:22:32  *** Progman has quit IRC
00:23:44  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuG0
00:23:59  <Samu> omg, so small
00:38:12  *** Thedarkb-T60 has joined #openttd
00:59:28  *** Wormnest has joined #openttd
01:45:59  *** Wormnest has quit IRC
02:05:39  *** Speedyn has quit IRC
02:10:17  *** Speedy` has joined #openttd
02:28:06  *** glx has quit IRC
02:54:14  *** Samu has quit IRC
03:55:24  *** D-HUND has joined #openttd
03:58:47  *** debdog has quit IRC
04:05:00  *** Thedarkb1-T60 has joined #openttd
04:11:09  *** Thedarkb-T60 has quit IRC
05:30:11  *** HerzogDeXtEr has joined #openttd
05:43:13  *** triolus has joined #openttd
06:01:57  *** sla_ro|master has joined #openttd
06:02:10  *** triolus has quit IRC
06:18:53  <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on issue #6173: Add support for SDL2 https://git.io/fhuR3
06:20:06  *** triolus has joined #openttd
07:19:59  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on issue #6173: Add support for SDL2 https://git.io/fhuRx
07:22:04  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on issue #6173: Add support for SDL2 https://git.io/fhuRh
07:25:13  *** andythenorth has joined #openttd
07:26:17  <peter1138> Morning.
07:26:48  <andythenorth> moin
07:27:07  <peter1138> LordAro, I thought someone (a dev) had already made an SDL2 patch.
07:27:56  <peter1138> Just, incomplete. About 5-6 years ago, too. And it wasn't me.
07:28:24  <andythenorth> oof
07:28:34  <andythenorth> big diff, but I want this feature https://github.com/OpenTTD/OpenTTD/pull/7073
07:28:39  * andythenorth tests it
07:29:11  <peter1138> It's not that big, but does need review.
07:29:40  <peter1138> Has some weirdly named functions and variables.
07:29:45  <peter1138> FlowsLock
07:29:47  <peter1138> t_2dm
07:30:37  <Gustavo6046> looks pretty cool
07:31:00  <andythenorth> I just worry about side effects
07:31:27  <peter1138> What sort of side effects.
07:31:48  <andythenorth> for example, it seems to modify town behaviour for bridges
07:31:53  <andythenorth> so do towns fail to grow?
07:31:56  <andythenorth> etc
07:32:15  <andythenorth> it's not just fixing isolated behaviours one at a time
07:32:42  <peter1138> *nod*
07:32:57  <andythenorth> it also plants extra river tiles
07:33:07  * andythenorth tests it anyway
07:34:24  <Gustavo6046> 150 wpm gang
07:34:29  <Gustavo6046> ops wrong channel
07:35:42  <andythenorth> eh the 7073 patch works really well
07:35:58  <andythenorth> the extra river tiles make rivers more random and interesting
07:38:32  <peter1138> It sounds like 2 separate features combined.
07:40:14  <andythenorth> hmm
07:40:20  <andythenorth> no way to know if the patch causes this https://dev.openttdcoop.org/attachments/download/9242/Unnamed,%2001-01-2000%233.png
07:40:47  <andythenorth> river generation is imperfect anyway, but more seem to be failing to find a route to sea
07:40:57  <andythenorth> 'more' is very subjective :(
07:42:04  <andythenorth> same map https://dev.openttdcoop.org/attachments/download/9243/Unnamed,%2001-01-2000%234.png
07:43:54  <andythenorth> pretty reproducible actually
07:44:06  * andythenorth tries master
07:44:57  <LordAro> peter1138: yeah, i seem to recall something like that as well. maybe rubidium?
07:45:35  <andythenorth> probs this thread if any https://www.tt-forums.net/viewtopic.php?f=32&t=69650
07:46:37  <andythenorth> presumably we're discussing this? o_O https://github.com/nikolas/OpenTTD/tree/sdl2
07:50:31  <andythenorth> hmm samu's patch
07:50:36  * andythenorth had better read it
07:50:47  <andythenorth> I guess it's banning rivers from contiguous slopes
07:52:51  <andythenorth> I'm guessing this https://github.com/OpenTTD/OpenTTD/pull/7073/commits/04bcae89d4ddece7d5730f739e4550fae363dba5#diff-2ade113f29822f3daae5176be52a7e6eL1012
07:53:09  <peter1138> I was going to try to make a GLFW driver at one point.
07:54:05  <andythenorth> :o
07:54:08  <andythenorth> :)
07:54:49  <andythenorth> oof 7073 is really frustrating
07:55:05  <andythenorth> it produces rivers that are (1) more navigable and (2) mostly look better
07:55:18  <peter1138> But?
07:55:19  <andythenorth> but also (3) are too often broken, with no route to sea
07:55:29  <peter1138> Ah.
07:55:30  <andythenorth> and look stupid because they stop at the top of a hill
07:56:25  <andythenorth> https://dev.openttdcoop.org/attachments/download/9244/Unnamed,%2001-01-2000%235.png
07:56:57  <andythenorth> https://dev.openttdcoop.org/attachments/download/9245/Unnamed,%2001-01-2000%236.png
07:57:03  <peter1138> Yers
07:57:10  <andythenorth> so yeah, another useful looking patch, that tries to boil the ocean and fails
07:58:03  <andythenorth> samu is 28% of open PRs
07:58:13  <andythenorth> there's not one that I can try and push through yet
08:00:06  <LordAro> preferring the navigable paths might help?
08:00:23  <LordAro> rather than only
08:01:24  <peter1138> Urgh, this laptop display... Keep having to change the angle of it. TN panels :(
08:03:06  <andythenorth> does anybody make a decent laptop anymore?
08:10:50  <peter1138> Eh, this one is about 8 years old :p
08:11:42  <andythenorth> probably has working keys
08:11:46  <andythenorth> a feature mine lacks
08:16:50  * peter1138 hungers.
08:17:44  <peter1138> But there's not a lot of food here. Ketchup on weetabix is a bit weird.
08:18:55  <andythenorth> deliveroo
08:19:00  <andythenorth> order porridge
08:22:48  <LordAro> peter1138: that's literally the worst thing i've ever heard of
08:23:08  *** Wolf01 has joined #openttd
08:23:18  <Wolf01> o/
08:23:29  * LordAro aways on a bike
08:23:56  <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth requested changes for pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuEu
08:31:45  <DorpsGek_II> [OpenTTD/OpenTTD] michicc commented on pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuEo
08:32:50  <peter1138> :D
08:33:01  <peter1138> I didn't try it either.
08:33:17  <peter1138> I did have just plain weetabix, though.
08:34:17  <peter1138> + hot water.
08:34:50  * andythenorth wants porridge
08:38:06  <peter1138> Turns out plain weetabix with hot water is actually nice. Who knew?
08:39:52  <andythenorth> basically porridge :P
08:41:25  <peter1138> Wheat rather than oats.
08:43:25  <andythenorth> "details"
08:43:32  * andythenorth has porridge
08:50:40  *** nielsm has joined #openttd
08:51:40  <nielsm> morning
08:51:48  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #6980: GDI engine for font glyph rendering as a replacement for FreeType https://git.io/fhuEb
08:52:38  <nielsm> I've come closer to why VS is constantly rebuilding some files, set msbuild logging level to "detailed" and found this:
08:52:39  <nielsm>          ..\src\music\emu\opl_ks.cpp will be compiled because it was not found in the tracking log.
08:52:39  <nielsm>          ..\src\music\emu\opl_nuked.cpp will be compiled because it was not found in the tracking log.
08:53:14  <nielsm> the tracking log is one of those .tlog files it writes, has name of a compilation unit on one line, followed by line of all other files it references on the following lines
08:53:28  <nielsm> for some reason those two files are never written to the tracking log as compilation units
08:54:23  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fhuEA
08:58:55  *** gelignite has joined #openttd
09:08:57  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7074: Change 81330b8d6e: Cached path only needs to be cleared if it was found to be invalid. https://git.io/fhuum
09:09:29  *** Progman has joined #openttd
09:10:47  <peter1138> I looked at caching NPF, it needs to delve much deeper due to how NPF works.
