IRC logs for #openttd on OFTC at 2018-04-16
⏴ go to previous day
01:53:24 *** Smedles has joined #openttd
03:06:54 <arahael> LordAro: That's 10.13.4, I think. :)
03:40:38 *** Darkvater has joined #openttd
03:40:38 *** ChanServ sets mode: +o Darkvater
05:03:38 *** muffindrake4 has joined #openttd
06:39:11 *** Wacko1976 has joined #openttd
06:51:33 *** ToffeeYogurtPots has joined #openttd
07:55:29 *** Arveen2 has joined #openttd
08:09:50 *** andythenorth has joined #openttd
08:35:17 *** Wacko1976 has joined #openttd
08:37:25 <peter1138> Have you heard the word, about the bird?
08:39:03 *** Thedarkb1-X40 has joined #openttd
08:45:32 *** Smedles has joined #openttd
08:45:55 <andythenorth> now we just need a better livery GUI
08:47:22 <Pikka> bright top pixel on -> yellow end?
08:48:18 <andythenorth> dunno about the steam coaches
08:48:23 <andythenorth> windows are tall
08:49:21 <Pikka> not bad... taller/narrower looks older
08:51:32 <andythenorth> also should Scooby Doo be solid CC?
08:52:31 <Pikka> dunno... I've been keeping the stripe so it all still matches with standard passenger and mail cars.
08:53:40 <peter1138> Better livery GUI? What did you envision?
08:54:05 <andythenorth> (1) usability stuff, get rid of the daft tickbox thing
08:54:24 <andythenorth> (2) 5 types of pax coach, but only 1 freight?
08:54:31 <andythenorth> and pax colour is also mail?
08:54:55 <andythenorth> I don't have a proposed solution, I following truebrain's rule, complain about problem :P
08:55:09 <andythenorth> I wondered about applying livery to groups
08:55:19 <andythenorth> and it's either simple: engine livery, wagon livery
08:55:40 <andythenorth> or it's full wtf: engine livery by type (visual effect), and wagon livery by cargo refit
08:55:52 <peter1138> Was that another thing I had a patch for? :S
08:55:57 <andythenorth> groups don't do anything :)
08:56:05 <andythenorth> no management of orders
08:56:12 <andythenorth> no making groups from 'vehicles using this station'
08:56:29 <andythenorth> groups are waste of space, but "hey, they're hierarchical now"
08:57:15 <andythenorth> liveries need a bit of fun applying imho
08:57:30 <andythenorth> it's not a place where 100% logic is required
08:57:31 <peter1138> They're not a waste of space, they just don't do everything you want them to do.
08:58:26 <andythenorth> how do we improve anything without dramatic unreasonable statements though? :P
09:01:09 <peter1138> Hmm, why do I have these old patches?
09:03:17 <peter1138> --- station_cmd.c (revision 2979)
09:04:38 <peter1138> CMD_SET_GROUP_LIVERY
09:04:42 <peter1138> Yeah, I have a patch for group liveries.
09:06:04 <peter1138> Maybe I should rebase it :p
09:36:06 <_dp_> peter1138, when your patch is older than some players
10:10:49 *** andythenorth has joined #openttd
10:28:52 <andythenorth> dunno about the buffet
10:30:17 <Pikka> first window full size?
10:32:04 <andythenorth> might need to be symmetrical
10:33:25 <andythenorth> doesn't have to be a buffet, might be something else
10:35:31 <Pikka> it's just a different-looking coach in the middle for visual interest
10:35:36 <Pikka> doesn't need to be prototypical ;)
10:36:06 <andythenorth> I have to repeat this for 5 other generations, so keep it simple eh
10:36:13 <Pikka> all narrow windows? hmmm
10:36:34 <andythenorth> I went on a french train with a playground coach in it once
10:36:38 <andythenorth> had a climbing frame
10:39:04 <andythenorth> who will finish first? :P
10:40:03 <Pikka> are your coaches all symmetrical? I'm wondering if early coaches having different corridor sides will confuse people
10:41:41 <andythenorth> they're not all symmetrical no
10:41:56 <andythenorth> but generally I'm finding symmetry looks better, and is faster to work with
10:47:54 <V453000> andythenorth and Pikka drawing trains? Did I just wake up 10 years ago? :D Sounds great!
10:49:59 <andythenorth> and OpenTTD is dying again V453000 !
10:50:04 <andythenorth> new github and everything
10:50:17 *** ToffeeYogurtPots has quit IRC
10:51:15 <V453000> I will just fix PURR tonight and lets see if I can proceed with some work on some trains too :)
10:54:27 <andythenorth> I rearranged all my spritesheets
10:54:45 <andythenorth> so that the columns and rows match photoshop 'press cursor' distance
10:55:45 <andythenorth> shift-cursor key
10:55:49 <andythenorth> when using marquee or move tool
10:56:03 <andythenorth> they used to be off by 1 or 2
10:56:11 <andythenorth> I wrote a PIL script to move everything
10:56:27 <V453000> not too surprised byu that :?D
10:56:33 <V453000> can't imagine moving everything by hand
10:56:40 <V453000> don't even want to guess how many sprites do you have
10:59:45 <andythenorth> V453000: only about 10 or 20 thousand
11:00:09 <V453000> could be worse but it's still a lot :)
11:00:26 <V453000> very nice wagons, I love the variety
11:00:40 <andythenorth> someone said horse is too flat
11:00:51 <andythenorth> I am -1 to random noise crap
11:01:08 <andythenorth> Pikka planning to add random noise? o_O
11:01:23 <V453000> I think it's in a good spot
11:01:31 <andythenorth> noise just damages the shape
11:01:32 <V453000> judging from these pictures here atm
11:01:39 <andythenorth> it's for people who can't actually draw shapes
11:01:48 <andythenorth> "wtf this is getting hard, I splatter some pixels"
11:02:20 <V453000> it's not like it's completely flat, you have a bunch of "random" pixels there, and the window variety does a lot
11:02:49 <Pikka> I've a little bit of noise on, eg, coach roofs. but not too much.
