IRC logs for #openttd on OFTC at 2018-03-18
⏴ go to previous day
00:00:36 *** JetFly|2 has joined #openttd
00:19:30 *** Supercheese has joined #openttd
00:38:29 *** Biolunar has joined #openttd
00:54:47 *** SeraphAngel has joined #openttd
01:20:42 *** FLHerne has joined #openttd
02:07:25 *** Supercheese has joined #openttd
02:09:11 <GT> Just returned to OTTD after a long time off. What is the best Train AI available?
02:09:42 <Eddi|zuHause> you mean the least worst?
02:10:02 <Eddi|zuHause> tbh, i have no idea, i never use them
02:11:44 <GT> Well, sometimes I do, but always end up a bit disappointed
02:12:54 <GT> Basic funtionality is there, but no AI shows any higher level of intelligence, like building a nice network,
03:01:25 *** Thedarkb has joined #openttd
06:37:31 *** Supercheese has joined #openttd
07:33:01 *** Alberth has joined #openttd
07:33:01 *** ChanServ sets mode: +o Alberth
07:54:28 *** sla_ro|master has joined #openttd
08:34:43 *** sla_ro|master has joined #openttd
08:48:42 *** Progman has joined #openttd
09:27:35 <nielsm> did the source migration to github get finished? or still some things to do?
09:31:10 <Alberth> I'd be surprised if it was completely finished :)
09:32:09 <Alberth> moving the source history to GH is just one of the many steps, I think
09:33:22 *** Biolunar has joined #openttd
09:35:40 *** Wacko1976 has joined #openttd
09:37:05 <LordAro> the source itself was finished copying
09:37:14 <LordAro> but still for testing purposes only
09:38:53 *** Biolunar has joined #openttd
09:38:57 *** HerzogDeXtEr has joined #openttd
10:09:28 *** HerzogDeXtEr has joined #openttd
10:24:52 *** synchris has joined #openttd
10:28:09 *** FLHerne has joined #openttd
10:37:34 *** andythenorth has joined #openttd
10:43:32 <andythenorth> randomly reversed engine sprites
10:43:47 <andythenorth> currently I duplicate the spritesheet and move the sprites manually
10:43:53 <andythenorth> I could automate that with PIL
10:44:21 <andythenorth> or I could have different action 1 for reversed
10:44:47 <andythenorth> PIL slows down the compile
10:45:01 <andythenorth> doing it in action 1 requires special cases that make the compile and nml more complicated
10:52:12 <andythenorth> doing it manually is no work at all right now :P
10:59:39 * andythenorth wonders why vehicle_is_flipped isn't documented in nml
11:10:05 <Alberth> var_variables.py also says it's C8 :)
11:13:24 <Alberth> you could also hack nml to reverse the sprite :p
11:16:45 *** FLHerne has joined #openttd
12:06:02 *** Wacko1976 has joined #openttd
12:21:57 *** sla_ro|master has joined #openttd
12:22:32 <andythenorth> I wonder if it was just missed in docs
12:22:35 <andythenorth> and if I should fix it
12:27:46 *** Wormnest has joined #openttd
12:42:29 <andythenorth> I'll test it works first :P
12:49:39 *** ToffeeYogurtPots has joined #openttd
13:24:14 <TrueBrain> hmmm .. I cannot make a private repo inside the OpenTTD space .. at least, that costs money :P
13:25:11 *** ToffeeYogurtPots has quit IRC
13:26:34 *** ToffeeYogurtPots has joined #openttd
13:26:34 *** dustinm` has joined #openttd
13:26:34 *** Smedles_ has joined #openttd
13:26:34 *** Maarten has joined #openttd
13:26:34 *** ccfreak2k has joined #openttd
13:26:34 *** TinoDidriksen has joined #openttd
13:26:34 *** Extrems has joined #openttd
13:26:34 *** greeter has joined #openttd
13:26:34 *** Warrigal has joined #openttd
13:26:34 *** mikegrb has joined #openttd
13:26:34 *** Tharbakim has joined #openttd
13:26:34 *** techmagus has joined #openttd
13:26:34 *** HeyCitizen has joined #openttd
13:26:34 *** nahkiss has joined #openttd
13:28:55 <andythenorth> TrueBrain: we have donations? :P
13:31:41 <TrueBrain> it is only 25 dollar yes :P But I have no clue if it solves my issue :D
13:47:54 <TrueBrain> yippie, I made my issue migration resumable
13:48:04 <TrueBrain> so now if github barks, I can just continue it at a later time
13:48:30 <TrueBrain> and I can pre-import, making FlySpray readonly for a much smaller amount of time :)
13:49:20 <TrueBrain> 275 lines of code btw ... lol
13:50:23 <LordAro> can you disable issue creation on the github side while you're doing it?