09:13:17  *** andythenorth has quit IRC
09:15:49  *** andythenorth has joined #openttd
09:25:15  <andythenorth> wow
09:25:22  <andythenorth> once again I am surprised by reddit
09:25:41  <andythenorth> https://www.reddit.com/r/openttd/ is mostly full of nice people asking reasonable questions
09:25:51  <andythenorth> what happened to 'reddit is the sewer of the internet'
09:25:52  <andythenorth> ??
09:28:45  *** tokai|noir has joined #openttd
09:28:45  *** ChanServ sets mode: +v tokai|noir
09:35:24  *** tokai has quit IRC
09:49:51  <DorpsGek_II> [OpenTTD/OpenTTD] michicc approved pull request #7074: Change 81330b8d6e: Cached path only needs to be cleared if it was found to be invalid. https://git.io/fhuuH
09:50:24  <andythenorth> michi_cc: you don't have merge rights? :o
09:51:40  <michi_cc> I have, but then my name is attached to the commit :) Devs with commit rights can generally eat their own dog food :p
09:52:13  <michi_cc> Especially if it's apparent they are around.
10:06:36  <DorpsGek_II> [OpenTTD/OpenTTD] michicc merged pull request #7074: Change 81330b8d6e: Cached path only needs to be cleared if it was found to be invalid. https://git.io/fhuum
10:06:44  <michi_cc> andythenorth: Happy now? ^^^
10:06:49  <andythenorth> :D
10:12:22  <andythenorth> BBL
10:12:24  *** andythenorth has quit IRC
10:40:19  *** sla_ro|master has quit IRC
10:44:37  <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7074: Change 81330b8d6e: Cached path only needs to be cleared if it was found to be invalid. https://git.io/fhugs
10:47:29  <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7072: Change: Add path cache for ships https://git.io/fhugW
10:56:00  *** frosch123 has joined #openttd
11:49:00  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuaI
11:52:09  * peter1138 returns.
11:52:34  <peter1138> I was around ;)
11:53:12  <peter1138> No wait. I *wasn't* around.
11:54:22  <nielsm> https://0x0.st/sFrZ.jpg  stop perching on my phone... silly bird
12:00:50  * LordAro reappear
12:00:52  <LordAro> it brr
12:01:09  <nielsm> 1>------ Up-To-Date check: Project: strgen_vs141.vcxproj, Configuration: Debug Win32 ------
12:01:09  <nielsm> 1>Project is not up-to-date: build input 'C:\Users\nielsm\Dev\vcpkg\scripts\buildsystems\msbuild\vcpkg.targets' is missing.
12:01:29  <nielsm> no what? why? that file exists and is readable and has a modify date in the past
12:07:35  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #6965: Add: Four alternative town cargo generation methods https://git.io/fhuwY
12:09:21  *** andythenorth has joined #openttd
12:09:51  <peter1138> Andy returns.
12:11:57  <peter1138> Hmm, liveries...
12:12:18  <peter1138> Need to fix that GUI.
12:14:23  <andythenorth> innit
12:14:30  <andythenorth> or I could just redesign it all
12:14:37  <andythenorth> :o I could learn the GUI code :P
12:16:04  <nielsm> more wtf https://0x0.st/sFrw.png
12:16:27  <nielsm> 1. why does it uppercase the path
12:16:34  <nielsm> 2. why does that prevent it from being opened?
12:18:00  <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuwA
12:18:14  <peter1138> andythenorth, it was redesigned :p
12:18:29  <peter1138> andythenorth, just my plan of having a resizable matrix didn't quite work.
12:18:34  * andythenorth finds the fork
12:18:40  <andythenorth> eh I have you upstream anyway
12:19:03  <andythenorth> group-livery?
12:19:16  <peter1138> It's out of date.
12:19:17  <andythenorth> ugh fucking .obs files :|
12:19:30  <peter1138> I have a local one with the sprite updates split up.
12:19:38  <peter1138> Maybe I should force-push. Hmm.
12:19:39  *** Thedarkb1-T60 has quit IRC
12:20:25  <peter1138> andythenorth, I have a fix for that but it'll break VS builds, and whatever other builds we have that don't use the Makefiles.
12:20:45  <andythenorth> oof
12:21:03  <peter1138> What methods of building are there on OSX?
12:21:15  <andythenorth> I only know about make?
12:21:18  <peter1138> I just need to translate the awk file into a vbs.
12:21:47  <andythenorth> awk is available for me
12:21:53  <andythenorth> dunno if it's installed by default
12:22:08  <peter1138> It probably is.
12:22:10  <andythenorth> probs
12:22:13  *** Thedarkb-T60 has joined #openttd
12:23:07  <peter1138> If not we could just write a helper program in C++ that does the work, and that should work on any platform.
12:23:26  <peter1138> The issue is if grfcodec is necessary.
12:25:36  <peter1138> Was there an xcode project file at some point or something?
12:26:23  <peter1138> Ah, looks like if there was, xcode called make anyway, so it's just Windows that needs tweaking.
12:26:49  <peter1138> It'll build without the .obs files but won't run.
12:32:22  <andythenorth> :P
12:35:14  <peter1138> Core 2 Duo laptops are slow :/
12:35:25  <peter1138> That NoCab save still runs horribly slow :D
12:35:37  <peter1138> Haven't tried it without the cache though.
12:37:04  <peter1138> Hmm, I'm missing a nice non-sample-based GM synth on Linux :/
12:37:39  <peter1138> There is a 100MB OPL3 soundfont which is not quite what I want.
12:37:52  <nielsm> merge my adlib-music branch? :3
12:38:00  <nielsm> it (mostly) works now
12:38:40  <nielsm> of course that needs adlib.cat
12:43:18  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7065: Change: Make ships stop and change direction slowly instead of instantly turning. https://git.io/fhuro
12:43:37  <peter1138> Yeah I only have OpenMSX here.
12:45:59  <andythenorth> I need a tame train nerd
12:46:04  <andythenorth> for advice
12:47:12  <nielsm> gah nope can't find anything showing why devenv.exe likes to uppercase specifically that vcpkg.targets filename
12:48:36  <nielsm> wtf-isms https://0x0.st/sFr5.png
12:50:12  <LordAro> how odd
12:50:21  * andythenorth had forgotten about windows
12:50:25  <andythenorth> C: and stuff
12:50:58  <andythenorth> probably the wrong tool to build civilisation with :P
12:52:28  <nielsm> okay, it seems to be caused by the git checkout having been made with WSL
12:52:59  <nielsm> which makes some weird case-sensitivity things in ntfs
12:53:23  <nielsm> I now made new directories for every level of the path and moved the contents into the new dirs
12:53:35  <nielsm> deleted the old now empty dirs, and renamed the new to the old names
12:53:38  <nielsm> then it works
12:53:47  <nielsm> it was not good enough to rename the original dirs
12:54:16  <nielsm> I'm going to submit a bug report for vs2019 preview
12:55:59  <LordAro> nice.
13:05:09  *** Thedarkb-T60 has quit IRC
13:07:30  <nielsm> and the issue was similar with the two CPP files constantly being rebuilt
13:07:46  <nielsm> moved them to a new dir and renamed that to match the old name
13:07:52  <nielsm> now they don't fail all the time
13:18:44  <peter1138> o_O
13:26:19  *** Thedarkb-T60 has joined #openttd
13:28:16  *** Thedarkb-T60 has joined #openttd
13:28:33  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7065: Change: Make ships stop and change direction slowly instead of instantly turning. https://git.io/fhWk0
13:30:07  *** gelignite has quit IRC
13:33:48  * LordAro continues being a webdev https://i.imgur.com/Lt8jvyl.png
13:34:56  <peter1138> Fix the underline on the top right!
13:35:02  <LordAro> already done :p
13:36:02  <peter1138> Ship it!
13:36:17  <LordAro> it looks a lot worse at the bottom of the page :p
13:37:45  <peter1138> The whole thing looks a bit dated these days.
13:38:04  <peter1138> Right, off to visit the MiL again.
