IRC logs for #openttd on OFTC at 2020-03-31
⏴ go to previous day
00:02:02 *** Arveen2 has joined #openttd
00:04:24 *** Xaroth65 has joined #openttd
00:15:01 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro opened pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5eU
02:30:06 <Eddi|zuHause> i wanted to complain in the german forum about summer time setting not working, and found a thread from 8 years ago, complaining about the same thing – by me
02:33:27 <FLHerne> 'summer time setting' ?
02:35:17 <Eddi|zuHause> you know, the opposite of winter time
02:37:04 <FLHerne> Ok, but since when has OTTD cared about that at all? (or why should it?)
02:37:29 <Eddi|zuHause> openttd doesn't, but the forum does (post times and stuff)
02:46:57 <dwfreed> Eddi|zuHause: the forum probably hasn't been updated in 8+ years
02:47:53 <Eddi|zuHause> dwfreed: i think the opposite is true, some update reset the changes that were made after the last complaint
06:59:42 *** ChanServ sets mode: +v tokai
07:30:29 *** yur3shmukcik[m] has quit IRC
07:30:45 *** KenjiE20 has joined #openttd
07:30:45 *** ZirconiumX has joined #openttd
07:30:45 *** lastmikoi has joined #openttd
07:30:45 *** gooodger has joined #openttd
07:30:45 *** natalie[m] has joined #openttd
07:30:45 *** joey[m] has joined #openttd
07:30:45 *** osvaldo[m] has joined #openttd
07:30:45 *** elliot[m] has joined #openttd
07:30:45 *** natmac[m] has joined #openttd
07:30:45 *** grag[m] has joined #openttd
07:30:45 *** yur3shmukcik[m] has joined #openttd
07:30:45 *** freu[m] has joined #openttd
07:30:45 *** pina[m] has joined #openttd
07:30:45 *** NGC3982 has joined #openttd
07:30:45 *** Decapita1 has joined #openttd
07:30:45 *** dihedral has joined #openttd
07:30:45 *** tyteen4a03 has joined #openttd
07:30:45 *** luaduck has joined #openttd
07:30:45 *** Alkel_U3 has joined #openttd
07:30:45 *** michi_cc has joined #openttd
07:30:45 *** liquid.oftc.net sets mode: +v michi_cc
07:31:45 *** ChanServ sets mode: +o orudge
07:58:43 *** sla_ro|master has joined #openttd
08:16:13 *** andythenorth has joined #openttd
09:04:11 *** D-HUND is now known as debdog
09:46:44 *** arikover has joined #openttd
09:53:48 <Samu> oh, i'm gonna test #8051
09:59:18 <Samu> I have a feeling it's gonna break start_date
10:12:58 <tuvok> hm i have big problems with traffic lights and trains
10:13:53 <tuvok> gives a manual with that?
10:24:36 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5Cf
10:28:17 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5CT
10:31:55 <Samu> I mean, random deviation for start_date, has to happen on constructor
10:32:52 <Samu> you set a random deviation for 182 days in main menu, then you start a new game, and the deviation didn't occur
10:33:08 <Samu> it will occur when the script is starting, which is pointless by then, it has already waited 182 days
10:34:04 <LordAro> put that in the issue please
10:45:53 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5CV
10:51:26 <Samu> my bad, you don't really set a random deviation for 182 days
10:51:36 <Samu> you set a start_date of 182 days, what I meant to say
11:17:04 <tuvok> hm i think iam to stupid for traffic signals and trains
11:26:32 <LordAro> tuvok: you can ignore all the presignal stuff, just use PBS signals
11:26:43 <LordAro> presignals are for advanced users only
11:44:04 <Eddi|zuHause> we should just remove presignals
12:02:00 *** gelignite has joined #openttd
12:47:41 *** andythenorth has left #openttd
13:27:07 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5Rb
13:28:52 *** msmarquim has joined #openttd
13:30:51 <Samu> this is so confusing, even i get confused
13:50:16 *** arikover has joined #openttd
13:57:16 *** andythenorth has joined #openttd
14:22:58 <Samu> this just in! I think i found a bug with random deviation itself, the interval
14:23:38 <Samu> if value is 30, and random deviation is 10, the resulting value is inside [20, 39]
14:23:44 *** andythenorth has left #openttd
14:24:05 <Samu> shouldn't it be [20, 40]?
14:28:27 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5ur
14:29:50 <LordAro> Samu: it's not uncommon for ranges to have an exclusive upper bound
14:30:15 <LordAro> i'd probably say that it should be documented, but probably not worth changing
14:31:06 <Samu> random range had a similar bug
14:31:28 <LordAro> should probably be consistent
14:32:14 <Samu> btw, have you seen #7661 - it reworks start_date to be more sane
14:33:34 <Samu> start_date is usually the problem in all this
14:41:06 <Samu> "When the game is started, the value for the setting will be randomized between user_specified_value - random_deviation and user_specified_value + random_deviation."
14:47:21 <Eddi|zuHause> if ranges have an open and a closed boundary, you can join and split them without fancy +1 maths
14:47:31 <Eddi|zuHause> but random deviation should probably be symmetrical
14:50:02 <DorpsGek_III_> [OpenTTD/OpenTTD] nielsmh commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5zb
14:58:04 <Samu> commit message looks good? Fix: [Script] Random deviation upper bound range is inclusive
15:03:50 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick opened pull request #8052: Fix: [Script] Random deviation upper bound range should be inclusive https://git.io/Jv5gD
15:39:35 *** supermop_Home has joined #openttd
16:07:00 <Eddi|zuHause> <Samu> https://docs.openttd.org/ai-api/classAIInfo.html says "random_deviation If this property has a nonzero value, then the actual value of the setting in game will be user_configured_value + random(-random_deviation, random_deviation)." <- that is not a good description as random() usully has an open upper bound
16:08:45 <Eddi|zuHause> so it technically describes the old behaviour, but since we determined to change the behaviour, this description must also change
16:10:38 <LordAro> depends if random() is referring to anything specific
16:10:45 <LordAro> but should probably be clearer anyway
16:13:13 <Samu> ok, english is not my forte
16:14:17 <LordAro> that's a pretty funny sentence really
16:16:00 <Eddi|zuHause> it's a really bad description, because it uses technical language which it never specified
16:16:59 <Samu> so between 20 and 40, inclusive
16:18:03 <Samu> how can I make the descriptionit clearer
16:18:23 <Eddi|zuHause> probably something like "... will be randomized in the range [user_configured_value - random_deviation, user_configured_value + random_deviation] (inclusive)"
16:28:24 *** gelignite has joined #openttd
16:33:26 <Samu> while we're at it, this document is a bit outdated
16:33:42 <Samu> GetAPIVersion only has 1.3 as max
16:36:03 <Samu> custom difficulty no longer exists
16:44:22 <Samu> I have no means to test doxygen, but I think it's going to show ok
16:47:39 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick updated pull request #8052: Fix: [Script] Random deviation upper bound range should be inclusive https://git.io/Jv5gD
16:52:21 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick updated pull request #8052: Fix: [Script] Random deviation upper bound range should be inclusive https://git.io/Jv5gD
16:53:29 <Samu> how do i check doxygen ? wanted to see how it's displayed
16:54:19 <nielsm> you install doxygen on your own machine and run it
16:54:46 <Samu> oh really? :( no website to look for?