13:50:31 <TrueBrain> there are 14k comments and 7k tasks .. of which most are closed... that makes 21k comments (including close comments), 7k create issue, en 7k close issue .. so 35k calls
13:50:38 <TrueBrain> I can make 1 every 3 seconds
13:50:56 <TrueBrain> @calc 35000 * 3 / 3600
13:50:56 <DorpsGek> TrueBrain: 29.1666666667
13:51:00 <TrueBrain> 29 hours for the initial import
13:51:34 <TrueBrain> the only thing this import script does not track, is reopening of a ticket
13:51:59 <TrueBrain> and also for example a comment made after closing
13:52:02 <TrueBrain> is in the wrong order of time
13:54:34 *** sim-al2 has joined #openttd
13:56:33 *** Thedarkb has joined #openttd
13:58:16 *** SeraphAngel has joined #openttd
14:00:13 <Alberth> it's not by definition wrong :p
14:06:20 <andythenorth> not sure how much history has to be perfect
14:06:25 <andythenorth> we're not going back in time :P
14:12:52 <TrueBrain> okay .. so resumable issue import is running on a 3 second delay .. lets see if that avoids the Abuse stuff
14:13:01 <TrueBrain> and we will check in 24 hours if all issues are imported :P
14:15:36 *** ChanServ sets mode: +v tokai
14:16:00 <_dp_> hi, are you finally moving to github? :)
14:16:08 <_dp_> been out of the loop for a while ^^
14:19:15 <andythenorth> TB has pulled the pin
14:19:18 <andythenorth> well not quite yet
14:20:01 <TrueBrain> frosch123 did in fact; I am just a puppet
14:20:08 <_dp_> well, I guess it's still something
14:21:03 <_dp_> I completely missed 1.7.2 so kinda waiting for it now :)
14:36:35 <andythenorth> peter1138: any way to reverse first vehicle of consist when it's purchased? Articulated CB can reverse trailing units, but not first unit I think.
15:02:03 *** supermop has joined #openttd
15:02:28 *** SeraphAngel has joined #openttd
15:20:36 *** Stimrol has joined #openttd
15:21:55 <supermop> still dont get this:
15:22:18 <supermop> if i have a blue truck, and want it to have a random color
15:22:32 <supermop> where do those palette names go
15:22:53 <supermop> in the definition of the sprite group? in some switch?
15:23:42 <supermop> sorry if i;m being dumb but there is no little code example for recoloring on the wiki so im clueless
15:24:53 <supermop> nfi how to say 'use this spritegroup with this palette'
15:29:34 <andythenorth> I am literally looking at that in Horse right now
15:29:44 <andythenorth> due to removing the old, pixel-processing way
15:30:30 <andythenorth> supermop: you've found the 'colour_mapping' cb yes?
15:57:15 *** Thedarkb has joined #openttd
16:02:25 *** frosch123 has joined #openttd
16:08:21 *** Wacko1976_ has joined #openttd
16:13:03 *** gelignite has joined #openttd
16:29:04 <TrueBrain> ah, github replied; there is another API to use for this stuff
16:31:32 <TrueBrain> just annoying I couldnt find that myself :P
16:32:07 <frosch123> does it have other restrictions?
16:32:13 <frosch123> or what is the idea behind two apis?
16:32:32 <TrueBrain> basically, what I expected, the current API triggers emails
16:32:37 <TrueBrain> so they dont want you to spam it like I do
16:32:42 <TrueBrain> as it might trigger a lot of emails
16:32:54 <TrueBrain> of course, I dont care about notifications, so I want an API where that is disabled
16:32:59 <TrueBrain> they have one (not official yet, basically)
16:33:06 <TrueBrain> meaning you have less restrictions
16:34:37 <andythenorth> do trains have some user bits I can use?
16:34:41 <TrueBrain> it also allows me to import comments at once; so 1 call for a complete issue
16:34:53 <andythenorth> I want to reliably know which sprite a vehicle has randomised
16:35:05 <andythenorth> or maybe I can depend on the random reliably
16:35:40 <TrueBrain> hmm, this API suggests that you can create tickets with older timestamps
16:35:42 <TrueBrain> that is interesting
16:35:52 <andythenorth> it's for migrations?
16:36:04 <TrueBrain> so yes, that includes migrations
16:36:26 <TrueBrain> main difficulty is that it is async
16:36:35 <TrueBrain> so you create an issue, then you have to poll if it was created
16:37:11 <TrueBrain> so time to fiddle with this!
16:38:16 <TrueBrain> means I have to rebuild my script a bit ... owh boy :D
16:44:12 <TrueBrain> sadly, the import API is a one-shot API
16:44:29 <TrueBrain> that means putting FlySpray in read-only and then migrating
16:44:42 <TrueBrain> would take ~2 hours to do the migrations, by first estimates
16:44:56 <TrueBrain> but okay .. lets first get it to work :)
16:44:58 <LordAro> why do you need to put it in readonly mode?