13:38:09  <LordAro> F
13:43:29  *** Flygon has quit IRC
13:51:03  <nielsm> btw has ottd been Elfring'd yet? https://github.com/numpy/numpy/issues/12673
13:51:49  <LordAro> oh yes
13:56:13  <frosch123> oh, there is a verb for that guy :p
13:57:28  <LordAro> https://lkml.org/lkml/2017/11/28/690 "That said, I'm considering Markus Elfring being a human." "Thanks for this view."
13:57:31  <LordAro> heh
13:58:24  <LordAro> nielsm: https://github.com/OpenTTD/OpenTTD/issues?utf8=%E2%9C%93&q=is%3Aissue+author%3Aelfring+ but yes
13:59:28  <LordAro> istr they appeared here as well
14:01:01  <LordAro> http://webster.openttdcoop.org/index.php?channel=openttd&date=1523836800#1523908299 indeed
14:01:34  <LordAro> it honestly could just be a well trained markov chain
14:03:22  *** Thedarkb-T60 has quit IRC
14:27:46  *** Thedarkb-T60 has joined #openttd
14:27:57  <nnyby> elfring on sourceforge as well >_< https://sourceforge.net/p/cppcheck/discussion/development/thread/e325c49216/  what a truly annoying bot
14:35:47  *** Samu has joined #openttd
14:37:24  <Samu> gi
14:47:43  <andythenorth> so I have to draw this for Horse
14:47:45  <andythenorth> but
14:47:49  <andythenorth> it's ugly AF https://photos.smugmug.com/Trains/Daves-APT-P-travels/i-GHx9Hg3/0/cbde1531/L/APT-P_Quintinshill_10-83-L.jpg
14:47:58  <andythenorth> and I am not motivated to do it :P
14:48:13  <nielsm> that's a dirty train
14:49:32  <Samu> Data type of TileOffsByDiagDir.
14:49:36  <Samu> data type?
14:51:27  <Samu> ah, TileIndexDiff?
15:06:03  <Samu> Split into two functions. I.e. one that does the above tile checks and the main recursive function.
15:06:08  <Samu> but but...
15:06:15  <Samu> i'm gonna repeat variables
15:06:35  <Samu> gonna repeat delta_mid and dir
15:06:41  <Samu> i thought repetition was bad
15:10:09  <andythenorth> of vars?
15:11:23  <Samu> https://paste.openttdcoop.org/pvjabr0nl
15:11:49  <Samu> 3 and 6 are repeated at 23 24
15:12:33  <Samu> i will use them for the main function
15:23:18  *** dvim has joined #openttd
15:29:14  <LordAro> "OpenTTD is officially supported on the following operating systems." ... "Solaris"
15:29:17  <LordAro> hmm.
15:29:32  <LordAro> also missing MacOS
15:36:44  <Samu> michi_cc what do you mean by variable name? https://github.com/OpenTTD/OpenTTD/pull/7073#discussion_r249269002
15:40:41  <LordAro> Samu: presumably not descriptive enough
15:41:24  <Samu> i need more names?
15:41:39  <LordAro> "t2_dm" does not say what the variable is for
15:41:45  <LordAro> t_2dm*
15:41:54  <Samu> tile_2 times delta middle
15:43:31  <LordAro> regardless, i think you should look at the comments further down first, before making minor modifications like this
15:44:06  *** Wormnest has joined #openttd
15:45:34  <andythenorth> So cab car as engine?  Or middle car as engine?  https://www.apt-p.com/APTConfigurations.htm
15:45:35  <andythenorth> I have code for either
15:46:32  <nielsm> you mean which actually provides power?
15:46:54  <nielsm> is it modular or single-part articulated?
15:47:46  *** sla_ro|master has joined #openttd
15:47:57  <andythenorth> it's built as cabs + middle cars
15:49:36  <andythenorth> http://bundles.openttdcoop.org/iron-horse/nightlies/v6957-3007/docs/html/trains.html#helm_wind_cab
15:50:54  <nielsm> zero-capacity power cars in the middle are more "interesting" perhaps
15:51:04  <nielsm> question is whether any AI would be able to build the train
15:51:18  <andythenorth> hmm I hadn't thought about the capacity :D
15:52:02  <nielsm> I'd say just be traditional with zero-capcity powered driving cars at each end and unpowered cars along the entire rest of the train
15:52:39  <andythenorth> yeah
15:53:20  <andythenorth> hmm, less ugly train I could draw https://c1.staticflickr.com/4/3734/10926416136_d6cb59dd9c_b.jpg
15:53:23  <andythenorth> never got built IRL
15:53:37  <Samu> are these names better https://imgur.com/uMLQvtf
15:55:33  <nielsm> maybe the later versions of the train could also get limited and then full capacity in the cabs, to reflect minituarization or whatever you'd call it
15:56:21  <andythenorth> yes
15:56:28  * andythenorth wondering if 2 generations are enough
15:56:36  <andythenorth> horse generations are usually 30-40 years
15:56:45  <andythenorth> but these trains are 20 years for some reason I forget :P
15:56:55  <andythenorth> less drawing = win
15:57:46  <nielsm> maybe a tick-tock upgrade path?
15:58:02  <nielsm> where a better passenger car becomes available halfway before the major cab upgrade
15:58:27  <andythenorth> higher capacity?
15:58:35  <andythenorth> TGV-Duplex style
15:59:07  <nielsm> for example
16:02:40  <nielsm> imo you should make one of the later generation have all-axles-powered configuration :)
16:02:58  <nielsm> less power on each individual unit, but even power per weight regardless of length
16:03:14  <andythenorth> yeah they're coded that way on all
16:03:27  <andythenorth> it's unrealistic for the early one, but eh
16:03:33  <andythenorth> game world
16:06:33  <andythenorth> hmm
16:06:48  <andythenorth> the middle cars aren't necessarily introduced at same time as cabs :x
16:07:15  <nielsm> right I think we talked about that earlier too :/
16:08:42  <Samu> making the river walk to the sea means i need to touch mode code
16:09:06  <Samu> unrelated to mine
16:09:11  <nielsm> Samu, not necessarily make it always go to the sea
16:09:22  *** Thedarkb-T60 has quit IRC
16:09:30  <nielsm> not don't kill the previous behavior
16:09:36  <andythenorth> the point of the comments is that rivers already go to the see
16:09:39  <andythenorth> sea *
16:09:43  <andythenorth> so let's not regress
16:11:18  <nielsm> https://user-images.githubusercontent.com/1780327/51436871-eab15480-1c8c-11e9-85cf-4c0340265c56.png
16:11:22  <nielsm> that one's a good example
16:11:25  <Samu> on more difficult terrain, rivers simply end up into lakes or are not created at all
16:11:40  <nielsm> the river just sort of flows around the top of the hill
16:11:49  <nielsm> that looks really dumb
16:12:54  <andythenorth> I didn't post the most pathological
16:13:04  <nielsm> but really, river generation is really bad in general imo
16:13:09  <nielsm> they should be wider
16:13:13  <nielsm> they should erode terrain
16:13:14  <andythenorth> https://dev.openttdcoop.org/attachments/download/9242/Unnamed,%2001-01-2000%233.png
16:13:27  <andythenorth> ^ just looks broken
16:13:37  <andythenorth> rivers don't hang around at the top of hills :)
16:13:46  <andythenorth> nor do they fail to connect when 1 tile apart
16:14:21  <Samu> that last screenshot, it's 4 rivers
16:14:59  <Samu> hmm not sure what I can do
16:15:28  <Samu> it's part due to pathfinder and part my demands
16:15:36  <andythenorth> I didn't post it to github because it's too pathological
16:15:56  <andythenorth> the ridge on the hill makes it very hard for the pathfinder
16:16:14  <nielsm> instead of trying to find a path that allows locks to be built everywhere
16:16:16  <andythenorth> but I was able to trigger many many others in testing that *should* have gone to sea :)
16:16:37  <nielsm> keep pathfinding as it is, but try to place some additional river tiles to make locks possible
16:16:41  <andythenorth> what we don't know is whether rivers should be able to terradorm
16:16:45  <andythenorth> terraform *
16:16:54  <Samu> rivers don't terraform
16:17:16  <Samu> I wouldn't even know how to do  that
16:17:23  <Samu> it would be during pathfinding
16:17:32  <andythenorth> hmm, how handy, I already drew an unused sprite for this high speed train :P
16:19:47  <Samu> someone with more experience is needed if pathfinding while terraforming occurs
16:21:20  <Samu> i don't know if that is possible nielsm
16:21:42  <Samu> i already place additional river tiles to make locks possible, but that alone isn't enough
16:22:13  <Samu> i had to ensure there's a path to the next lock nearby
16:22:32  <Samu> terrain slopes are heavily conditioned
16:22:51  <Samu> and that's for 90 degrees allowed
16:23:09  <Samu> haven't done anything if 90 degrees are forbidden, it would be even worse
16:24:00  *** Thedarkb-T60 has joined #openttd
16:26:39  <Samu> i had fat rivers in an old version of the patch, sec let me find
16:28:57  <Samu> http://i.imgur.com/jTxft7U.png
16:29:01  <Samu> they looked like that
16:32:53  <nielsm> I kind of like that
16:33:04  <Samu> https://dev.openttdcoop.org/attachments/download/9242/Unnamed,%2001-01-2000%233.png rivers aren't connected because they're 2 rivers, I only ensure the connection per river
16:33:26  <andythenorth> yes
16:33:35  <andythenorth> but it only looks weird because neither reached the sea
16:33:42  <andythenorth> but it's a hard case on that hill
16:33:43  *** gelignite has joined #openttd
16:33:48  <nielsm> the entire production of rivers needs to be rethought imo
16:33:55  <andythenorth> +1
16:34:36  <nielsm> I usually just play with them off in landscape generation because they're just thin squiggles of annoyance littered around, not actually useful for anything
16:35:29  <Samu> locks make everything hard
16:35:58  <andythenorth> well locks should be 1 tile
16:36:09  <andythenorth> but nobody wants to commit to that, because savegames
16:36:12  <nielsm> are rivers produced via pathfinding?