17:05:24 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro approved pull request #8052: Fix: [Script] Random deviation upper bound range should be inclusive https://git.io/Jv5oz
17:14:54 <Samu> hi nielsm how's ending_year going?
17:15:16 <nielsm> not done anything more on it
17:15:38 <nielsm> it's not a critical bug by any means
17:16:15 <Samu> there is a possibility of an overflow, if you wait 800k+ years
17:17:31 <Samu> the _date doesn't get capped to year 5 million, it goes over
17:18:57 <nielsm> oh... if the player is insane and sets the end year to the same as MAX_YEAR
17:19:36 <nielsm> can we just call that an easter egg
17:19:52 <Samu> i tested what happened when it overflows
17:19:59 <nielsm> or otherwise promote it to a feature
17:20:07 <Samu> it's gonna end crashing openttd with weird errors
17:20:47 <Samu> i have a save, if you want to
17:20:59 <nielsm> it's clear it will misbehave
17:21:44 <nielsm> I just think you've already caused me to waste more time (and wasted more of your own time) on a bug nobody will ever experience in practice
17:22:20 <Samu> Unnamed, 5879610-07-11.sav the last day before the overflow :p
17:22:27 <Samu> i like to test this stuff hehe
17:51:39 *** tuvok-_ has joined #openttd
18:25:15 <Zuu> I know this may be controversial, but do you know any other open source game than OpenTTD that is perhaps a bit more general that can be played online multiplayer?
18:26:00 <LordAro> define "a bit more general"
18:26:58 <Zuu> These days I've ended up being in the board of a local game club which mostly meet and play games mostly as an excuse to be social and meet and due to c* we need to try to host something online.
18:27:56 <Zuu> It can be yazzy or some well known board game or some variant of that. That plays for 20-40 minutes or so.
18:29:00 <LordAro> does OpenRCT2 have multiplayer yet?
18:29:28 <Zuu> Thanks for the suggestion/link.
18:31:00 <Zuu> Oh I've completely left out of OpenRCT/FreeRCT.
18:31:25 <LordAro> FreeRCT died several years ago
18:31:41 <Zuu> Yeah. At least I've implemented fences for it :D
18:34:33 <arikover> Zuu: it is available on the Ubuntu Software Center, and I think a multilpayer game can be set online... not sure.
18:35:02 <Zuu> arikover: thanks for the suggestion. (fun fact: I've wored on a computer implementation of settlers of catan in Visual Basic which had hot seat multiplayer)
18:41:04 <arikover> Zuu: hot seat multiplayer reminds me of HM&M2... fun times.
18:48:03 <arikover> Zuu: so in the meantime I installed Pioneers from Ubuntu bananas, and there is a Pioneers Server thing. Of course the game is turn-based, so there is a lot of waiting, but it works.
18:48:45 <Zuu> That might be alright as that allows for the others to talk while waiting for the person deciding what to do.
18:49:36 <arikover> Zuu: ... the main problem being the maximum number of players: 8.
18:50:47 <Zuu> I know all comes down to that we should play OpenTTD :-p
18:58:11 <LordAro> Zuu: well, 1.10.0 release tomorrow :)
18:58:51 <Wolf01> They said no april fools, so release now :>
18:59:23 <LordAro> i'm expecting it to be tagged tonight, tbf
18:59:50 <Samu> make cluelessplus build air
19:00:47 <Zuu> Great that the tradition is kept. Swedish road administration used to release model updates on or around 1 april but this year the schedule has been changed.
19:02:36 <Zuu> Clueless Plus can build air if you toggle a swich. The switch is there because it is not as good to not go bankrupt if you enable it. :-p
19:03:19 <Samu> I remember I had a question for you, but can't remember which
19:03:36 <Samu> now that you're here, I don't remember, of course
19:08:08 <Samu> i think it was about the road pathfinder? let me see
19:13:39 <TrueBrain> run while you can Zuu , run! :P
19:13:52 <Zuu> TrueBrain: Great advice :-)
19:14:53 <TrueBrain> now you only have yourself to blame :)
19:15:11 <Samu> AIRoad.CanBuildConnectedRoadPartsHere can return -1
19:15:37 <Samu> line needs to be changed to AIRoad.CanBuildConnectedRoadPartsHere(cur_node, path.GetParent().GetTile(), next_tile) > 1
19:15:46 <Samu> can you change it? it says you're project manager
19:16:37 <Samu> then i guess superlib would also needs fixed
19:16:50 *** andythenorth has joined #openttd
19:17:03 <TrueBrain> owh no, an andythenorth
19:17:51 <TrueBrain> right, lets add some more regression tests ..
19:18:44 <andythenorth> oof I am accidentally going to work the rest of the day :P
19:18:58 <Samu> CanBuildConnectedRoadParts can in fact return -1, 0, 1 and 2
19:19:03 <andythenorth> New Zealand call at 20.30 UK time, and some website I am building also
19:19:10 <andythenorth> are we releasing anything? D
19:19:24 <TrueBrain> is it the end of the month?
19:20:27 <Zuu> Samu: I think I inherited the common road path finders after probably TrueBrain at some point. But I've not been active coding anything OpenTTD related for years.
19:20:41 <andythenorth> TrueBrain nearly!
19:20:50 *** tokai|noir has joined #openttd
19:20:50 *** ChanServ sets mode: +v tokai|noir
19:21:10 <TrueBrain> yes, blame the author, such an easy escape :P :P :P :P :P
19:21:21 <Zuu> I probably still have the credentials so I can push an update if it is needed.
19:21:27 <TrueBrain> while I am looking at hacks to get around BaNaNaS doing uniqueid detection :P
19:21:37 <Zuu> TrueBrain: Think this time you took the blame on you.