11:03:39 <Pikka> there's enough variety in the shading that nothing looks "flat" anyway
11:04:32 <Pikka> not for UKRS3, or anything else I'm working on at the moment
11:04:48 <V453000> so classic pixel pushing? :)
11:05:14 <Pikka> and some procedural generation with Processing
11:05:15 <V453000> so, working on UKRS3? What are the ideas for it if I may ask?
11:05:25 <V453000> :0 procedural generation? :0
11:06:00 <peter1138> andythenorth, new carriages.
11:06:07 <peter1138> andythenorth, nice carriages.
11:06:23 <peter1138> Now do some dirty variations ;p
11:06:45 <Pikka> openttd needs more weathering or something
11:07:07 <andythenorth> weathering script in PIL
11:07:13 <andythenorth> would totally work
11:07:26 <V453000> darken all pixels by 1 index
11:07:28 <andythenorth> apply 7 random pixels of dirt to each sprite
11:07:53 <andythenorth> V453000 that won't work because Livery Aging isn't allowed because 32 day callback is all wrong
11:08:06 <andythenorth> someone closed that issue anyway :)
11:08:52 <V453000> Pikka: omg nice, will read, thanks :)
11:10:56 <V453000> so drawing 32bpp x2 eh :) very nice
11:11:37 <V453000> I'll probably keep trying to smash my head against 3D for my next trains
11:11:44 <Pikka> it seems to be the sweet spot... big enough for more detail, small enough that the pixel-drawing style still works
11:12:44 <peter1138> I regret making 4x ;(
11:12:58 <V453000> very nice, yeah x4 is crazy much
11:12:58 <peter1138> It only started out as a zoom level originally.
11:13:27 <V453000> how well do the graphics work in x1?
11:14:07 <Pikka> x1 is dots moving around a screen at modern resolutions anyway :P but yes, not bad.
11:14:57 <V453000> yeah I changed my screen as well and everything is a bit smaller ._.
11:15:32 <peter1138> I might find myself a high resolution monitor.
11:19:37 <V453000> I feel like no matter what I do with 3D rendered stuff, it won't look better than your 32bpp x2 hand drawn stuff ... might be influenced by the fact that your stuff is simply awesome art :D
11:21:10 *** WWacko1976-work has joined #openttd
11:22:52 <peter1138> Hmm, 31" 8K monitor...
11:24:13 <peter1138> I may need to upgrade my RAM again.
11:24:58 <Pikka> :) well it's all different, V... hand-drawn vs rendered, "realistic" vs fantastical...
11:27:02 <V453000> yeah my things might not even look like trains almost at all
11:27:32 <V453000> regardless I think I need so many different sprites that hand drawing is simply not viable for me atm
11:30:33 *** Stimrol has joined #openttd
12:20:29 <Eddi|zuHause> i have two 2GB DDR2-1000something RAMs lying around here that my friend gave me from his computer that died
12:20:59 <Eddi|zuHause> no idea what i would use those for
13:06:58 *** FLHerne has joined #openttd
13:13:32 *** TheMask96 has joined #openttd
13:13:55 <peter1138> I was using that sort myself up until a couple of months ago.
13:22:25 <Eddi|zuHause> i think it's slightly too low for putting into my current computer
14:05:38 <peter1138> So when I was upgrading my PC I decided to put off getting a new GPU until a month later.
14:05:44 <peter1138> Then the prices went up :(
14:21:57 <Eddi|zuHause> i was considering getting a new PC for the past year, but never got around to doing it
14:22:13 <Eddi|zuHause> i could use a laptop
14:38:17 *** snail_UES_ has joined #openttd
14:41:04 *** andythenorth has joined #openttd
14:44:23 <Sacro> Eddi|zuHause: then you could leave the hause
14:45:04 <Eddi|zuHause> i'd have to think of a new nick
14:45:29 <Sacro> I can't think of many german places to go
14:52:20 *** ChanServ sets mode: +v tokai
14:53:58 *** sim-al2 has joined #openttd
14:54:08 <Eddi|zuHause> it's more complicated than that, you also have to change the preposition
14:54:36 <Eddi|zuHause> bei/auf/über/...
14:54:51 <Eddi|zuHause> possibilities are endless
14:56:15 <Eddi|zuHause> "zu" only works for going to a place, not staying at a place. except for home
14:56:42 <Eddi|zuHause> (and a few special cases, too)
14:57:48 <Eddi|zuHause> ("zu Tisch" comes to mind, although that is a rather old fashioned phrase)
14:59:09 <Eddi|zuHause> (although that rather describes the process of eating rather than staying at a table)
15:47:27 *** sla_ro|master has joined #openttd
16:04:10 *** Alberth has joined #openttd
16:04:10 *** ChanServ sets mode: +o Alberth
16:50:26 *** andythenorth has joined #openttd
17:01:07 *** synchris has joined #openttd
17:18:39 <andythenorth> diesel engines are going to get a 2nd livery, which shows when the engine is flipped
17:18:44 <andythenorth> I was trying to avoid this :P
17:27:34 *** Wormnest has joined #openttd
17:34:07 <peter1138> Just spent an hour diagnosing code that failed to work.