16:36:12  <Samu> if they're 1 tile, my patch wouldn't even be needed
16:36:16  <Samu> yes
16:36:34  <nielsm> then make routes that produce lock-capable cost less
16:36:49  <nielsm> or rather, penalize downward flow that doesn't allow lock placement
16:37:23  <nielsm> so you'll still get a path from spring to sea, just with preference for paths that allow locks
16:37:27  <Samu> well, ok, must understand then what exactly how the pathfinder works in this case
16:37:47  *** Gustavo6056 has joined #openttd
16:37:59  <Samu> i dont think it pathfinds from start to end
16:38:10  <Samu> it pathfinds from a slope to the next slope flowing down
16:38:33  <Samu> and continues to do so, until it reaches the sea or gets stuck, then transforms the last part into a lake
16:38:40  *** Gustavo6046 has quit IRC
16:38:40  *** Gustavo6056 is now known as Gustavo6046
16:39:01  <Samu> it's not only 1 pathfind
16:39:02  <Samu> call
16:39:19  <nielsm> oh well
16:39:44  <nielsm> that's just more material in favor of scrapping current river generation completely
16:40:46  <andythenorth> there was a big rainfall rivers patch
16:40:50  <andythenorth> but eh
16:40:56  <andythenorth> what I am curious about
16:41:01  <andythenorth> is why we start at springs and pathfind down
16:41:09  <andythenorth> why don't we start at coast, and pathfind up>?
16:41:11  <Samu> but maybe turning the pathfind into 1 entire call would be a good idea
16:41:39  <andythenorth> we don't need to go between two points
16:41:40  <nielsm> andythenorth I think it only becomes harder to produce a river in reverse
16:41:46  <andythenorth> why? o_O
16:41:52  <andythenorth> it's just a turning algorithm
16:43:01  <andythenorth> don't need a fixed destination
16:43:04  *** dvim has quit IRC
16:43:16  <andythenorth> just dice roll for a turn, weight by surrounding tile heights
16:44:06  <andythenorth> also, above height level 1, occasionally lower the next tile if ascending
16:44:21  <nielsm> by pathfinding upwards you somewhat lose the ability for multiple rivers to merge, for example
16:44:42  <nielsm> which is imo something you really want to have
16:44:43  <andythenorth> nah, just cache a tile as a fork point
16:44:46  <andythenorth> and then come back to it
16:45:14  <andythenorth> a river is just a tree
16:45:28  <andythenorth> easier to start from top node and work down
16:45:35  <andythenorth> cache the nodes
16:45:44  <andythenorth> iterate each branhc
16:45:46  <andythenorth> branch *
16:47:11  <nielsm> I think you've signed up for this now :P
16:47:35  <andythenorth> I am not a good enough programmer :P
16:48:05  <andythenorth> the current rivers got done because Rubidium crudely repurposed existing pathfinder
16:48:15  <andythenorth> ruthlessly :)
16:48:30  <andythenorth> all the other attempts I have seen are boil-the-ocean reality simulators
16:48:37  <andythenorth> faffing about with flow and gravity and stuff
16:49:00  <andythenorth> but we just want to generate a tree, and render it into tiles
16:52:42  <andythenorth> oh :) https://en.wikipedia.org/wiki/Strahler_number#River_networks
16:53:03  <andythenorth> https://en.wikipedia.org/wiki/Stream_order
16:55:20  <andythenorth> now we just need someone who likes trees :P
16:55:49  <nielsm> to make a river you need a tree, eh
16:56:06  <nielsm> sounds like a beaver thing
16:57:12  <andythenorth> :P http://www.thecodelesscode.com/case/233
17:06:52  <Samu> RIVER_HASH_SIZE
17:06:56  <Samu> what is a hash :(
17:07:12  <Samu> static const uint RIVER_HASH_SIZE = 8; ///< The number of bits the hash for river finding should have.
17:07:21  <nielsm> a sum of something
17:07:39  <nielsm> which is usually difficult to bring back to the original input value to the hash function
17:07:59  <nielsm> but is useful to quickly compare whether two large inputs are likely to be identical
17:08:44  <nielsm> hash tables and cryptographic hashes
17:08:51  <nielsm> are the most common
17:13:46  <Samu> @calc 1 << 8
17:13:46  <DorpsGek> Samu: Error: Something in there wasn't a valid number.
17:13:53  <nielsm> 256
17:14:13  *** Thedarkb-T60 has quit IRC
17:14:18  <Samu> :) ty
17:14:25  <nielsm> 1 2 4 8 16 32 64 128 256  <- 8 doublings
17:14:34  *** Thedarkb-T60 has joined #openttd
17:14:45  <planetmaker> @calc 1>>8
17:14:45  <DorpsGek> planetmaker: Error: Something in there wasn't a valid number.
17:14:51  <planetmaker> hm
17:14:58  <planetmaker> @calc shift(1,8)
17:14:58  <DorpsGek> planetmaker: Error: 'shift' is not a defined function.
17:15:08  <planetmaker> @calc 2**8
17:15:08  <DorpsGek> planetmaker: 256
17:15:22  <nielsm> learning the series of powers of 2 is a good value in software dev :)
17:15:41  <planetmaker> it's something which simply sticks after a while, I think
17:15:45  <nielsm> yea
17:16:11  <planetmaker> I guess the values up to 2**16 are pretty commen :)
17:16:14  <planetmaker> *common
17:16:18  <nielsm> well, I have 1-17, 20, and 24 remembered, the rest are usually not that useful
17:16:28  <planetmaker> yepp
17:17:09  <planetmaker> though not sure I get every digit right for things > 2**12
17:17:14  <Samu> 1 << RIVER_HASH_SIZE becomes num_buckets inside AyStar
17:17:25  <Samu> 256 buckets
17:18:50  <nielsm> yea it's a hash table
17:22:05  <Samu> seems to be using NPF code
17:22:11  <Samu> not yapf
17:22:59  <planetmaker> @calc 3 + sqrt(-1)
17:22:59  <DorpsGek> planetmaker: 3+i
17:23:03  <planetmaker> :)
17:24:07  <planetmaker> @convert 1 mph to kph
17:24:07  <DorpsGek> planetmaker: 1.609344
17:25:06  <Samu> @convert 1 kph to 1 kp-ish h
17:25:06  <DorpsGek> Samu: Error: 1kp-ishh is not a valid unit.