19:22:16 <Zuu> Ah that uniqueid and script name soup :-)
19:23:46 *** Progman has joined #openttd
19:23:54 <Samu> there has been a recent fix regarding that function on OpenTTD side, so it's kind of relevant if it was also fixed in the library
19:24:29 <Samu> it still works, doesn't appear to be breaking AIs, but, it's accepting -1 as true
19:26:19 <Zuu> Hmm pushing to bananas is probably the easy part. Figuring out how to auth with devzone and how to use hg is probably more complicated.
19:27:22 <Zuu> TrueBrain: according to devzone, Yexo is the original author and not you.
19:27:34 <TrueBrain> pretty sure that is sort-of a lie :P
19:27:37 <TrueBrain> but I am fine with it :D
19:27:49 <DorpsGek_III_> [OpenTTD/OpenTTD] andythenorth commented on issue #7644: Mysteriously poor performance on macOS https://git.io/fjii3
19:28:07 <Zuu> Hm.. although the first commit says "import verison 2" so there is still a verison 1 not in the repo that could be yours. :-p
19:30:11 <TrueBrain> that is how things went in these days :)
19:31:48 <Samu> due to a bug in openttd side, -1 was all that it was returning when loading a game with a different map size than the current
19:31:59 <Zuu> Is devzone still the source of thruth and I should figure out how to auth with it or are things moving over to github these days?
19:32:29 <TrueBrain> I do not think anyone currently is actively planning to migrate projects like that to GitHub
19:32:37 <TrueBrain> but if you like, I would welcome if you do :)
19:32:59 <Zuu> Or if you Samu have a ssh-key setup I can give you write access to the repo so you can patch the code and then I can figure out how to deploy to bananas.
19:33:17 <Samu> i dont have an account on openttdcoop
19:34:36 <Zuu> You have a private and a public key. The public key needs to be authorized on opentttdcoop. Today such things usually is done by adding it to your user account, but this is an older system where I think an admin needs to authorize your public key.
19:35:27 <Zuu> But it could also be that I'm authorized and I just have used a specific ssh key for openttd-devzone and have broken my .ssh/config.
19:36:39 <TrueBrain> I said it before, and I say it again: tusd is implemented poorly ... the "post-create" call can be done before the "pre-create" call is
19:36:41 <TrueBrain> you read that right
19:36:59 <TrueBrain> euh .. sorry: "post-finish" can be called before "post-create"
19:39:17 <TrueBrain> also, something that is shitty: if you open 2 Excel files in 2 windows, make changes in one, make changes in the other, go back to the first and hit undo .... it undos the change of the other window!!!
19:39:20 <glx> any way to mark when a step is done ?
19:39:22 <TrueBrain> that fucked me over real good today
19:42:14 <TrueBrain> hmm .. a "main.nut" is required, not?
19:42:34 <glx> only classes are required I think
19:42:48 <Samu> it looks for a file with that name
19:42:54 <glx> but I need to check the source
19:43:15 <Zuu> TrueBrain: main.nut is requried for OpenTTD but not for Excel :-p
19:43:24 <TrueBrain> sorry, I switched subject :P
19:44:12 <Zuu> Yeah it is strange how undo works in Excel. It has bugged me too. Or that when you have multiple windows of the same file it is a bit quirky.
19:44:28 <glx> ah yes scanner search for main.nut
19:49:17 <spnda> Can I somehow obtain OpenTTD 1.8 or maybe even older?
19:49:48 <TrueBrain> glx: isn't that simply replacing "info.nut" with "main.nut", and loading that?
19:50:01 <TrueBrain> knock yourself out :)
19:54:12 <DorpsGek_III_> [OpenTTD/OpenTTD] glx22 commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5MM
19:58:53 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5Mx
20:00:03 <TrueBrain> b'/* A train: \xf0\x9f\x9a\x86 */\n'.decode("latin-1") gives no error
20:00:06 <TrueBrain> that is now what I want
20:00:39 <TrueBrain> decode("ascii") fixes my problem
20:00:53 <TrueBrain> but I guess that means some people cannot upload anymore :P
20:01:12 <TrueBrain> we made it a bit difficult in OpenTTD .. requiring a BOM to use utf8
20:01:53 <TrueBrain> I want to actively refuse those uploads, so people can fix it before it hits BaNaNaS
20:06:19 <TrueBrain> "File contains UTF-8 characters but doesn't contain UTF-8 BOM. OpenTTD won't load this file correctly. Please save the file with 'UTF-8 BOM' encoding."
20:06:23 <TrueBrain> are people going to understand this?
20:06:58 *** rotterdxm has joined #openttd
20:07:34 <LordAro> might want to consider linking to some wikipedia page about file encodings?
20:07:37 <LordAro> but other than that, reads fine
20:07:47 <Zuu> Samu: Is it ok that I log your report as an issue on devzone?
20:08:04 <Zuu> regarding the road path finder.
20:09:59 <Zuu> Samu: You won't get further than me without the help of a devzone admin to authorize your public key. If you're on windows you can enable the optional feature "openssh client" and you'll get ssh-keygen.
20:11:39 <Zuu> TrueBrain: It is probably enough for people who are aware of UTF-8 (BOM) or use an editor like notepad++ which have line-encoding options for BOM character. But most people would not be aware of it without an online search.
20:12:03 <TrueBrain> as long as they can find it with a search, I am fine, tbh
20:12:18 <Zuu> But that probably strengthens the case for actively checking for it and rejecting files encoded wrongly.
20:13:13 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro closed pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5eU
20:13:14 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro commented on pull request #8051: Fix: Don't rerandomise a script's settings when loading from a save https://git.io/Jv5D4
20:13:16 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro commented on pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/Jv5DB
20:13:18 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro closed pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/fjtsp
20:13:52 <LordAro> TrueBrain: i'd also be in favour of "fixing" OTTD's behaviour :p
20:14:02 * Zuu accidently killed their chat client when trying to make a thumbs up emoji :-p
20:14:36 <TrueBrain> LordAro: well, fixing OpenTTD is not as easy as you would think
20:14:43 <TrueBrain> currently we have some people who used latin-1
20:15:04 <TrueBrain> so how I for example now detect what was meant:
20:15:10 <TrueBrain> I check for BOM, if so: decode as utf-8
20:15:18 <TrueBrain> if not, decode as latin-1, and try to decode as utf-8
20:15:27 <TrueBrain> succeed with utf-8? Check if it gives the same result as latin-1
20:15:32 <TrueBrain> no? Someone forgot the BOM
20:15:48 <LordAro> sounds very complicated :p
20:15:57 <TrueBrain> but okay, detection works :)
20:16:14 <LordAro> personally i'd just do utf8 decode -> if fail, fallback to latin-1
20:16:55 <TrueBrain> as long as you do that for the whole file, I guess
20:17:23 <LordAro> that is the -ve side to it
20:17:30 <LordAro> but... how big are the scripts really?
