IRC logs for #openttd on OFTC at 2018-08-11
⏴ go to previous day
00:31:16 *** chomwitt has joined #openttd
03:05:58 *** Supercheese has joined #openttd
03:30:43 <snail_UES_> michi_cc: thanks for looking into that
03:52:50 *** ToBeFree has joined #openttd
05:52:23 *** haudrauf has joined #openttd
06:10:20 *** Supercheese has joined #openttd
06:21:22 *** chomwitt has joined #openttd
06:48:02 *** Alberth has joined #openttd
06:48:02 *** ChanServ sets mode: +o Alberth
07:18:56 *** chomwitt has joined #openttd
07:29:19 *** chomwitt has joined #openttd
07:40:40 *** chomwitt has joined #openttd
07:42:41 <k-man> i need a compact terminus design. for 8 platorms
07:42:47 <k-man> any suggestion of which to build
07:58:32 *** andythenorth has joined #openttd
08:06:12 <peter1138> That doesn't look compact.
08:23:22 <Alberth> k-man: a large part of the fun can be found in figuring out the layout of such things, try building something, and improve it
08:23:31 <Alberth> and yes, path signals can be used.
08:26:29 <SpComb> plus making it fit in whatever space it is that you have available, not just green-field
08:38:24 <andythenorth> today I can poke at nml
08:38:37 <andythenorth> give or take some chores I have to do :)
08:41:31 <peter1138> Today I will be... cycling.
08:41:53 <nielsm> also fyi I'll be difficult to reach for the next 3 weeks, I'm traveling to japan on vacation from wednesday
09:21:16 <Alberth> basically it avoids the "plop pre-designed thing down, add long straight tracks, buy trains, done!"
09:22:28 <Alberth> for extra challenge don't use flat landscape, and don't change it
09:24:35 <Alberth> peter: play with aircraft? copy/paste out of the box then!
09:26:27 <peter1138> Nah, haven't got time to fire up X-Plane this morning.
09:28:26 <Alberth> time seems such a scarce thing, nobody seems to have any
09:31:02 <andythenorth> I don't have time to discuss that
09:32:57 *** chomwitt has joined #openttd
09:52:44 <Wolf01> Mmmh, the script timer doesn't work at all
09:53:29 <Wolf01> Yesterday worked, today it performed join exactly as the same time of the identify command
10:03:35 *** agentw4b has joined #openttd
10:18:11 <TrueBrain> LordAro: ^^ is funny :D I cannot see any way to write a regression test :)
10:18:37 <TrueBrain> as it would require building a rail where the railtype is more than 4 bits :)
10:19:00 <LordAro> i thought the issue was building any rails?
10:19:11 <TrueBrain> I doubt it is currently broken :P
10:19:12 <LordAro> that's how i read it, anyway
10:19:29 <TrueBrain> a << 4 | 1 << 8 ... becomes a << 6 | 1 << 8
10:19:42 <TrueBrain> so all I can see, that bits 4,5 were 0, instread of the desired value
10:19:50 <TrueBrain> which only happens for railtypes that use those bits
10:20:15 <TrueBrain> he said: could not build rail tracks, but I doubt that is true
10:20:40 <LordAro> i feel there's an easy way to find out... :>
10:21:04 <TrueBrain> ah, no, you could test rails to the other axis
10:21:13 <TrueBrain> that is one bit that did change
10:23:39 <TrueBrain> LordAro: no, you are aboslutely right. Tracks will work for what ever AXIS 0 is
10:23:43 <TrueBrain> just not for any other :P
10:24:11 *** agentw4b has joined #openttd
10:24:28 <TrueBrain> so I hope/assume the regression set has one with 0 bits :)
10:24:34 <TrueBrain> and it just needs one with non-zero bits :P
10:26:55 <LordAro> bin/ai/regression/tst_regression/main.nut:1007
10:26:57 <TrueBrain> in the next PR you see lovely the regression hit on a cost change, rightfully :D
10:27:31 <TrueBrain> 4 years no change to regression? Oh my ....
10:28:14 <TrueBrain> leave that location for JGR if you like LordAro :) Might make it a bit easier :)
10:28:38 <TrueBrain> but indeed, only one direction is build .. silly :D
10:30:32 *** agentw4b has joined #openttd
10:30:52 <TrueBrain> owh, the AI API hasnt really changed in 4 years .. that would also explain why regression isnt touched that long :P
10:31:40 <TrueBrain> I need to learn DorpsGek_II to mark things as rotten, that after 3 months he closes stuff after notifying the user every month or so about it
10:33:08 <TrueBrain> or mine, it seems :P
10:33:18 <LordAro> pretty sure all of peter1138's stuff was just waiting on a review
10:33:31 <TrueBrain> he himself didn't label it as such, so .. *shrug*
10:33:48 <TrueBrain> second one for sure needs a few lines of extra code
10:34:02 <LordAro> i think they predate your PR labels :p
10:34:16 <TrueBrain> ITS RIGHT THERE! :P
10:35:17 <TrueBrain> (and you can add labels to PRs that exist before the label did :P)
10:40:30 *** Supercheese has joined #openttd
10:44:03 <TrueBrain> so LordAro, how is cmake going? :D
10:44:43 <TrueBrain> wiki in a docker? :(
10:45:13 <TrueBrain> you are as bad in this as I am :P
10:57:12 <TrueBrain> do I read this right ... the NML version is the version of Python it is running on? :D
10:57:20 <TrueBrain> ah, no variable reusage :D
10:57:52 <TrueBrain> ah, no git support (yet)
10:58:44 <Alberth> I thought we have nml 4 already?