17:34:10 <peter1138> Then it started working.
17:38:46 <andythenorth> usually when I do that, I'm in the wrong repo :P
17:38:58 <andythenorth> remarkably often when I have multiple checkouts
17:49:30 *** TheMask96 has joined #openttd
17:49:37 <andythenorth> I guess we wouldn't allow 4 flipped states? o_O
17:51:36 <Alberth> I can't quite imagine what 4 flipped states would mean
17:52:04 <andythenorth> it's just a counter on how many times user has flipped the vehic;e
17:52:49 <andythenorth> it's another silly idea for arbitrary, user-editbale storage per vehicle
17:52:59 <andythenorth> similar to cargo subtypes, and just as bad
18:03:58 <Eddi|zuHause> if that's not a BAD FEATURE waiting to happen, then i don't know
18:04:53 <Eddi|zuHause> on a related note: is the flipped state kept through autoreplace?
18:06:23 <andythenorth> I doubt it somehow
18:07:54 <andythenorth> colour me surprised :P
18:08:18 <Eddi|zuHause> now try autoreplacing with engines that cannot be flipped :p
18:09:28 <andythenorth> that's why I'm surprised it works at all
18:09:35 <andythenorth> I guess something just copies as many props as possible
18:15:35 <andythenorth> peter1138: how about default company colour settings per vehicle in buy menu? o_O
18:18:51 *** Progman has joined #openttd
18:20:09 <Alberth> multi-company buy menu or so?
18:21:41 *** frosch123 has joined #openttd
18:21:51 <andythenorth> 'buy this vehicle in red'
18:21:56 <andythenorth> 'buy this vehicle in pink'
18:22:02 <andythenorth> I know the train fans want the real liveries
18:22:17 <andythenorth> CC is more fun, by miles
18:33:12 <TrueBrain> frosch123: #6730, assuming he is asking because he wants to implement it, is tha tsomething we accept PRs for?
18:33:54 <frosch123> that guy has forked cppcheck, i guess they check every project on github :)
18:34:42 *** andythenorth has joined #openttd
18:34:54 <TrueBrain> but does he make any chance if he would make those PRs? :)
18:35:04 <DorpsGek> andythenorth: supermop was last seen in #openttd 6 days, 2 hours, 23 minutes, and 26 seconds ago: <supermop> should look good out of the box
18:35:30 <frosch123> TrueBrain: for "strdup" certainly not :)
18:35:58 <frosch123> i would rather use std::string everywhere, than checking success of memory allocation
18:36:45 <frosch123> i wouldn't even know what ottd should do if strdup fails
18:37:28 <LordAro> TrueBrain: i get that warning with gcc7.3
18:37:47 <TrueBrain> LordAro: I just dislike issues with too little information
18:38:00 <TrueBrain> learning peopl ehow ot write proper bug reports is also a goal in life :)
18:38:21 <LordAro> discussed it with frosch123 before, it's probably spurious
18:38:33 <LordAro> there's several lines in that file that are very similar, but only that one triggers it
18:38:39 <frosch123> yes, i also though thar lordaro came with that before :)
18:39:25 <frosch123> TrueBrain: actualy, rb already removed all usages of strdup, it only exists in depend.cpp
18:40:09 <TrueBrain> frosch123: wrote a nice reply (I tihnk) :)
18:40:23 <TrueBrain> in general I am tempted to close all these kind of tickets like that; let me know ifyou agree with the content or not :)
18:41:19 <frosch123> you should have written lordaro instead of frosch :)
18:42:06 <TrueBrain> I just hope the content of tickets improves .. so far I only have seen ambigous reports :(
18:43:36 <TrueBrain> really considering an issue template, if this continues :)
18:43:38 <LordAro> needs a CONTRIBUTING.md
18:44:18 <andythenorth> but I want the one that looks best, not most realism
18:45:04 <andythenorth> I am glad you said that
18:45:16 <andythenorth> reinforces my prejudice :)
18:45:42 <LordAro> TrueBrain: find the hackmd.io link from yesterday!
18:46:03 <TrueBrain> LordAro: I think I wrote enough bla for one week :P Now it is up to you and andythenorth :)
18:46:20 <andythenorth> I will write some in a bit, once chores are done
18:46:38 <andythenorth> if I get stuck on some sections I'll leave them out and ship a v1
18:46:44 <andythenorth> something is better than nothing here
18:48:53 *** Thedarkb has joined #openttd
18:56:29 *** agentw4b has joined #openttd
18:59:47 <TrueBrain> the GitHub went from 50 unique visitors to 300 after the forum post :) seems people read forums :)
19:00:05 <TrueBrain> owh, no, not forums
19:00:07 <TrueBrain> was via the website
19:00:17 <TrueBrain> so people read news posts :)
19:01:50 <TrueBrain> frosch123: coding style bla, what is your opinion on #6726 ? :)
19:03:15 <TrueBrain> LordAro: my ICC license is in btw
19:07:54 <Wolf01> peter1138: I read the comments about the TODOs, I think some of them aren't really TODOs, or they were just put there for later refactoring, I would focus on the ones with the subtypes and the SE one (if really needed to check roadtype availability)
19:12:24 <peter1138> I just went through and posted where there are TODOs.