20:17:30 <Zuu> Hm.. are there files with mixed utf-8 and latin that we need to support? Oo
20:17:58 <LordAro> Zuu: that would be a case of "this is not valid utf8, so must be latin-1"
20:17:58 <TrueBrain> well, on a positive note, I only found a few uploads where people used latin-1
20:18:04 <TrueBrain> and almost always with the pound-sign :P
20:18:14 <LordAro> if there are invalid characters, that's the author's fault
20:18:38 <LordAro> s/invalid characters/incorrect characters after latin-1 decode/
20:20:49 <TrueBrain> but okay, at least at upload we are now somewhat prtected from this :)
20:24:20 <TrueBrain> // function GetShortName() { return "WRON"; }
20:24:20 <TrueBrain> function GetShortName() { return "REGR"; }
20:24:25 *** WormnestAndroid has quit IRC
20:24:25 <Zuu> Regarding script size: SuperLib *.nut is about 9000 lines.
20:25:51 <TrueBrain> these are all anti-Zuu lines of code btw :P
20:25:56 <TrueBrain> his tricks no longer work on us!
20:25:59 <TrueBrain> (well, Rb wrote that for musa :P)
20:26:01 *** WormnestAndroid has joined #openttd
20:26:35 <Zuu> LordAro: 324 k chars (using cat *.nut | wc -m)
20:27:12 <LordAro> yeah, nothing too significant really
20:27:38 <Zuu> TrueBrain: So no blame on me? :-)
20:34:17 <TrueBrain> okay, that are a few more regression cases \o/ Still a few more to go, but at least this is growing nicely now :)
20:37:18 <TrueBrain> hmm .. readme.txt and license.txt .. what encoding do they have in OpenTTD ?
20:39:17 *** gelignite has joined #openttd
20:39:36 <TrueBrain> it supports other languages, so I guess that is not possible
20:39:39 *** rotterdxm has joined #openttd
20:41:35 <TrueBrain> yeah, had to be valid UTF-8
20:56:27 *** frosch123 has joined #openttd
21:00:01 <rotterdxm> is it weird if i set the IRC client colorscheme to match my NML language syntax file that i use in notepad++
21:00:24 <rotterdxm> because i don´t want to be right in that case
21:02:02 <TrueBrain> 70% coverage .. getting there ..
21:02:54 <rotterdxm> is this the new BaNaNas API you´re working on?
21:03:43 <rotterdxm> you´re trying to make it backwards compatible with existing uploads?
21:03:52 <rotterdxm> with all their own flavors of weirdness
21:04:00 <TrueBrain> no, I am building the validator
21:04:04 <TrueBrain> to only allow that what is valid
21:04:07 <TrueBrain> so people don't upload crap :)
21:04:15 <rotterdxm> *whistles innocently*
21:04:31 <Samu> Zuu, sorry, I was having dinner. I'm new to SSH stuff
21:04:33 <rotterdxm> so what does coverage mean in this context?
21:04:37 <TrueBrain> okay ... I need a heightmap and a scenario (savegame) that is as tiny as possible .. hmmmmmm
21:04:38 <Samu> actually, i've never used it before
21:05:02 <TrueBrain> rotterdxm: I wrote an API for BaNaNaS, and I am now writing test-cases to validate I did everything as expected. coverage in this case means the amount of lines executed by my test-set
21:05:18 <TrueBrain> that means I have 30% code that is not triggered by the test-case, most likely indicating I need to add more tests :)
21:05:22 <Zuu> Samu: It is no problem. The point is that even if you sort it out you will be at no further point than I am.
21:05:26 <TrueBrain> for example, I don't test all API endpoints yet
21:05:47 <supermop_Home> andythenorthi want beer
21:06:01 <TrueBrain> I keep reading these things in the voice of homer
21:06:46 <supermop_Home> new goal: deliver 800,000 liters of beer to Supermop by April 1 2020
21:08:23 <rotterdxm> ah i see, i just googled that term. interesting, Truebrain
21:08:37 <rotterdxm> so ideally 100% of the code is triggered
21:08:38 *** arikover has joined #openttd
21:08:52 <TrueBrain> 100% is not realistic
21:09:00 <TrueBrain> but at least till the point I don't see any lines I could trigger
21:09:10 <TrueBrain> the benefit is, if some upload ever breaks anything, we can fix it
21:09:18 <TrueBrain> and still know that nothing else is now broken (even more)
21:09:29 <rotterdxm> that is a huge upside
21:09:55 <rotterdxm> how long have you been working on this?
21:10:13 <TrueBrain> the API? First commit was 5th of March, I believe
21:10:58 <TrueBrain> and we have 30 days left to get it in production :D
21:12:10 <rotterdxm> ain´t no pressure without pleasure
21:12:24 <rotterdxm> how come there is a 30 day deadline?
21:14:16 <TrueBrain> the VPSes hosting the current BaNaNaS CDN expire at that point
21:14:20 <TrueBrain> as in, they are EoL
21:14:55 *** arikover has joined #openttd
21:18:36 <rotterdxm> oh whoa. okay so you´re replacing that whole... ecosystem(?) entirely?
21:19:40 <rotterdxm> what happens if you do not make the deadline ?
21:20:10 <rotterdxm> should i start panic downloading newGRFs x)
21:21:15 <TrueBrain> rotterdxm: nah; I just blame Xaroth65 and we can move on
21:21:56 <rotterdxm> it´s good to have a solid cascading blame schema prepared well in advance
21:22:24 <TrueBrain> always rotterdxm , always :)
21:22:26 <Samu> makes it easier to understand what's solved
21:27:13 <Samu> turns out it affects *all* AIs :p
21:27:19 <andythenorth> I have a thick skin usually
21:27:21 <Samu> but not to the point of breaking them
21:27:25 <andythenorth> I could offer blame-as-a-service
21:27:36 <andythenorth> I might need 'service suspended' days sometimes though
21:28:06 <Samu> they just build sub-optimal paths
21:29:07 <Zuu> Samu: I don't rememeber how does squirrel threat numbers when casted to boolean?
21:29:42 *** arikover has joined #openttd
21:29:55 <Zuu> So pre conditions failed is regarded as true. But why does it call it with wrong input data?
21:30:38 <TrueBrain> haha, I now did the stupiest thing ever ... I am generating regression-tests from existing BaNaNaS uploads, and feed that to the new BaNaNaS API, to see if they validate ... slow as fuck :P Because it is restarting the server every time :)
21:30:43 <Samu> the input data was correct, but the bus caused it to be returning -1 all the time after loading a different map size
21:31:47 <Zuu> So when the bug in OpenTTD is fixed the script works fine?