10:59:05 <TrueBrain> the git version on github suggests no version I can find :D
10:59:23 <TrueBrain> the version is fully auto-detected, from what I can tell .. and only supports hg, from what I can spot
10:59:51 <Alberth> makes sense, fr0sch is still working on it afaik
11:01:01 <TrueBrain> ah, commit with a tag 0.4.5 :)
11:01:09 <TrueBrain> seems the tags never made it onto github yet
11:01:16 <TrueBrain> guess frosch should push the tags :D
11:01:27 <TrueBrain> (possibly he doesnt know you need to do that separately :)
11:01:47 <Alberth> tbh, it's silly it doesn't happen by itself
11:02:22 <TrueBrain> happens if 'tags' are just another ref .. but yeah
11:02:52 <TrueBrain> it always feels to me that is an implementation detail which ended up in the workflow :P
11:03:39 <Alberth> git-devs are very much technically correct, and not user-oriented
11:04:22 <TrueBrain> hmm .. so do I experiment with GCP some more .. havent heard back yet from them, which means exactly nothing in either direction ..
11:04:36 <TrueBrain> always tricky .. how much effort do you put into something that might not happen :D
11:04:53 <Alberth> depends on the amount of fun you're having :p
11:05:55 <TrueBrain> there is also the risk that if you find out how awesome it is, the slap in the face when it doesnt happen is too great :P
11:06:21 <Alberth> hmm, good point, beter not get too excited :p
11:09:48 <Alberth> venturing into action A, breaking the idea that each sprite is self-contained :)
11:17:21 <Alberth> so far, I have been decoding 5 sprites in 2 weeks, not very fast :p
11:17:35 <TrueBrain> what are you trying to do? :)
11:19:45 <Alberth> at the highest level, extend nml with station support, below that, understand chips code (an example), below that understand actions, below that, decode grf from its bytes
11:20:06 <TrueBrain> :D I never understood NewGRF .. so I can only imagine :D
11:20:22 <Alberth> it's just assembly language :p
11:23:31 <TrueBrain> so when are we porting AIs to NewGRF? :D
11:25:09 <michi_cc> After we've ported OTTD to a modern language :p
11:25:53 <TrueBrain> I think he means Go :P
11:26:28 <Alberth> ah, go is silly, no non-deterministic choice in receive
11:26:42 *** Wormnest has joined #openttd
11:27:13 <michi_cc> node.js of course, inside a browser based execution environment :)
11:27:35 <TrueBrain> YES! That sounds like a perfect idea! Let me reopen a few issues ;)
11:32:22 <TrueBrain> lol, I am reading how GCP did their CI pipeline .. it is pretty nifty
11:32:31 <TrueBrain> you don't need a Jenkins or something fo rit
11:32:40 <TrueBrain> and it is designed to rune very step in a Docker
11:32:53 <TrueBrain> finally a company that understand that running code in a shared env is a stupid idea if you allow random PRs :)
11:35:01 <TrueBrain> so they made it stupidly easy to run multiple steps, each step in their own docker
11:37:28 *** madmax28 has joined #openttd
12:24:41 *** Progman has joined #openttd
14:15:33 *** chomwitt has joined #openttd
14:25:04 *** wodencafe has joined #openttd
14:49:12 <peter1138> Hmm, can I make git run a regression test before accepting a commit? (Locally, of course)
14:50:04 <peter1138> I may need to rephrase that, it's got to be possible. My real question is, has anyone already done it and can share the incantation so I don't have to figure it out myself? :)
15:06:02 <LordAro> peter1138: there's the ottd githooks repo
15:06:22 <LordAro> nothing that goes as far as running regression tests though, i think
16:08:47 *** frosch123 has joined #openttd
16:15:03 <frosch123> TrueBrain: tags are pushed
16:15:58 <frosch123> i thought "--all" was all :p
16:16:21 <TrueBrain> nothing ever is .. -Wall are not all warnings, and git push--all doesntpush tags :D
16:16:29 <TrueBrain> always something with these programs :P
16:19:23 *** DanMacK has joined #openttd
16:52:51 *** snail_UES_ has joined #openttd
17:40:19 <andythenorth> should I play tanks?
17:45:31 <snail_UES_> andythenorth: how about laying down a good plan for new cargoes? :)
17:49:39 <andythenorth> that requires the nml patch :P
17:52:52 <andythenorth> 16 in / 16 out at industry
17:53:39 <snail_UES_> you can still code in NFO :p
19:17:56 *** tokai|noir has joined #openttd
19:17:56 *** ChanServ sets mode: +v tokai|noir
19:32:11 *** Alberth has joined #openttd
19:32:11 *** ChanServ sets mode: +o Alberth
20:33:05 *** Supercheese has joined #openttd
22:23:15 <andythenorth> nielsm: so where's your nml fork? o_O
22:24:08 * andythenorth has played tanks until literally all fun has been drained
22:24:14 <andythenorth> and then some more :x
22:45:18 *** andythenorth has joined #openttd
22:45:26 <andythenorth> computer crashed :(
23:13:32 <michi_cc> snail_UES_: You need to have 10 04 in your Action 0 (i.e. bit 2 of the railtype flags set).
continue to next day ⏵