19:12:53 <TrueBrain> lol, 50 euro per month to rent a virtual mac :D
19:13:11 <TrueBrain> or 1 dollar per hour
19:13:17 <andythenorth> there are $1 / hour versions
19:13:20 <andythenorth> oh you found it :P
19:13:31 <andythenorth> how many hours / month we need?
19:13:48 <TrueBrain> 2 hours per release or so
19:13:55 <andythenorth> I could just hang an old mac off my network, but eugh
19:14:00 <andythenorth> then I am responsible
19:14:09 <TrueBrain> which is the biggest issue :)
19:14:15 <andythenorth> I can donate $20 / month for mac host easy
19:14:18 <peter1138> Still having issues cross-compiling it? :(
19:14:27 <TrueBrain> that is not an issue that will be solved
19:14:47 <TrueBrain> as fundamentally, having a 3rd party implement fixes, never results in good results :)
19:14:55 <frosch123> TrueBrain: what is the question about 6726? the one you already answered?
19:15:12 <TrueBrain> I was just wondering if there is a bigger line in OpenTTD source there
19:15:24 <TrueBrain> I know we used to do it like that; but that was 10 years ago :D
19:15:40 <frosch123> we have LAST for the last valid item, and END for the first invalid item
19:15:53 <frosch123> as long as those are not mixed, i am fine
19:16:11 <TrueBrain> peter1138: can we introduce unit tests? :D
19:16:49 *** muffindrake has joined #openttd
19:17:11 <peter1138> I just wanted some way to see it working without me having to think about it :p
19:17:51 <TrueBrain> or what were you tihnking about?
19:18:39 *** agentw4b has joined #openttd
19:20:50 <andythenorth> much time ago :|
19:22:03 <Wolf01> I would really like unit tests
19:22:48 <Wolf01> At least to be sure the 43563467476 cases of a simple function which checks if something is allowed to do in a tile work as intended
19:23:43 <Wolf01> And I don't speak for the functions with 1e238 cases, there would be really helpful
19:25:19 <peter1138> TrueBrain, something that uses the new string codes, I guess.
19:25:34 <frosch123> Wolf01: that is all around "town always build road"
19:25:48 <peter1138> I can see the code probably works, but I've not tried it.
19:25:52 <frosch123> there must be some road available so town can grow
19:26:19 <frosch123> when i was working on that branch towns always built ROAD, so ROAD was always required
19:26:20 <Wolf01> Ok, then I could have handled the cases on the branch
19:26:51 <Wolf01> Still missing a lot of things, but the core part for towns is there
19:28:36 <frosch123> everytime i see a wot ad, i think of andy
19:29:56 <andythenorth> it was the ads that got me playing
19:30:02 <andythenorth> but now I've stopped
19:30:17 <andythenorth> I only have limited tolerance for being sent abusive DMs
19:58:25 *** FLHerne has joined #openttd
20:06:14 *** HerzogDeXtEr has joined #openttd
20:23:05 *** FLHerne has joined #openttd
20:24:08 <Thedarkb> Did you sort out the library issue on the Debian/Ubuntu build machines?
20:47:45 *** sim-al2 is now known as Guest128
20:47:46 *** sim-al2 has joined #openttd
20:49:58 <andythenorth> so many open tabs
20:50:11 <andythenorth> one of them has lordaro's CONTRIBUTING paste in it
20:58:19 <TrueBrain> smee I finally found a good wy to make an SDK :)
20:58:56 <TrueBrain> but extracting XCode TAKES FOR EVER :)
21:00:14 <peter1138> TrueBrain, how do i pull a PR locally?
21:00:38 <TrueBrain> peter1138: google.com :) Something with refs/pull/<number>/head .. never sure about syntax
21:00:45 <TrueBrain> git fetch upstrean refs/pull/1111/head ?
21:03:16 <peter1138> Ah, git fetch upstream refs/pull/6726/head:6726 < makes a local branch called 6726.
21:03:28 <peter1138> Thanks for the hint :p
21:03:55 <TrueBrain> FETCH_HEAD .. local branch .. tomato tomat :)
21:09:41 <peter1138> Hmm, what's the optimal -j for parallel compiling these days?
21:09:53 <peter1138> cores, cores * 2, cores + 1, ...
21:10:38 <peter1138> what about hyperthreading?
21:10:42 <TrueBrain> if your system has HT, cores is sufficient. If it doesnt have HT, add a few more
21:10:55 <TrueBrain> HT on its own will keep your CPU busy enough
21:11:05 <debdog> (except one has many cores but little ammount of RAM)
21:11:25 <TrueBrain> little ram? *looks at his machine, looks at debdog, looks back at his machine* NAH! :D
21:11:58 <TrueBrain> that would mean compiling a single file takes more than 1GB :P
21:12:04 <debdog> so, you prolly not one of these exceptions
21:14:07 <TrueBrain> finding out which Xcode dropped 10.12 SDK feels like downloading every version :(
21:15:02 <TrueBrain> owh well, 10.13 should work .... :D
21:22:15 <TrueBrain> andythenorth: which clang version does your OSX have?
21:22:54 <andythenorth> Apple LLVM version 9.0.0 (clang-900.0.39.2)
21:23:01 <andythenorth> Target: x86_64-apple-darwin16.7.0
21:24:16 <LordAro> now try to work out what actual LLVM version number that corresponds to
21:26:30 <TrueBrain> 16.7 is weird .. 10.11 was darwin15 .. so 12 and 13 are 16? OSX is funny
21:36:18 <andythenorth> we got any other licenses besides GPL v2?