17:25:11  <Samu> kek
17:27:54  <Samu> static inline uint TileHash(uint x, uint y)
17:27:54  <Samu> {
17:27:55  <Samu> 	uint hash = x >> 4;
17:27:55  <Samu> 	hash ^= x >> 6;
17:27:55  <Samu> 	hash ^= y >> 4;
17:27:55  <Samu> 	hash -= y >> 6;
17:27:55  <Samu> 	return hash;
17:27:57  <Samu> }
17:28:05  <Samu> too complex for me
17:28:37  <nielsm> it's not important what the hash is
17:28:48  <nielsm> it's just used to speed up certain lookups
17:29:08  <nielsm> it shouldn't have any bearing on the actual algorithm
17:38:49  *** dvim has joined #openttd
17:40:49  *** Thedarkb1-T60 has joined #openttd
17:42:56  <DorpsGek_II> [OpenTTD/website] LordAro requested changes for pull request #27: Update: README.md is now more verbose and even contains a FAQ https://git.io/fhuyh
17:46:34  *** Thedarkb-T60 has quit IRC
17:59:03  <DorpsGek_II> [OpenTTD/OpenTTD] ldpl commented on pull request #6965: Add: Four alternative town cargo generation methods https://git.io/fhuSA
18:05:22  <Samu> xtree
18:06:13  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #6965: Add: Four alternative town cargo generation methods https://git.io/fhu9s
18:06:24  <Samu> C:\Program Files (x86)\Microsoft Visual Studio17\Community\VC\Tools\MSVC.16.27023\include\xtree
18:06:31  *** Borg has joined #openttd
18:06:32  <Samu> doesn't look like it's part of openttd
18:06:37  <Borg> toskete!! I mean.. hi
18:06:52  <Borg> what 'unmasked' means in GRF debug window for callbacks?
18:07:06  <Borg> I changed some tiles via act0.. and now.. animations seems to not working at all
18:08:27  <DorpsGek_II> [OpenTTD/OpenTTD] ldpl commented on pull request #6965: Add: Four alternative town cargo generation methods https://git.io/fhu9W
18:13:56  <DorpsGek_II> [OpenTTD/OpenTTD] ldpl commented on pull request #6965: Add: Four alternative town cargo generation methods https://git.io/fhu9o
18:16:59  <Borg> ! *00 09 03 01 1D 08 1D 09 1D 0A 0A 08
18:17:12  <Borg> property 11 not touched...
18:17:46  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #6872: Add: Houses can accept up to 16 different cargo types via NewGRF https://git.io/fhu97
18:18:02  <Samu> what's a breadth first search
18:18:08  <Samu> another pathfinder?
18:18:12  <LordAro> Samu: sort of
18:18:39  <Samu> tile finder it seems, and it's marking tiles with xtree
18:18:43  <LordAro> https://en.wikipedia.org/wiki/Breadth-first_search reminder that OTTD does nothing especially original, in terms of algorithms
18:18:46  <Samu> what is xtree
18:19:06  <LordAro> i have no idea
18:19:17  <LordAro> other than it being part of VS, given the path you posted
18:19:22  <LordAro> i'm sure google will tell you
18:25:29  <Samu> 	class set
18:25:29  <Samu> 		: public _Tree<_Tset_traits<_Kty, _Pr, _Alloc, false>>
18:25:29  <Samu> 	{	// ordered red-black tree of key values, unique keys
18:25:39  <Samu> now in english plz
18:25:49  <Samu> hehe
18:27:28  <Samu> std::set<TileIndex> marks;
18:28:16  <nielsm> LordAro: I don't see the alignment things you're pointing to, they look aligned to me?
18:28:19  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #6926: Change: Allow dock to be constructed in more locations https://git.io/fhuHI
18:28:36  <DorpsGek_II> [OpenTTD/website] TrueBrain updated pull request #27: Update: README.md is now more verbose and even contains a FAQ https://git.io/fhuqd
18:29:28  <TrueBrain> now it is up to andythenorth :P
18:29:56  <TrueBrain> btw, DigitalOcean mailed back .. basically, they response: look here, a ticket, vote for it if you want it. Owh, and here, a guide to do exactly what I said in my question that I did not want to do that
18:30:02  <TrueBrain> so .. not super useful reply :(
18:30:12  <LordAro> sounds about right
18:30:16  <TrueBrain> so I am doing it the most ugliest way I can think of, the IPv4/IPv6 bla
18:30:31  *** Zino_ has joined #openttd
18:31:25  <TrueBrain> at least they did read my questions; so at least it is a good support :) Just the answers .. I found those too by googling :) Was hoping a bit more in-depth, I guess
18:31:25  <andythenorth> oof
18:31:43  * andythenorth is going for drinks, no PRs for me :P
18:31:48  <andythenorth> soz
18:31:56  <LordAro> nielsm: https://i.imgur.com/c6pG9D4.png this is what github shows me (see the addition of the line)
18:31:57  *** Zino_ has quit IRC
18:31:59  <TrueBrain> no rush andythenorth
18:32:04  <LordAro> the review boxes make it hard to see
18:32:06  <TrueBrain> I wont fix the IPv6 stuff before next weekend anyway
18:32:19  <andythenorth> also... "unfinished" :P https://dev.openttdcoop.org/attachments/download/9247/high-speed-pax.png
18:34:03  *** glx has joined #openttd
18:34:04  *** ChanServ sets mode: +v glx
18:34:47  <nielsm> LordAro: https://0x0.st/sFiA.png
18:34:52  <TrueBrain> hmm .. Azure Pipelines broke .. "2019-01-20T18:34:35.2769753Z ##[error]Error: Moved Permanently" .. on GitHub release publish
18:34:54  <TrueBrain> that is just annoying
18:36:40  <LordAro> nielsm: how interesting, i shall blame GH
18:36:49  <LordAro> unless there's a hidden tab character in there or something silly
18:37:08  <nielsm> commit checker would have caught tabs
18:38:08  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #6928: Fix #5713: Use pathfinder to find closest ship depot https://git.io/fhuH0
18:38:52  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #6872: Add: Houses can accept up to 16 different cargo types via NewGRF https://git.io/fhuHg
18:40:48  <Samu> how come there are conflicts? :(
18:42:06  <Samu> no conflicts here, only there?
18:42:19  <Samu> must test again
18:43:58  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuqz
18:44:48  <DorpsGek_II> [OpenTTD/website] LordAro dismissed a review for pull request #27: Update: README.md is now more verbose and even contains a FAQ https://git.io/fhuyh
18:45:04  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh updated pull request #6872: Add: Houses can accept up to 16 different cargo types via NewGRF https://git.io/fNanB
18:45:35  <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain opened pull request #7075: Fix: [AzurePipelines] the repository OpenTTD-CF is renamed to CompileFarm https://git.io/fhuHX
18:45:47  <Samu> why don't you use svn style revision? :|
18:45:56  <Samu> svn still exists
18:46:35  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7075: Fix: [AzurePipelines] the repository OpenTTD-CF is renamed to CompileFarm https://git.io/fhuHD
18:46:57  <Samu> At revision: 24318
18:47:09  <TrueBrain> LordAro: well ... it seems the redirect is a bit weird. Uploads failed now on the old URL, but still worked for a bit of time
18:47:12  <TrueBrain> not sure how they do that :P
18:47:18  <TrueBrain> better safe than sorry :D
18:47:20  <LordAro> interesting
18:47:45  <TrueBrain> only a dismissal of the README? Not even an approve? :P
18:48:33  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7073: Feature: Generate lock ready rivers upon world generation https://git.io/fhuH9
18:51:39  <LordAro> TrueBrain: have some patience! :p
18:51:50  <TrueBrain> NO!
18:51:51  <TrueBrain> :D
18:52:00  <LordAro> how rude
18:52:05  <TrueBrain> :(
18:52:07  <TrueBrain> I know :(
18:52:14  <LordAro> :p
18:52:24  <TrueBrain> its cold outside
18:53:22  <DorpsGek_II> [OpenTTD/website] LordAro approved pull request #27: Update: README.md is now more verbose and even contains a FAQ https://git.io/fhuQe
18:53:26  <TrueBrain> :D \o/
18:53:29  <TrueBrain> tnx LordAro!