21:32:01 <Zuu> But in older non-fixed OpenTTD versions the AI get in trouble?
21:32:32 <Samu> they don't get in trouble, because most of the time the library was also incorrectly using -1 as true
21:33:32 <Samu> in short, it doesn't break AIs
21:33:37 <Samu> they still build their roads
21:33:55 <Zuu> But if we fix the library to take -1 as false, then it will break AIs? :-p
21:35:15 <Samu> ah, of that i'm not sure, didn't test
21:36:00 <Samu> wait a minute... I did test that!
21:36:29 <Samu> it doesn't break them, it builds with a different path
21:37:29 <Zuu> Is this fix (backported) in 1.10 branch?
21:39:48 <Zuu> So then the problem will solve by itself soon. And the script code could maybe use "> 0" for more correctness. Although it can be argued both ways if production code should check for erors like pre-condition failed.
21:41:48 <Samu> i tested the > 0 on my AI library
21:41:55 <TrueBrain> okay, seems we have some AIs which have UTF-8 but don't have the BOM .. let me see ...
21:41:59 <Samu> tends to build "smarter" roads
21:42:21 <TrueBrain> cool, an AI from LordAro :)
21:42:23 <Samu> not saying the old method was wrong, it was also fine
21:42:37 * Zuu hopes not to be found guilty by TrueBrain
21:42:47 <LordAro> TrueBrain: well of course, i did it correctly :p
21:42:51 <Samu> fine in the sensse of "the path was complete and built correctly"
21:43:12 <TrueBrain> LordAro: pound-signs! :P
21:43:15 <LordAro> i'm fairly certain the squirrel code reads all encodings correctly anyway, doesn't it?
21:43:23 <Zuu> But given the many AIs/scripts I've uploaded it is not unlikely that I've done wrong. :-p
21:43:31 <TrueBrain> LordAro: now you tell me? Really? :P
21:43:41 <LordAro> i don't know what you're doing :p
21:44:00 <LordAro> i fixed something in that area some months ago...
21:44:10 <TrueBrain> if there is no BOM, it is read as ASCII
21:44:24 <TrueBrain> and with ASCII they mean latin-1, but who is counting
21:44:37 <TrueBrain> so I am guessing your info-lines look off
21:44:47 <LordAro> i feel like i would've noticed my info lines looking incorrect
21:44:49 <TrueBrain> "The company now has more than £"
21:45:01 <TrueBrain> would you mind double-checking this?
21:45:30 <LordAro> if it's just read as a sequence of bytes, i'm reasonably certain ICU/whatever just handles whatever it's given
21:45:51 <TrueBrain> so it is even more screwed than we thought?
21:46:37 <Samu> TrueBrain, didn't we check that with company value gs?
21:46:42 <TrueBrain> if you think that LordAro , I could really use confirmation of that claim :) As it kinda depend what we let through on BaNaNaS :)
21:46:48 <Samu> it also makes use of the £
21:46:59 <rotterdxm> so this is sort of like a brexit
21:47:01 <TrueBrain> no, yours was in latin-1
21:47:19 <TrueBrain> LordAro: as if your claim is correct, we don't need any encoding validation tbh
21:48:13 <TrueBrain> would also mean I don't get why we do all this BOM stuff in the code :D
21:48:46 <LordAro> (on Windows/MinGW, at least)
21:48:52 <TrueBrain> I was not out to win anything, to be clear :)
21:49:00 <TrueBrain> I just like to have the validation to be correct ;)
21:49:13 <TrueBrain> okay, so I am rejecting your AIs in that case :D
21:49:28 <Zuu> Also this is not the right win as it means more work. But at least not wasted work.
21:50:09 <TrueBrain> (to be clear, I am taking all BaNaNaS content as-is, that is already uploaded. Not rejecting anything already there :P)
21:50:43 <Samu> LordAro, look at the bright side, you had a crash report to fix!
21:50:44 <glx> yeah the idea is to reject new invalid stuff
21:51:01 <rotterdxm> that is excellent news
21:51:41 <LordAro> TrueBrain: i'm not entirely sure things should be rejected purely based on a possibly incorrect encoding
21:51:50 <glx> anyway wrong encoding only means weird displayed strings
21:51:50 <LordAro> it works fine, after all
21:51:58 <TrueBrain> LordAro: we just established it is not "possible" incorrect, it is incorrect
21:52:09 <LordAro> the string might not actually be used :p
21:52:09 <TrueBrain> and "works fine" can lead to other issues
21:52:12 <LordAro> could be in a comment
21:52:14 <TrueBrain> better to tell people to fix their issues?
21:52:51 <TrueBrain> well, other-way around: if you have the BOM, it really should be valid utf-8
21:52:54 <TrueBrain> otherwise we run into issues
21:52:58 <TrueBrain> so that is as simple as that
21:53:17 <TrueBrain> the other way around, I am willing to dismiss that as a warning-only
21:53:21 <LordAro> yes, no disagreements there
21:53:56 <TrueBrain> for which I currently have no control-flow, so I hate that suggestion :P
21:55:26 <TrueBrain> the one thing I am on the fence about, this also holds for strings displayed via GS
21:55:33 <TrueBrain> which are user-visible, not only for debugging
21:56:04 <Zuu> Especially if the strings come from eints in a language that the author doesn't read themself.
21:57:05 <TrueBrain> I found a readme with a similar issue :)
21:59:42 <TrueBrain> several readme.txt which are saved in latin-1 .. and OpenTTD loads them as utf-8
21:59:46 <Zuu> Most helpful for authors would probably be if both OpenTTD and BaNaNaS used the same validation rules so you can detect the problem early. Not sure if a default .editorconfig for AI/GS would help autors from makning misatkes.
22:00:07 <TrueBrain> most helpful, honestly, would be if OpenTTD accepted utf-8 files :P
22:01:04 <Zuu> If this was a game on steam, we'd set that requirement on next release and content authors would have to update :-p
22:01:33 <Zuu> Ie that all content must be in utf-8 (drop latin-1)
22:01:52 <TrueBrain> "CRITICAL Expeced entry 'errors' to be empty, but it is not; found: '['Unique ID should be exactly four character.']'"
22:02:12 <Zuu> But I guess we care about old stale content. :-)
22:03:12 <TrueBrain> lol ... I forgot, GS of course has their strings not in Squirrel, but in language files
22:03:17 <TrueBrain> so I guess that is fine-ish
22:03:21 <TrueBrain> also, they don't pass validation :P
22:03:24 <Zuu> Is it possible to go by ie API version. If it is 1.10 or higher, file must be utf-8?