21:36:49 <TrueBrain> hmm .. andythenorth, which OSX version are you running?
21:36:58 <TrueBrain> I assumed 10.13, but new information suggestions different :D
21:37:04 <andythenorth> 10.13 is well dodgy
21:37:04 <TrueBrain> should have asked :P
21:37:12 <andythenorth> Apple clown shoes
21:37:33 <andythenorth> I might have to rewrite CONTRIBUTING.md from scratch
21:37:59 <TrueBrain> the file is licensed?
21:38:17 <andythenorth> everything is licensed everywhere, etc
21:38:18 <TrueBrain> GitHub too has suggestions btw
21:39:46 <andythenorth> I'm starting to go off GPL
21:39:55 <andythenorth> I was a big fan, but it's problematic
21:40:09 <TrueBrain> MIT is a bit nicer license
21:40:14 <TrueBrain> or CC- variants, if you want to be more picky
21:40:18 <TrueBrain> but GPL .. is always .. icky
21:40:48 <andythenorth> I discovered my newgrfs were violating it
21:40:56 <andythenorth> so I've had to armchair lawyer some stuff
21:41:17 <TrueBrain> if you touch anything GPL .. CONCRATZ! YOU ARE NOW GPL TOO :P
21:44:30 <andythenorth> yeah, so what am I going to do?
21:44:38 <andythenorth> not use fonts in my project because WOO FONTS
21:44:48 <andythenorth> not provide html documentation because WOO BOOTSTRAP IS MIT :)
21:47:48 *** gelignite has joined #openttd
21:51:39 <LordAro> i thought i recognised the name from somewhere - elfring also hit freerct with a load of useless issue reports
21:52:31 <TrueBrain> and he seems unable to read a response I guess *shrug*
21:52:54 <TrueBrain> it almost feels like a bot tbh :)
21:52:56 *** elfring has joined #openttd
21:53:04 <TrueBrain> talking of which :)
21:53:18 <TrueBrain> LordAro: frosch123 told me to point him to you :)
21:53:50 <andythenorth> what's the deal with 80 char formatting in CONTRIBUTING.md?
21:53:55 <peter1138> Yeah, that comment is... not relevant to what you wrote.
21:53:59 <andythenorth> currently it's mixed, and looks daft
21:54:02 <TrueBrain> andythenorth: ditch it; new rule: newline at the dot
21:54:07 <TrueBrain> that makes reviewing SO MUCH EAISER
21:54:45 <andythenorth> TrueBrain: "at the dot"? o_O
21:55:10 <TrueBrain> elfring: we welcome any PR within sanity :) Talk to LordAro if you want to see if your PR has merrit before creating it :) Saves a lot of time for both you and us :)
21:55:21 <TrueBrain> andythenorth: yes :) So 1 sentence per line
21:55:26 <TrueBrain> means a diff means something
21:55:32 <andythenorth> 80 chars is so old
21:55:33 <TrueBrain> not saying we should :) Just saying it really helped me :)
21:55:44 <TrueBrain> well, the issue is if you change 1 word, you have to change the whole block :D
21:55:52 <andythenorth> we had a thing at work about 80 chars for a while so someone could read it in their phone or something
21:56:06 <andythenorth> or because they had 25 bash windows open inside emacs
21:56:10 <andythenorth> anyway we sacked that
21:56:14 <andythenorth> makes for bad python code
21:56:19 <TrueBrain> I have an ongoing discussion .. tempted to buy better screens for those people :P
21:56:25 <elfring> Which contributors will care for more complete exception handling?
21:56:33 <TrueBrain> for Python I do 120 .. to keep people a bit in line :)
21:56:42 <TrueBrain> elfring: same question, same answer
21:57:09 <peter1138> I was wondering about getting a 4K monitor...
21:57:18 <peter1138> Then I realised I'd probably want three.
21:57:25 <TrueBrain> I want a 27" 2k .. but it is so hard to find a good one :(
21:57:27 <peter1138> And *then* I saw that an 8K monitor exists...
21:58:39 <andythenorth> TrueBrain: those people don't work with me anymore ;)
21:58:57 <TrueBrain> you didnt like the pun? 120 chars? In line? :)
21:59:18 <andythenorth> CONTRIBUTING.md !!!!
21:59:21 <elfring> LordAro: Would you like to clarify exception handling any more?
21:59:31 <andythenorth> I'm sure some links are broken
22:00:25 <TrueBrain> elfring: no, Open Source works differently; what error flow do you intend to correct?
22:02:06 <elfring> TrueBrain, LordAro: Static source code analysis can point some open issues out in this case.
22:03:18 <TrueBrain> elfring: I am not looking for reports of what is wrong; I am looking for people to fix those issues via Pull Requests :)
22:03:46 <TrueBrain> we run Coverity over our code; we do static analysis :)
22:04:07 <TrueBrain> just not every warning is worth the time of fixing :)
22:04:28 <TrueBrain> LordAro: indeed, GCC 7.2 is throwing that error; guess I should add it to the CI soon :P
22:05:23 <elfring> TrueBrain, LordAro: There are a few design approaches available to improve the affected software situation.
22:05:30 <andythenorth> just needs "WRITE ABOUT CI HERE" removed or written
22:05:48 <TrueBrain> *looks around* nope andythenorth :)
22:05:48 <andythenorth> then I PR it as a single commit
22:06:10 <andythenorth> does it really need details today?
22:06:15 <TrueBrain> elfring: are you planning on writing code to apply those approaches? :)
22:08:35 <elfring> TrueBrain, LordAro: Which warnings did you get by the tools Coverity and GCC for the mentioned issue?