18:53:29  <DorpsGek_II> [OpenTTD/website] LordAro closed issue #26: Update README https://git.io/fhutH
18:53:30  <DorpsGek_II> [OpenTTD/website] LordAro merged pull request #27: Update: README.md is now more verbose and even contains a FAQ https://git.io/fhuqd
18:53:38  <nielsm> LordAro, types fixed for the house accepts :)
18:54:03  *** HerzogDeXtEr has quit IRC
18:54:08  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #6872: Add: Houses can accept up to 16 different cargo types via NewGRF https://git.io/fhu97
18:54:29  <LordAro> nielsm: still needs someone who actually knows GRFs to approve it, imo :>
18:54:40  <nielsm> meh!
18:54:52  <nielsm> !poke frosch123
18:54:52  <TrueBrain> so nobody?
18:54:53  <TrueBrain> :P
18:56:05  <planetmaker> I kinda feel the same. Though if you want I can become the "meh, I just approve it" guy :P
18:56:48  <LordAro> planetmaker: i thought that was my job
18:57:22  <nielsm> it's basically the same pattern as the industries 16 in/out patch which was already merged :)
18:57:37  <nielsm> except houses are simpler
18:57:43  <TrueBrain> "trust me, I know what I am doing"
18:57:46  <TrueBrain> :D
18:57:48  <nielsm> exactly!
18:58:38  <planetmaker> hehe. exactly that :P
18:59:40  <planetmaker> nielsm, does that work with town cargoes in all 16 slots?
19:00:08  * andythenorth wonders 
19:00:16  <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain merged pull request #7075: Fix: [AzurePipelines] the repository OpenTTD-CF is renamed to CompileFarm https://git.io/fhuHX
19:00:23  <planetmaker> *town effect cargoes
19:00:23  <andythenorth> when will 'vehicle ctrl-click counter var' patch be done? :P
19:00:54  <andythenorth> pretty simple, it's just a byte, incrementing each type player ctrl-clicks the vehicle
19:01:02  <andythenorth> rolls over to 0 after 16 clicks
19:01:06  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #6935: Prevent water construction where it could block nearby ships https://git.io/fhuQs
19:01:07  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro closed pull request #6935: Prevent water construction where it could block nearby ships https://git.io/fxno7
19:01:25  <andythenorth> vehicle_is_flipped is then mod(2) of ctrl-click-counter
19:02:17  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #6980: GDI engine for font glyph rendering as a replacement for FreeType https://git.io/fhuQc
19:02:45  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #6980: GDI engine for font glyph rendering as a replacement for FreeType https://git.io/fhuQW
19:03:09  <TrueBrain> hmm .. GitHub has a bit of issues to show of branches if they are already merged
19:03:12  <TrueBrain> I guess because of the rebasing
19:03:16  <TrueBrain> it still thinks it is new
19:03:17  <TrueBrain> annoying
19:03:27  <LordAro> mm
19:05:01  <Samu> oh no LordAro :|
19:05:20  <Samu> ok, what about limiting it only to industries?
19:06:07  <Samu> they can automatically spawn with a 0.00001% chance to dead-lock ships
19:06:12  <Samu> kek
19:06:34  <LordAro> yeah... not worth the effort, i'm afraid :p
19:06:38  <Samu> usually firs industries
19:06:48  <andythenorth> why don't we cache contiguous blocks of water?
19:06:52  <andythenorth> in a pool?
19:06:57  <LordAro> andythenorth said in the original issue that he was unable to trigger it, so it must be something like a frame perfect issue
19:07:10  <LordAro> andythenorth: a swimming pool
19:07:14  <andythenorth> :P
19:07:32  <andythenorth> it would be expensive to compute once, but trivial to update
19:07:43  <andythenorth> there are n contiguous blocks on any map
19:07:58  <andythenorth> construction: adds, merges, or changes blocks
19:08:11  <LordAro> nielsm: any chance of a review for #7022 ?
19:08:27  *** Borg has left #openttd
19:08:29  <andythenorth> if a block is contiguous a ship can route to anything on it
19:08:39  <andythenorth> a ship is attached to a block
19:09:09  <andythenorth> ships can only route to depots, docks, bouys etc in their block
19:09:14  <andythenorth> any others, effectively don't exist
19:09:27  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7025: Add #6887: Option to show zone inside local authority boundary of towns https://git.io/fhuQy
19:12:08  <andythenorth> hmm if some metadata about contiguous blocks could be exposed
19:12:17  <andythenorth> I could stop port industries building in tiny lakes :P
19:12:29  *** dvim has quit IRC
19:12:33  <andythenorth> the blocks could be calculated when placing the tiles
19:12:55  <andythenorth> also bye
19:12:56  *** andythenorth has quit IRC
19:16:20  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fhZ7Z
19:17:28  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #6935: Prevent water construction where it could block nearby ships https://git.io/fhu7f
19:22:08  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #7029: #6315 Rail fences in snow or desert https://git.io/fhu7t
19:23:41  <Samu> I don't understand how could there be conflicts
19:23:53  <Samu> just retested here, it built, didn't conflict
19:24:23  <LordAro> did you update your copy of master first?
19:24:42  <Samu> oh, no, ok let me try that
19:25:41  *** Gabda has joined #openttd
19:25:53  <Gabda> hello!
19:26:19  <Samu> ah, there are conflicts *now*, there weren't any before
19:26:22  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhu7c
19:26:29  <LordAro> Samu: almost as if master has been changing!
19:27:47  <DorpsGek_II> [OpenTTD/OpenTTD] Gabda87 commented on pull request #7025: Add #6887: Option to show zone inside local authority boundary of towns https://git.io/fhu78
19:28:31  <Gabda> Can I your opinion on PR 7025 vs 7047?
19:28:37  <Gabda> *ask your
19:28:53  <LordAro> @topic get 3
19:28:53  <DorpsGek> LordAro: Don't ask to ask, just ask
19:29:51  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh requested changes for pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhu7g
19:30:01  <nielsm> there :)
19:30:09  <LordAro> lol
19:30:45  <nielsm> going to study it in detail later
19:31:01  <Samu> ah
19:31:09  <Samu> v->dest_tile was changed
19:31:21  <Samu> v->SetDestTile
19:33:16  <Samu> the cache stuff was merged, i see
19:33:43  <Samu> im surprised there's only 1 conflict
19:33:55  <Samu> and it's not even yapf related
19:34:29  <TrueBrain> hmm ... how do I trigger a release based on the finishing of another release .. hmmmmmm
19:43:47  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7040: Fix: Round up deltas for smooth scrolling, so that target will eventually be reached https://git.io/fhu5O
19:43:59  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro merged pull request #7040: Fix: Round up deltas for smooth scrolling, so that target will eventually be reached https://git.io/fhnn8
19:44:14  * peter1138 is home.
19:44:38  <Eddi|zuHause> it's not what it looks like
19:46:12  <Samu> wow, i can finally run openttd in debug mode with 5000 ships
19:46:16  <Samu> nice improvement
19:46:31  <peter1138> You're welcome.
19:46:47  <peter1138> And only with YAPF, of course.