22:03:45 <TrueBrain> LordAro: ^^ sounds like a solid idea? :)
22:04:35 <LordAro> except that you need to read the file to know which API version it is? :p
22:04:50 <Zuu> TrueBrain: Mostly GS uses their external string files but it is still possible to build strings only in GS (just not mix with the external ones) and display to the user if I'm not mistaken.
22:05:10 <LordAro> seems a bit backwards to read the file using old OTTD magic, then decide if it got it correct or not
22:05:12 <TrueBrain> lol, all Zuu 's GS fail validation ... GetShortName() fails :P
22:05:23 <TrueBrain> LordAro: it only needs to read info.nut once
22:05:27 <TrueBrain> to know for the whole set
22:06:31 <Samu> my GS had the £ in the info.nut
22:07:03 <Zuu> TrueBrain: What have I done?? :-p
22:07:19 <Zuu> NoCarGoal: function GetShortName() { return "CAGO"; }
22:07:38 <TrueBrain> function GetShortName() { return /*"1234"*/ SELF_SHORTNAME; }
22:07:59 <LordAro> TrueBrain: it doesn't really sounds like you'd be simplifying anything, tbh
22:08:33 <TrueBrain> LordAro: forcing everything to be UTF-8, without BOM etc, does simplify it for users, I think
22:08:43 <TrueBrain> but, it is a suggestion I like, take it for ever you like :)
22:08:43 <Zuu> TrueBrain: which one is that?
22:09:04 <LordAro> TrueBrain: sure, i don't disagree
22:09:04 <TrueBrain> version from 2012 :P
22:09:23 <TrueBrain> hmm .. some people uploaded a readme_settings.txt
22:09:31 <LordAro> but other forms still have to be supported, so there's no particular reason why the "new" UTF-8 should be enforced
22:09:32 <Zuu> Ah.. I guess I was a bit too creative back then.
22:09:36 <TrueBrain> the game only allows readme.txt / readme_nl.txt / readme_nl_NL.txt, as far as I know
22:10:14 <TrueBrain> LordAro: and I guess that is the suggestion: don't support other forms. Only support UTF-8
22:10:18 <TrueBrain> 1 world, simple, clean
22:10:21 <Zuu> Commit message of NoCarGoal version 5: "Version 5: Provide the short name in a way that bananas don't get confused"
22:10:42 <TrueBrain> Zuu: it is no longer confused. It now rejects it :P
22:10:52 <TrueBrain> we talked about writing a Python wrapper around Squirrel
22:10:55 <TrueBrain> so we could execute it
22:11:04 <LordAro> TrueBrain: but they've still got to be read to determine whether we can support them or not
22:11:09 <TrueBrain> for GameScripts, do languages have to be in the lang/ folder?
22:11:12 <TrueBrain> or can they be where-ever?
22:11:35 <TrueBrain> as I have one here where they are in the root folder ..
22:12:04 <TrueBrain> ah, one upload out of many
22:12:09 <TrueBrain> so most likely a boo-boo, which BaNaNaS accepted
22:12:13 <TrueBrain> good, it no longer does :)
22:12:21 <Zuu> TrueBrain: What a lovely task you have to wander through all oddities people have succeeded to upload :-D
22:12:22 <LordAro> i'm not seeing any particular advantage to dropping support for other encodings, when for at least some level, other encodings would have to be implemented
22:12:47 <TrueBrain> Zuu: I have to judge for every of them who was in the wrong ... and as you can see about the encoding talk, many opinions too :D
22:13:38 <TrueBrain> like .. I have a readme_de.txt that is saved as latin-1
22:13:41 <TrueBrain> I wonder how the game shows it
22:13:49 <TrueBrain> what does it do when it fails to load a readme as utf-8
22:14:35 <LordAro> displays lots of questionmarks, probably
22:14:54 <Samu> I didn't know you could have multiple readme files for every language
22:15:22 <frosch123> TrueBrain: str_validate(p, this->text + filesize, SVS_REPLACE_WITH_QUESTION_MARK | SVS_ALLOW_NEWLINE);
22:15:48 <TrueBrain> frosch123: and those flags do what I expect them to do?
22:16:17 <frosch123> whitespace like tabs and formfeed become blanks, other stuff becomes ?
22:16:52 <TrueBrain> I guess here too, the same question: is this a warning, or should we flatout refuse those uploads :)
22:17:00 <frosch123> also, since when does ottd support readme.txt.gz and .xz?
22:17:11 <TrueBrain> I was surprised as much as you are, when I read that blob
22:17:19 <TrueBrain> someone took the effort to implement that
22:17:25 <TrueBrain> because .. they could, I guess?
22:17:34 <frosch123> TrueBrain: non-utf8 users should be auto-banned
22:17:47 <TrueBrain> hahaha, banning them, that is even better :D
22:18:09 <frosch123> it's ridiclous that win10 still defaults to 8bit encodings
22:18:29 <TrueBrain> can't say I disagree ... UTF-8 everything
22:19:05 <TrueBrain> today I found these: 🚢 🚌 ✈
22:19:51 <TrueBrain> no :) I added them to my regression set
22:19:59 <TrueBrain> so check UTF-8 stuff
22:20:14 <TrueBrain> for some reason, I am still getting used to the fact icons exist inUTF-8
22:20:21 <TrueBrain> on one hand it makes me happy
22:20:26 <TrueBrain> on the other .... not sure what to think
22:20:44 <frosch123> aren't they the majority meanwhile?
22:21:06 <frosch123> pretty sure the emoticon font was the biggest last time i checked
22:21:23 <TrueBrain> like .. why are there cat-versions in there?
22:21:37 <nielsm> because japan wanted it
22:22:04 <frosch123> TrueBrain: the fun starts with the prefix-characters
22:22:47 <LordAro> you didn't know about emoji?
22:22:51 <frosch123> for genders, ethnicies, and what more
22:22:59 <TrueBrain> no no nono noooooooo
22:23:09 <TrueBrain> okay, I can see that usecase
22:23:15 <TrueBrain> not the vomit one :P
22:23:33 <frosch123> ah, and for some reason they are postfix
22:23:51 <LordAro> and 10 and 11 and 12 and ...
22:23:52 <TrueBrain> Unicode has versions? Owh boy ...
22:24:12 <TrueBrain> how do you type this shit ...