22:09:02 <TrueBrain> elfring: how is that relevant?
22:10:03 <elfring> TrueBrain: I am looking for acceptance possibilities for affected software design aspects.
22:10:22 <TrueBrain> elfring: do you plan on implementing those possibilities in code?
22:10:29 <andythenorth> TrueBrain: [distracting you here] "Continuous integration (CI) tools monitor pull requests, and help us identify build and code quality issues. The results of CI tests may show on the pull request."
22:10:34 <andythenorth> can I ship that?
22:11:01 <andythenorth> better than "WRITE ABOUT CI HERE"
22:11:11 <TrueBrain> The results of the CI tests will show on your pull request. By clicking on Details you can further soom in; in case of a failure it will show you why it failed. In case of success how awesome you were.
22:11:32 <TrueBrain> possibly rephrase a bit
22:11:34 <TrueBrain> but you get the jizz :)
22:11:43 <TrueBrain> it is not MAY show .. it WILL show :)
22:11:49 <elfring> TrueBrain: Higher level development tools can help more to improve also this software, can't they?
22:11:55 <TrueBrain> (without a success,i you cannot accept pull requests andythenorth :))
22:12:13 <TrueBrain> elfring: not the question; I ask if YOU are willing to contribute with code?
22:12:29 <TrueBrain> elfring: as we all know things to improve; so only telling us it should etc, does not really help
22:12:46 <andythenorth> someone going to teach me interactive rebase for this .md file?
22:12:52 <andythenorth> or I just do one big diff against trunk? :P
22:14:40 <elfring> TrueBrain: It depends. - I hope that advanced tools can be used so that some work can eventually be reduced.
22:15:30 <TrueBrain> elfring: you talk in riddles and in circles, sorry :(
22:15:40 <TrueBrain> elfring: it almost feels you are trying to sell something .. :(
22:15:47 <__ln__> Oh, elfring is here and I'm out of popcorn.
22:16:10 <TrueBrain> __ln__: be nice; wait, that is not in your dictionary. be not-your-usual-self!
22:17:06 <elfring> TrueBrain: You might notice that I am trying to improve several software components in a similar way for a while.
22:17:34 <TrueBrain> elfring: tools are a means to an end. Without people to be willing to dedicate their time to follow up on it, tools are just that: tools
22:17:50 <TrueBrain> elfring: a shovel in my garden can do wonders. But without me operation, it isjust a piece of wood with a piece of iron attached
22:18:15 <TrueBrain> elfring: we need people who not only tells us: you should deal with a result value of strdup (better yet: dont use strdup), but who make a Pull Request to do something with it
22:18:54 <TrueBrain> elfring: are you that person for us?
22:18:54 <elfring> TrueBrain: But you can also let a machine perform the digging for you, can't you?
22:19:11 <TrueBrain> elfring: a machine can tell me all the errors in the world; with no human to correct them, they mean nothing to me
22:19:54 <TrueBrain> (and a machine that digs my garden on its own, with no humans, does not exist as of yet ... maybe in a few years :D)
22:20:09 <TrueBrain> scary .. I come home one day: A SWIMMING POOL!
22:20:44 <andythenorth> it's linked from CONTRIBUTING
22:20:57 <elfring> Can better exception handling help you more automatically?
22:21:09 <TrueBrain> andythenorth: no, that is git; at least, a quick glance gives me that impression
22:21:15 <TrueBrain> andythenorth: just everyone mostly ignores it :D
22:21:38 <TrueBrain> elfring: and we are back in a circle :) Not sure if it is the language, or that you are reading off a script :(
22:23:18 <andythenorth> we gonna have LICENSE.md?
22:23:23 <andythenorth> github likes it if we do
22:23:29 <TrueBrain> rename our COPYING, you say?
22:23:59 <andythenorth> TrueBrain: I'll link COPYING
22:24:03 <TrueBrain> I think this conversation has come to an end elfring; sorry.
22:24:20 <andythenorth> there's a paradigm to debate
22:24:42 <andythenorth> my experience of paradigm-driven developers is universally negative so far :|
22:24:55 <TrueBrain> the moment you start a build, and after 10 files cmake still says: 0% done .. oh-oh
22:27:19 <TrueBrain> this really is going to take a long time to build :D But at least it is building what I expect him to build .. which is nice ..
22:28:46 <andythenorth> let's see if I did it right :P
22:28:52 <andythenorth> I am bad at bureaucracy
22:29:11 <andythenorth> yeah broken ignore
22:29:32 <andythenorth> I clicked a magic git button, but it was the wrong magic
22:32:20 <TrueBrain> please rebase :D (most likely one of the last time that you are obligated to do that, but .. you use a too old master that doesnt fix that yet :D)
22:33:14 <glx> anyway it's always nice to create a PR from an up-to-date clone
22:33:21 <elfring> How do you think about to encapsulate error detection and corresponding exception handling as a reusable aspect in your software?