19:47:25  <Samu> i need to check whether caching while searching for depot is worth it now
19:48:09  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7072: Change: Add path cache for ships https://git.io/fhu58
19:51:07  <Samu> it returns the depot location, which was just pathfound
19:51:16  <Samu> that path could have been cached
19:51:33  <Samu> otherwise, shipchosetrack will pathfind again next tile
19:51:46  <Samu> pathfinder is called again, and this time it caches path
19:51:58  <Samu> so i guess the answer is yes
19:53:06  <Samu> what if I was only interested in knowing the depot location
19:53:09  <Samu> hmm
19:53:24  <Samu> caching the path to depot would screw the ship
19:53:34  <Samu> must think
20:02:24  <nielsm> https://github.com/OpenTTD/OpenTTD/pull/7022  --  turns out it's because the vbscript doesn't handle unix line endings, and my copy is checked out with unix line endings
20:02:49  <nielsm> if I unix2dos the file it generates zero changes to the file
20:03:14  *** Thedarkb1-T60 has quit IRC
20:03:15  <LordAro> heh
20:03:28  <LordAro> both scripts should probably be able to cope with both
20:04:14  <nielsm> or at least have the vbscript abort if it looks fishy
20:04:26  *** Gabda has left #openttd
20:05:09  <Eddi|zuHause> i'm pretty sure you can force git to use windows line endings
20:05:33  <DorpsGek_II> [OpenTTD/OpenTTD] ldpl commented on pull request #7025: Add #6887: Option to show zone inside local authority boundary of towns https://git.io/fhu5Q
20:05:48  <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhu55
20:06:04  <nielsm> Eddi|zuHause yea you can
20:06:07  <nielsm> I just haven't
20:06:11  <nielsm> but maybe I should :P
20:06:36  <glx> hmm nice noe
20:06:39  <glx> *one
20:09:17  <Samu> ok no need for me to do anything
20:10:28  <Samu> because sending a ship to depot is not just 1 simple step apparently
20:10:33  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7065: Change: Make ships stop and change direction slowly instead of instantly turning. https://git.io/fhude
20:10:46  <peter1138> Samu, what are you trying to achieve?
20:11:14  <Samu> was wondering whether caching was worth it for my patch
20:11:21  <Samu> find ship depot using pathfinders
20:11:26  <Samu> but it's not
20:11:56  <peter1138> Ah for finding the depot.
20:12:07  <peter1138> Once it's found it, it'll be cached.
20:12:09  <Samu> it goes through some DoCommand stuff
20:12:09  <peter1138> So yeah, probably not.
20:12:14  <glx> nielsm: probably because line 179
20:12:30  <Samu> DoCommand stuff involves multiplayer crap
20:12:48  <glx> I though it was a smart idea to skip temporary file usage
20:13:13  <peter1138> If we can do that for the language files (.tmp) then that'd be good ;)
20:13:22  <peter1138> I don't really like that 25MB "temporary" file sitting around.
20:13:34  <Samu> speaking of multiplayer, is this multiplayer sync ready? :p
20:13:41  <peter1138> Is what?
20:14:00  <Samu> caching while joining a multiplayer game won't cause desync?
20:14:13  <peter1138> It should be fine, the cache is actually saved in the savegame.
20:14:30  <Samu> nice, nice
20:15:11  <LordAro> peter1138: i wonder whether it was worth caching such a thing
20:15:14  <LordAro> saving*
20:15:15  <peter1138> I suppose technically the caches could've been saved on saving, but... uh... I went overboard :p
20:15:25  *** Progman has quit IRC
20:15:25  <peter1138> LordAro, well it's done now.
20:15:28  <LordAro> indeed
20:16:15  *** Progman has joined #openttd
20:18:26  <Samu> what was the other request from LordAro
20:20:37  <Samu> how do I mention the commit conflicting with mine?
20:20:40  <Samu> 81330b8d6edee68c38717462737fbfca6420701d ?
20:20:54  <peter1138> Why do you need to mention it?
20:20:55  <LordAro> don't mention it
20:20:56  <Samu> https://github.com/OpenTTD/OpenTTD/commit/81330b8d6edee68c38717462737fbfca6420701d
20:21:03  <Samu> ok
20:21:31  *** dvim has joined #openttd
20:22:37  <peter1138> What do you do with conflicts is you resolve them.
20:23:14  <nielsm> when you rebase you resolve the conflict as part of the rebase, there is no additional commit
20:23:25  <nielsm> when you merge you resolve the conflict as part of the merge commit
20:23:38  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #6931: Change: Prevent town growth from blocking ships https://git.io/fhud0
20:24:14  <LordAro> btw, i'm adding "needs-triage" labels to PRs that need discussion about whether they're a Good Feature
20:28:44  <peter1138> Should I revert the savegame change?
20:29:25  <LordAro> it is a quite a bit of code that perhaps isn't all that necessary
20:29:29  <LordAro> but ultimately up to you :p
20:31:01  <LordAro> nielsm: oh, i know what caused the alignment issue - github (or maybe firefox) uses ligatures in its monospace font...
20:31:52  <DorpsGek_II> [OpenTTD/OpenTTD] George-VB commented on issue #6908: Request: Persistent storage for vehicles https://git.io/fhudi
20:32:40  *** frosch123 has quit IRC
20:32:58  <Samu> so, a "blind search" is technically a first breadth search?
20:33:21  <LordAro> depends where you got "blind search" from :p
20:33:41  <Samu> breadth first search
20:33:57  <Samu> it searching all tiles for one that matches the condition "I am a depot"
20:34:03  <LordAro> but "breadth-first" search is, generally, a very suboptimal algorithm for finding the shortest path, though it is guaranteed to actually find the shortest path
20:35:15  <Samu>  that's what i made then
20:35:37  <Samu> pathfind suboptimal for the shortest path to a tile that is "I am a depot"
20:35:43  <Samu> english
20:37:35  <DorpsGek_II> [OpenTTD/OpenTTD] glx22 updated pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhsWa
20:38:23  <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhudF
20:38:26  <peter1138> Heh, it saves about 1 second of high load ;/
20:38:43  <peter1138> Oh well, at least we can saveload some std:: stuff if we ever need to.
20:38:44  <LordAro> peter1138: surprisingly high, how many ships?
20:38:49  <peter1138> 5000.
20:38:59  <LordAro> ah, lol
20:39:01  <peter1138> Yeah.
20:39:12  <LordAro> how long is the load in total?
20:39:38  <peter1138> Depends on the CPU :p
20:39:48  <peter1138> I'm a high-end CPU so..
20:40:04  <LordAro> so does the amount of time saved :p
20:40:10  <LordAro> how much for you?
20:41:01  <peter1138> Hmm?
20:41:45  <peter1138> Anyway.
20:42:03  <peter1138> Now we *have* the path cache, I'm tempted to partially revert the change that lead to #7062.
20:42:04  <LordAro> what percentage of the loading time is "saved" by saving the ship paths in the save
20:42:13  <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on issue #6908: Request: Persistent storage for vehicles https://git.io/fhudx
20:42:31  <LordAro> peter1138: good plan though
20:44:12  <Samu> I have a trash pentium cpu laptop from 2001 somewhere, can only run 32-bit versions of openttd on windows xp
20:44:29  <Samu> have it somewhere
20:44:35  <Samu> would have to find it
20:44:59  <LordAro> i suspect you're not going to win the "who has the oldest computer" argument amongst the people here :p
20:45:05  <LordAro> far too many greybeards
20:45:36  <glx> I still have working pentium 133MMX
20:46:26  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7061: Fix #6583: Rework Tileheight handling (patch by adf88) https://git.io/fhuFJ
20:46:40  <glx> well was working last time I used it ;)
20:48:36  <peter1138> I've got a Sun SparcStation in the loft.
20:48:40  <peter1138> Might be a 2.
20:49:00  <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #6872: Add: Houses can accept up to 16 different cargo types via NewGRF https://git.io/fhuFO
20:49:13  <peter1138> 40 Mhz MIPS CPU.
20:49:25  <glx> hehe
20:49:33  <LordAro> very nice
20:54:45  <peter1138> Also a P90 laptop. Maybe a 486 laptop, not sure.