22:24:48 <TrueBrain> fucking hell .. this is short of insane
22:24:50 <LordAro> otherwise you memorise the codepoints :p
22:24:50 <frosch123> TrueBrain: copy+paste from kcharselect
22:25:02 <TrueBrain> right ... something I could have done without knowing :P
22:25:04 <LordAro> i've never seen someone discover emoji before
22:25:23 <LordAro> here, have a safety vest 🦺
22:25:27 <TrueBrain> I knew emojis were part of the unicode, but I did not know they went THIS FAR with it
22:25:38 <frosch123> TrueBrain: there are new ones every year
22:25:59 <frosch123> and people fight over them because they look different on ios and android
22:26:01 <TrueBrain> in 2020, we are getting one for lungs
22:26:19 <frosch123> and since they are used to transport emotions, people are very upset very fast
22:26:33 <frosch123> since minor differennces in the glyphs appear very different in expression
22:26:45 <frosch123> also you missed the gun drama
22:26:47 <TrueBrain> it is a font ......
22:27:18 <LordAro> 2013 Google is my favourite
22:27:19 <TrueBrain> okay, that is just funny
22:27:23 <frosch123> TrueBrain: 🔫 <- that was a big deal some years ago
22:27:33 <TrueBrain> I understand why, tbh
22:27:37 <frosch123> it was a real gun first, then it got replaced by toy guns
22:27:47 <TrueBrain> I seem to get the twitter-variant
22:27:54 <frosch123> which causes very different reactions when sender/receiver had different real/toy versions
22:28:10 <nielsm> except microsoft's turned from a scifi zapper into something more real gun-ish
22:28:15 <TrueBrain> wanna go 🔫🔫 with me outside? Yeah, I can understand that :)
22:28:17 <frosch123> oh, lordaro has a better link :)
22:29:37 <TrueBrain> this is just insane
22:29:41 <TrueBrain> people have too much time
22:30:27 <TrueBrain> and Tom Scott is awesome :)
22:30:49 <TrueBrain> okay, all NewGRFs validate, except for some readme/license/changelogs .. no surprise there
22:30:53 <frosch123> imagine you are part of unicode, and you have to discuss and approve this shit
22:31:26 <TrueBrain> "dear, what did you do today?" - "well, we talked for 8 hours if the poop emoji should allow the skintone postfix"
22:31:57 <LordAro> i have little doubt the unicode consortium hates emoji and just wants to go back to encoding languages that only 3 people have ever heard of
22:32:07 <LordAro> but if they don't, apple will just make stuff up themselves
22:32:08 *** Zuu is now known as Guest20962
22:32:20 <TrueBrain> Zuu, stay or go .. but you making us nervous :)
22:32:21 <nielsm> yes surely they'd rather pick up the han unification matter again
22:32:22 *** Guest20962 is now known as Zuu
22:33:07 <Zuu> My client keept crashing upon join so I've now found a new one.
22:33:28 <nielsm> anyway I'm off to sleep, careful not to get fooled tomorrow
22:33:51 <TrueBrain> funny, I have a NewGRF that works fine on Windows, and fails on Linux, to show the correct language for its readmes :)
22:33:52 <nielsm> I prepared a small joke for my coworkers tho :D
22:34:01 <LordAro> relatedly, anyone want to approve (& merge?!) #8027 ?
22:34:01 <TrueBrain> (lowercase language-code, instead of uppercase)
22:34:30 <LordAro> TrueBrain: tempted to call that an OTTD issue
22:34:33 <TrueBrain> I can YOLO it if you like?
22:34:46 <TrueBrain> LordAro: filesystem issue, tbh
22:35:11 <Zuu> Any way, what I tried to say when my client crashed was that on my Windows 10, when I open notepad and want to save a new file it now defaults to utf-8 (no BOM).
22:35:19 <LordAro> well yes, but there's no particular reason why OTTD shouldn't accept rEaDmE_nL.TxT
22:35:30 <TrueBrain> on windows, it does :D
22:35:44 <LordAro> *on systems with sane filesystems too
22:35:48 <TrueBrain> BaNaNaS now rightfully refuses uploads with that, btw
22:36:03 <frosch123> LordAro: titlegame?
22:36:31 <TrueBrain> someone is awake :D
22:36:47 <TrueBrain> clearly me, glx and andythenorth were not :P
22:38:11 <TrueBrain> very tempted to fix all those NewGRFs with wrong encoding ..
22:38:19 <TrueBrain> but I am not going to repack all those entries :P
22:38:35 <LordAro> better make sure it works before pushing...
22:38:41 <LordAro> not the one i voted for
22:42:05 <DorpsGek_III_> [OpenTTD/OpenTTD] SamuXarick updated pull request #7661: Codechange: Rework 'start_date' parameter for AIs as a game setting https://git.io/fjMsZ
22:42:34 <frosch123> LordAro: were assertions already disabled somewhere/somewhen?
22:42:36 <Zuu> ah.. title game competition. I missed it but looking through them I would have voted for another one than the winning one. Too crowded for my taste.
22:42:47 <LordAro> frosch123: i don't think so..
22:43:08 <LordAro> is that just a hack in he configure script?
22:46:14 <LordAro> ah crap, and translations need backporting
22:46:28 <LordAro> luckily it's only 21:45
22:46:46 <TrueBrain> I have the same issue, indeed
22:46:57 <TrueBrain> (mainly because it is 22:45 here :P)
22:47:09 <TrueBrain> backporting with my python script is easy btw
22:47:19 <TrueBrain> but yeah .. needs doing :)
22:47:37 <frosch123> you can also forget them, and put them into 1.10.1
22:49:41 <frosch123> andythenorth: someone send 9 screenshots to info@, do you want them?
22:49:59 <TrueBrain> hmm, I just realised that my regression tool kind is a musa replacement .. it does nearly the same, except in a YAML format instead of an .ini
22:51:57 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #8027: Prepare for 1.10.0 release https://git.io/Jv59A
22:52:09 <LordAro> just as well i had some commits ready to dismiss that review :p
22:52:30 <LordAro> frosch123: remind me of your translation backport script?
22:53:08 <TrueBrain> it might kill your puppy; don't hold me responsible plz
22:53:47 <LordAro> i don't see a kill(puppy) in it
22:53:51 <LordAro> so it's probably good
22:54:13 <andythenorth> frosch123 are they nice screenshots? :)
22:54:14 <TrueBrain> frosch123: because when I needed it for 1.9, nobody had one that worked with git .. so I wrote my own
22:54:35 * andythenorth considers changing website to say 'submit PR for screenies'
22:55:09 <TrueBrain> frosch123: and it seems mine takes into account changes in english.txt
22:55:23 <TrueBrain> that was one of the issues with 1.9, where master changed english.txt entries, which should NOT be backported ;)
22:55:27 <LordAro> TrueBrain: wow that was simple
22:55:32 <LordAro> and seemed to work first time!