22:34:13 <andythenorth> ok so I have to learn rebase now
22:34:55 <TrueBrain> git rebase upstream/master
22:35:21 <andythenorth> I was close by guessing :P
22:35:26 <andythenorth> not close enough
22:35:51 <andythenorth> ugh git push -f is going to get me killed at work
22:36:00 <andythenorth> one day I'll forget and do it by accident
22:36:06 <TrueBrain> it also means they have a bad pre-commit filter :D
22:36:18 <TrueBrain> so you can always just say it was their fault :P
22:36:26 <andythenorth> well it's my company
22:36:31 <andythenorth> so it's always my fault, ultimately
22:37:07 <TrueBrain> I wonder if we should make the GPL stuff more verbose, like many other projects
22:37:11 <TrueBrain> so much administration
22:37:33 <andythenorth> we need planetmaker or someone
22:37:44 <TrueBrain> there are bots that do that work :)
22:38:00 <TrueBrain> very nice work on the CONTRIBUTION.md
22:38:00 <__ln__> elfring: Please make concrete suggestions about improvements in the form of pull requests.
22:38:02 <TrueBrain> I really like it :)
22:38:14 <andythenorth> TrueBrain: team game eh
22:39:17 <TrueBrain> also nicely done with credits where credits are due :)
22:39:27 <TrueBrain> and your first successful rebase \o/ :D
22:40:04 <TrueBrain> did frosch123 also already read it? :)
22:40:22 <TrueBrain> before I go: merge, andhe goes: WHAT THE FUCK :D
22:40:31 <TrueBrain> (he never does that; but I like to think I would :P)
22:41:05 <elfring> __ln__: How are the chances for pull requests around better exception handling?
22:41:35 <andythenorth> frosch gave comments
22:42:42 <TrueBrain> LordAro: does the fix you showed also work for this GCC 7 bug? Or are those 2 different things?
22:43:12 <LordAro> TrueBrain: it does, but frosch123 didn't think it was the Right Thing(tm) to do
22:43:23 <LordAro> (the other commit is the clang bug)
22:46:54 <TrueBrain> ah, so yeah, the values makes sense
22:47:03 <TrueBrain> 256 bytes between the two values :)
22:49:20 <TrueBrain> it seems GCC is blind for the if statement just before
22:51:46 <TrueBrain> well, if frosch123 is commenting, I also have one more andythenorth :D
22:52:09 <andythenorth> now I now how to push -f :P
22:52:23 <TrueBrain> and git commit --amend? :D
22:53:02 <andythenorth> how does that work in the PR though?
22:53:12 <andythenorth> i.e. what happens to your comments?
22:53:41 <TrueBrain> and you can still see our comments, jus tnot in that commit anymore
22:53:54 <TrueBrain> you make a new commit
22:53:59 <TrueBrain> so you can see the diff there
22:54:02 <TrueBrain> honestly, works as well
22:54:07 <TrueBrain> just who accepts your PR needs to squash
22:54:15 <TrueBrain> but .. back to the squash vs rebase dialog :D
22:57:44 <TrueBrain> @base 10 16 18446744073709551361
22:57:44 <DorpsGek> TrueBrain: FFFFFFFFFFFFFF01
22:57:48 <TrueBrain> @base 10 16 18446744073709551615
22:57:48 <DorpsGek> TrueBrain: FFFFFFFFFFFFFFFF
22:59:22 <andythenorth> do I need to comment on PR at all?
22:59:51 <TrueBrain> emails go out, and if you say: frosch123, I am sure you just did :P
23:00:57 <TrueBrain> so GCC think the range is between -1 and -254?
23:01:03 <TrueBrain> so length is always 0?
23:01:42 <frosch123> TrueBrain: no Werror then :)
23:02:08 <TrueBrain> I wonder how it makes that conclusion
23:02:43 <TrueBrain> as two if statements clearly show it is not possible for length to be 0
23:05:19 <TrueBrain> guess optimizations play a part there
23:06:20 <TrueBrain> I would need to install GCC 7 to debug that further .. just weird :P
23:07:06 <TrueBrain> possibly the warning is just out of context
23:07:12 <TrueBrain> because if you have 2 bytes A and B
23:07:26 <TrueBrain> than A - B can range from 255 to -254
23:07:33 <TrueBrain> and maybe it only reports on the -1 .. -254
23:08:15 <TrueBrain> given length - statspec->lengths is used twice
23:08:25 <TrueBrain> putting it in a variable both solves the issue as makes the code prettier
23:08:29 <TrueBrain> sadly, LordAro went for the cheap solution :D
23:09:16 <TrueBrain> would also explain the memset a bit lower; as multiplying it with a few puts the number back into what "is allowed"
23:09:21 <TrueBrain> by what-ever stupid definition
23:09:24 <TrueBrain> it is just a bogus warning
23:13:18 <LordAro> i'll see about fixing it more properly
23:14:25 <TrueBrain> do you have GCC 7 LordAro?
23:14:38 <TrueBrain> can you check if 6733 still warns? :)
23:17:17 *** sim-al2 has joined #openttd
23:18:01 <TrueBrain> I did something that broke compiling locally :D
23:19:11 <TrueBrain> ah, something added a wrong include ..
23:19:25 <TrueBrain> tnx LordAro; now I just hope frosch123 does like my solution :D
23:20:14 <TrueBrain> THERE! A CONTRIBUTING.md :D
23:23:36 <frosch123> did you force push to fake the pr number into the message? or did you anticipate the next number?
23:24:22 <TrueBrain> hm? 6729 is the bug report, not?
23:24:27 <TrueBrain> 6733 is the fix? What am I missing?
23:26:31 <TrueBrain> okay .. going to try eints ...