20:55:18  <Samu> oh well then u guys can test it
20:55:35  *** Wormnest has quit IRC
20:56:37  <glx> I'm not even sure we can still compile a win9x version
20:57:28  *** D-HUND is now known as debdog
20:57:41  <LordAro> i imagine it's been a while since anyone tried
20:57:53  <LordAro> hardest part is probably finding a new enough compiler
20:57:56  <glx> old MSYS used to work for that
20:59:16  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #6928: Fix #5713: Use pathfinder to find closest ship depot https://git.io/fhZzm
20:59:44  <glx> maybe MSYS2 could too with the right defines to force older win32 api
21:00:25  <glx> I think it defaults to XP SP2
21:00:59  <glx> but that may not be enough
21:01:18  <glx> might require an arch flag too
21:01:20  <Samu> the conflict has been solved https://github.com/OpenTTD/OpenTTD/pull/6928/files#diff-fa888affc2f63e5d03e59e88c34d30f3R214
21:01:22  *** gelignite has quit IRC
21:01:43  <glx> for older cpus
21:01:45  <Samu> the request was not attended
21:02:30  <Samu> i dont know what to do with nielsmh code, it does some weird thing
21:03:59  <Samu> a "never seen before" behaviour
21:04:42  <Samu> it displays the order it's following and it can change it suddenly into another
21:04:50  <glx> if it's really needed, we'll wait for a PR doing it
21:04:52  <Samu> it's by intention but it looks weird
21:05:05  <glx> and I won't do it :)
21:06:14  *** chomwitt has quit IRC
21:06:37  <Samu> then when searching for a depot where the pathfinder doesn't return "is lost" in that for loop, it's ... it's... suboptimal imo
21:07:47  <Samu> pathfinds N times, until it gets a "path found" from the pathfinder
21:08:14  <Samu> it's what I had at first for OPF, but i changed it, and there were gains in cpu usage
21:08:16  <nielsm> I'm not saying my suggestion is good code or correct, I'm not very acquiainted with the pathfinding
21:08:43  <nielsm> tried making something to demonstrate my idea
21:09:05  <nielsm> didn't get it working quite as intended
21:09:16  <nielsm> uploaded it regardless
21:09:40  <nielsm> didn't want to put too much effort into something I see as a non-issue
21:11:17  <Samu> i tested your approach with the 5000 savegame
21:11:47  <Samu> there were multiple stalls for at least 3 seconds each
21:12:03  <Samu> some 4-5 noticeable repeated stalls
21:16:49  <Samu> my old approach for opf was similar, it would pathfind for "every" depot
21:17:15  <Samu> and return the one with the smallest cost
21:17:53  <Samu> would stall once only, but would stall for 14,6 seconds
21:18:08  <glx> pathfind for every depot is intensive
21:18:28  <Samu> my new approach, the breadth first search stalls for 4.7 seconds
21:19:10  <Samu> if my prs weren't closed, i could make it go further down to 3.4 seconds
21:19:27  <Samu> or 2.1 seconds with 90 degrees forbidden :|
21:24:48  *** sla_ro|master has quit IRC
21:26:44  *** dvim has quit IRC
21:31:39  <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #6928: Fix #5713: Use pathfinder to find closest ship depot https://git.io/fhuby
21:35:20  <Samu> It is useful however to prefer 'lockable' slopes where possible. If none are available, allowing flow downslope is better than rejecting.
21:35:59  <Samu> andythenorth, that basically means, don't use the patch :(
21:37:12  <Samu> the river is generated without any terraform
21:37:21  <Samu> don't know what to do
21:37:56  <Samu> adapt the terrain by terraforming would probably be the answer, but how
21:39:08  <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7072: Change: Add path cache for ships https://git.io/fhuNs
21:40:41  <Samu> i'd need a different set of terraformstate rules
21:40:52  <Samu> than the default one
21:41:55  <Samu> TerraLockForm
21:41:58  <Samu> kek
21:42:29  <Samu> and it would have to be done earlier
21:42:52  *** Gustavo6046 has quit IRC
21:42:56  <Samu> during the terrain formation, the first steps very much
21:43:01  <Samu> hmm
21:43:03  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7056: Change #6060: Allow drawing dropdown lists with scrollbars above and fix scrolling movement https://git.io/fhnon
21:43:08  <DorpsGek_II> [OpenTTD/OpenTTD] PeterN closed issue #6060: Allow drawing dropdown lists with scrollbars above the widgets https://git.io/fhuN4
21:44:16  <Samu> who's the TerraGenesis expert here?
21:44:36  <peter1138> richk
21:44:38  <Samu> that's where I need to touch
21:51:25  <Samu> Sprite
21:51:48  <Samu> there are terrain forms in form of Sprite template?
21:51:57  <Samu> or
21:52:33  <Samu> where's the original generator code?
21:52:55  <Samu> seems like pieces of puzzle that fit together
21:53:25  <Samu> maybe i can create a puzzle piece that is lock-friendly
21:53:41  <peter1138> Yes, there are.
21:54:27  <Samu> but then i'm touching the original generator and not terragenesis :(
21:54:33  <peter1138> Yup.
22:01:35  *** Gustavo6046 has joined #openttd
22:02:39  <Samu> base perlin noise generator
22:02:58  <Samu> it's a heightmap?
22:03:06  <Samu> funny how things work in detail
22:05:24  <Samu> static const double extrapolation_factors[] = { 3.3, 2.8, 2.3, 1.8 };
22:05:30  <Samu> what's a double?
22:05:34  <Samu> first time i see it
22:05:47  <nielsm> :o
22:06:42  <nielsm> it's a IEEE 754 double precision floating point number
22:09:36  <peter1138> Shame I lost all my map tweaks :/
22:10:50  <peter1138> Oh! I just found it.
22:15:22  *** Flygon has joined #openttd
22:40:35  <Samu> pow.asm not found :|
22:49:28  <nielsm> are you trying to step the debugger into the pow() function?
22:49:31  <nielsm> youi know what it does, right?
22:49:40  <nielsm> it calculates x to the power of y
22:49:48  <nielsm> exponentiation
22:50:16  <Samu> was following code sin.asm
22:50:26  <nielsm> why...
22:50:33  <nielsm> it's just standard math library functions
22:50:41  <nielsm> you can assume those do what it says on the tin
22:50:44  <LordAro> you know what trigonometry is, right,
22:50:52  <LordAro> *?
22:51:38  <Samu> ah, so that's it
22:52:16  <LordAro> for the love of god, read some documentation
22:52:18  <LordAro> or google
22:52:49  <LordAro> or at least notice when a function is external to ottd
22:55:08  <nielsm> the only time you should need to read assembly when working on ottd is if you need to reverse engineer something from the original game for some reason
22:55:35  <LordAro> (which was done entirely more than 15 years ago)
22:55:44  <nielsm> (except for the music drivers!!)
22:55:45  <LordAro> more or less
22:55:58  <LordAro> nielsm: yeah, and no one cared :p
22:56:02  <nielsm> :(
22:56:12  <LordAro> <3
23:03:29  *** Wolf01 has quit IRC
23:09:21  *** snail_UES_ has joined #openttd
23:16:47  <Samu> ah, interesting discovery
23:17:27  <Samu> apparently it does really pathfind from spring to end
23:17:57  <Samu> have to examine this better
23:19:43  <Samu> amdsecgs.asm this one appears too often
23:19:57  <Samu> __security_check_cookie
23:20:02  <Samu> let me google
23:20:08  <nielsm> it's a debug thing
23:20:12  <nielsm> in visual c++
23:20:30  <nielsm> that checks whether there have been buffer overruns and such
23:22:44  *** Thedarkb1-T60 has joined #openttd
23:26:00  <Samu> return DistanceManhattan(spring, begin) > _settings_game.game_creation.min_river_length;
23:26:10  <Samu> distancemanhattan for determining river length?
23:26:41  <Samu> that's not right, imo
23:27:20  <Samu> should be number of tiles turned into river
23:27:24  <Samu> but.. k
23:30:12  <Samu> this is checked before pathfinder is called, hmm so it's a compromise
23:36:03  <Samu> FindSpring code is interesting, its not any random location where it decides to start a river
23:36:26  <Samu> must obey some tile heights of nearby tiles
23:47:09  <Samu> ah no, it doesn't pathfind from start to finish, it pathfinds in multiple segments
23:47:19  <Samu> I was right the first time
23:53:29  <Samu> on this river
23:53:32  <Samu> https://imgur.com/Ye07W4x
23:53:40  <Samu> the pathfinder is invoked 5 times
23:53:59  <Samu> and in 3 cases is to pathfind to the tile next door
23:54:02  <Samu> kek
23:55:52  <Samu> actually 2 case
23:55:54  <Samu> s
23:57:28  <Samu> pathfinds 4 times, not 5
23:57:40  <Samu> wait, 5
23:57:42  <Samu> it's 5
23:59:52  <peter1138> Okay, so I have a translations.vbs now

Powered by YARRSTE version: svn-trunk