22:55:51 <TrueBrain> happy it was useful :)
22:55:59 <LordAro> why did i end up with a second checkout when i did 1.9?
22:57:09 <frosch123> TrueBrain: mine does that as well
22:57:14 <LordAro> TrueBrain: dunno, but whichever script i used for 1.9.x required me make another checkout
22:57:31 <frosch123> when strings change in english, they should not be backported, also not in other languages
22:57:32 <LordAro> right, #8027 is now ready
22:57:54 <frosch123> yes, my script is svn-era, and needs two checkouts :)
22:58:22 <frosch123> andythenorth: no idea, they all look same-y, i guess 2 or 3 would have been plenty
22:59:37 <TrueBrain> frosch123: cool; so that means we do have 2 scripts to do the same :P
22:59:51 <frosch123> quite sure rb had a third
23:00:18 <TrueBrain> anyway, mine uses git and produces a diff btw, as a FYI. I just liked doing that :D
23:01:05 <frosch123> TrueBrain: you generate a diff yourself?
23:01:37 <frosch123> he, i did not edit .diffs since mq days
23:01:39 <TrueBrain> if memory serves me right, it gets the diff, and removes the lines that are changed in english.txt :P
23:02:52 <frosch123> ah,--recount makes it a lot easier
23:05:50 <LordAro> now the question is, when do we tag?
23:06:01 <LordAro> the time i set for the debian changelog was 12:00Z
23:06:41 <TrueBrain> when-ever you feel like releasing, tbh
23:07:18 <LordAro> good thing everything's automated, right?
23:07:25 <LordAro> oh, someone should probably write a news post
23:07:50 <TrueBrain> so you can wait till tomorrow, and have who-ever hit that release button
23:07:57 <TrueBrain> or release it when you merged
23:08:00 <TrueBrain> it really doesn't matter :)
23:09:16 <LordAro> alternatively we find the release-CI has broken somehow in the last 6 months and have to redo the tag :p
23:09:43 <TrueBrain> last release was in feb :P
23:10:00 <LordAro> ...i completely forgot about the RCs
23:10:51 <TrueBrain> first hit merge? :D
23:14:50 <LordAro> writing a first draft of the news post
23:16:16 <TrueBrain> nice work LordAro :)
23:24:51 <Samu> 8052 wasn't supposed to be backported?
23:25:00 <TrueBrain> I had a problem .. I used sed ... I now have two problems: sed 's/, '\''/, "/g;s/'\'', /", /g;s/'\'']/"]/;s/\['\''/["/'
23:26:00 <TrueBrain> 50% of the validations errors are UTF-8 related; lol
23:26:08 <TrueBrain> BaNaNaS really did not validate for any of it :)
23:28:15 <TrueBrain> I spot a typo .. can I edit this myself ..
23:28:45 <LordAro> i only used the webinterface anyway
23:29:09 <TrueBrain> there was a , missing
23:30:39 <Samu> LordAro, did you forget #8052, just wondering
23:31:25 <TrueBrain> unpacked, BaNaNaS is 15GB on disk, with 44k files
23:33:05 <LordAro> Samu: ah crap yeah, might as well
23:33:22 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro merged pull request #8052: Fix: [Script] Random deviation upper bound range should be inclusive https://git.io/Jv5gD
23:34:43 <Zuu> "on/offload" => load/unload ?
23:35:02 <Zuu> But maybe that is British to write on/offload?
23:35:12 <LordAro> Zuu: nah, load/unload is probably better
23:35:20 <Zuu> Or ship specific wording that I don't know.
23:36:48 <Zuu> I'm not authorized to github so it is probably easier that you fix it.
23:37:03 <DorpsGek_III_> [OpenTTD/website] TrueBrain dismissed a review for pull request #155: Add: News post for 1.10.0 https://git.io/Jv5HQ
23:37:22 <Samu> I gotta go, take care everyone
23:37:40 <arikover> website/#155 [l. 19] "imporovements" -> "improvements"
23:38:00 <DorpsGek_III_> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #8027: Prepare for 1.10.0 release https://git.io/Jv5Hq
23:38:22 <DorpsGek_III_> [OpenTTD/website] TrueBrain dismissed a review for pull request #155: Add: News post for 1.10.0 https://git.io/Jv5H5
23:38:27 <TrueBrain> spammy spam spam :)
23:39:23 <TrueBrain> LordAro: you know that this last-minute-PR-merge-and-backport is going to bite you in the ass, right? :D
23:40:11 <TrueBrain> why would that be a capital?
23:40:19 <TrueBrain> but the liar is not needed
23:40:29 <LordAro> TrueBrain: for Bad Reasons, i'm currently logged into work VPN to try to fix something because someone merged some changes that broke the Windows build
23:40:38 <LordAro> so yes, i know that feeling quite well
23:41:15 <Zuu> TrueBrain, Sorry. not capital. similiar => similar
23:41:35 <TrueBrain> why does my spellchecker not kick in, I wonder
23:41:45 <DorpsGek_III_> [OpenTTD/website] TrueBrain dismissed a review for pull request #155: Add: News post for 1.10.0 https://git.io/Jv5Hb
23:42:02 <TrueBrain> I like the mispelling of similair
23:42:06 <TrueBrain> I am going to remember that :)
23:45:15 <Zuu> I pasted the text in ms word. One thing it is good at is complaining about spelling. https://pasteboard.co/J1Hh3DC.png The remaining things like "roadtypes" and "titlegame" is more opiniated in my option. :-p
23:45:20 <TrueBrain> "yeah, I am going to run this script before going to bed" - 50 minutes later - ".. I am sure it is ALMOST done, right?"
23:45:58 <TrueBrain> favorite vs favourite .. guess you set it to US? :D
23:46:30 <TrueBrain> US english is not a language :P
23:46:36 * LordAro has to use US English at work due to Customers
23:46:43 <LordAro> i have no idea how to correctly use "license" anymore
23:47:40 <TrueBrain> I only recent learnt there is a difference
23:47:53 <TrueBrain> language is funny like that
23:52:47 <TrueBrain> hmm .. for some reason, a few secnarios fail their md5 check
23:53:45 <TrueBrain> and that seems to be caused by my code
23:54:50 <TrueBrain> "Junk at the end of file."
23:59:54 <TrueBrain> 345 validation errors, 39 of them are because of wrongly uploaded files (wrong name, or just shouldnt-have-been-there), 280 bad UTF-8, 17 because people copy/pasted basests too much, 5 because Zuu, leaving 4 with some random odds and ends .. not bad :) I can explain why all fail, so I am happy :)
continue to next day ⏵