23:27:25 <TrueBrain> it wants validation of Jenkins :D
23:27:37 <TrueBrain> I once wrote a bot which made the PR
23:27:40 <TrueBrain> waited for approval
23:27:50 <TrueBrain> but ... that is a lot of shit :P
23:29:11 <TrueBrain> now just administrators can ignore all requirements .. so behave frosch123 :P
23:29:23 <frosch123> he, just wanted to suggest unchecking that :p
23:29:37 <TrueBrain> just a bit unclear who 'administrators' are
23:30:11 <TrueBrain> it is called owner in most places
23:30:15 <TrueBrain> I ASSUME they mean the same
23:30:20 <TrueBrain> but ...... why call it differently?
23:30:57 <TrueBrain> okay, nightly commit is active again
23:31:28 <peter1138> That reminds me... I had to make my jenkin user a gitea admin before the API would work correctly. I'm sure that's not right :S
23:31:36 <LordAro> TrueBrain: worth pointing out that GH doesn't see "Codechange #nnnn" as a close marker
23:31:48 <andythenorth> such CONTRIBUTIONS expected now
23:31:57 <TrueBrain> LordAro: yup, I noticed :)
23:33:20 <TrueBrain> almost hit Squash, till I noticed LordAro approved; seems that is now enough for GH as Owner :D
23:33:51 <TrueBrain> hmm .. its green here :)
23:34:43 <TrueBrain> hmm .. that flag has nothing to do with it
23:34:52 <TrueBrain> seems Gardeners can also review :D
23:34:53 <frosch123> but when you try to confirm without approval or ci check, you have to confirm to abuse owner powers
23:35:14 <TrueBrain> owh, LordAro is now a Code Owner of course, because he made a contribution :)
23:35:28 <andythenorth> I can't merge PRs
23:35:34 <andythenorth> "You're not authorized to merge this pull request."
23:35:44 <TrueBrain> merge is restricted via another toggle
23:35:50 <andythenorth> I can comment and/or close
23:35:50 <TrueBrain> GitHub works in mysterious ways :)
23:36:41 <TrueBrain> I can dismiss reviews
23:36:44 <TrueBrain> but you need to tell WHY
23:37:38 <TrueBrain> happy eints 'just worked', really tnx frosch123 :)
23:38:28 *** gelignite has joined #openttd
23:41:09 <TrueBrain> anyway, frosch123, is #6733 acceptable, or do you want me to dig deeper why GCC does this wrong?
23:41:34 <frosch123> gcc will fix it somewhen
23:41:52 <frosch123> i think we always had weird warnings with some compilers
23:42:03 <frosch123> some got fixes with later compilers, some we disabled
23:42:09 <TrueBrain> yeah; that is why I tried to find something that both benefits the code, and solves the issue :)
23:42:35 <TrueBrain> but compilers indeed will be compilers .. so looking deeper is not really useful indeed
23:42:49 <TrueBrain> I remember the fixes during GCC4 we had to do :D
23:44:33 <TrueBrain> lot of cool new warnings in GCC 7 .. like showing which part of a printf is wrong :D
23:47:58 <frosch123> don't dismiss again :p
23:48:45 <frosch123> also: earlier on the contributing pr
23:48:55 <LordAro> i added -Weffc++ to see what happened
23:48:58 <frosch123> you had requested changes and i had approved, but merge was still blocked
23:49:10 <TrueBrain> frosch123: that is at least sane
23:49:12 <frosch123> so, request changes overrules approval :)
23:50:37 <TrueBrain> poor CI, has been busy for over an hour now :P
23:50:41 <TrueBrain> I really need to fix ccache
23:51:01 <TrueBrain> or leave a partial compiled or so
23:51:36 <TrueBrain> or AWS needs to reply with a nice offer :D
23:51:50 <TrueBrain> ugh, the OSX build is only at 55% ... this is taking a REALLY long time
23:51:51 <peter1138> I need to move mine, it's running on a Celeron at the moment.
23:52:09 <peter1138> What's it doing, bootstrapping a compiler?
23:52:23 <TrueBrain> no, that he already did
23:52:28 <TrueBrain> now he is compiling the compiler or something
23:52:43 <TrueBrain> wow ... flush is taking 300% CPU
23:52:47 <LordAro> TrueBrain: partial compiling is risky
23:53:36 <TrueBrain> why is it doing so much I/O for this .. holy crap ..
23:53:39 <TrueBrain> did not expect that ...
23:54:22 <TrueBrain> LordAro: our dependency is pretty good
23:54:24 <Wolf01> About partial compiling, with the new version of the tool I use to develop at work, they allow partial compiling of classes to solve the problem of circular dependencies -.-
23:54:36 <TrueBrain> only issue might be mtime
23:54:48 <TrueBrain> Wolf01: sounds scary :D
23:55:07 <Wolf01> Sounds insane, not scary
23:55:35 <LordAro> maybe make || (make mrproper && make) could work?
23:56:05 <TrueBrain> LordAro: that is the case in our current build system
23:56:15 <frosch123> btw, did someone from simutrans already complain?
23:56:16 <TrueBrain> but no need to test that every commit
23:56:29 <TrueBrain> and I kinda forgot about them
23:56:38 <LordAro> why would simutrans complain?
23:56:45 <TrueBrain> will write a message tomorrow telling them what was going on
23:57:09 <TrueBrain> hmm ... I closed my docker build, and flush is still at 300%
23:57:50 <peter1138> Last Simutrans build is 27th Feb, so...
23:57:57 <TrueBrain> when was their last commit?
23:57:59 <peter1138> I guess they don't change very much.
23:58:12 <LordAro> that's not *that* long ago...
23:59:09 <peter1138> Hm, according to the git mirror, yesterday.
23:59:22 <TrueBrain> so we are not compiling that version :D
continue to next day ⏵