16:45:26 <TrueBrain> its a one-shot import; so if someone would modify FlySpray after the ticket moved to GitHub, there is no way to fix that
16:45:31 <nielsm> so nothing new comes in while he's reading, is my guess
16:46:03 <LordAro> well i thought we were still testing?
16:46:16 <TrueBrain> how does that matter?
16:46:21 <TrueBrain> there is a point in time we are no longer testing?
16:46:49 <nielsm> flyspray will probably need to stay read-only after the import regardless
16:46:50 <TrueBrain> its called: thinking ahead :D
16:46:58 <LordAro> well it doesn't really matter if someone modifies flyspray while doing a test import
16:47:03 <TrueBrain> but so there is a window where you cannot create a bug, basically :)
16:47:08 <LordAro> i thought the repo would be deleted and recreated
16:47:12 <TrueBrain> no, LordAro, but now think ahead
16:47:17 <TrueBrain> if this is no longer a test
16:47:37 <TrueBrain> in general I try to avoid things that need to happen ina big-bang
16:47:42 <TrueBrain> I rather have incremental stuff
16:47:55 <TrueBrain> and although we are testing now, I am thinking about how the result would be
16:48:02 <TrueBrain> as I am testing the result process, not the endpoint
17:01:27 <TrueBrain> owh joy, a lovely API error I cannot parse :D
17:02:40 <TrueBrain> ha! Closed on Oct 2005 :D
17:03:37 <supermop> andythenorth: no, is that for the whole item or per sprite
17:03:55 <andythenorth> for the whole vehicle
17:04:04 <andythenorth> but it goes to a switch chain
17:04:13 <andythenorth> so you could probably isolate it somewhat
17:04:40 <supermop> so a flatbed with two subarus on it could paint one red and one blue?
17:05:24 <andythenorth> what colour are the subarus in the spritesheet?
17:06:14 <TrueBrain> its funny, adding the labels is recorded on thecurrent timestamp
17:06:18 <TrueBrain> I cannot influence that part :D
17:06:32 <andythenorth> ABANDON ALL HOPE
17:10:32 <TrueBrain> I need ~3 requests per issue now .. so that is 21k requests .. needs 5 hours
17:10:36 <TrueBrain> ugh .. 5 hours of read-only ..
17:11:07 <LordAro> can you not predownload everything from flyspray?
17:11:28 <TrueBrain> I don't understand your mindjump there?
17:11:35 <LordAro> where's the bottleneck?
17:11:43 <TrueBrain> 5k API requests per hour
17:12:20 <LordAro> and you have to get the information from flyspray before creating an issue out of it?
17:12:44 <TrueBrain> your mind made a jump that I cannot follow, so please explain what you are trying to resolve :)
17:12:45 <LordAro> so why not get all the information at once, store it somewhere
17:12:51 <TrueBrain> instead of continuing to lose me :D
17:12:53 <LordAro> then upload it to github
17:12:58 <TrueBrain> how does that solve anything?
17:13:01 <LordAro> less readonly flyspray
17:13:21 <LordAro> *unless* you're talking about the actual migration, rather than what you're doing currently#
17:13:26 <TrueBrain> okay, let me break it down for you again (as you clearly were the one missing the train somewhere):
17:13:31 <TrueBrain> WHEN we are going to do the real migration
17:13:34 <TrueBrain> this is what happens:
17:13:39 <TrueBrain> we put FlySpray in read-only
17:13:44 <TrueBrain> we start the migration
17:13:48 <TrueBrain> issues arrive at GitHub
17:13:53 <TrueBrain> if ALL issues are migrated
17:13:59 <TrueBrain> people can make new issues at GitHub
17:14:21 <TrueBrain> again, I am not talking about this testing; I am talking about the process to get to the end result
17:14:24 <TrueBrain> I am testing that process
17:14:44 <LordAro> yeah, that makes sense
17:14:48 <LordAro> i was thinking in test mode only
17:14:51 <TrueBrain> now my testing shows this "migration" part takes ~5 hours
17:15:10 <TrueBrain> which is a large amount of time to now need anyone creating issues on either system
17:15:31 <_dp_> to me 5 hours sounds like something no one would even notice xD
17:16:17 <_dp_> there are like what, 1-2 new bugs a week
17:16:42 <TrueBrain> main issue, I cannot put Create New Issue on GitHub in read-only
17:16:48 <TrueBrain> so I really need people to stay away from there :P
17:17:02 <TrueBrain> if someone does make an issue .. I have to remove the project, and start over
17:17:26 <TrueBrain> another annoying issue, as the API is async, I cannot just schedule all imports at once
17:17:38 <TrueBrain> as then the numbers dont have to line up
17:18:17 <TrueBrain> and it is mostly due to this: when you import an issue, you get an URL back to poll the status of the import
17:18:23 <TrueBrain> you can do 5000 API calls an hour
17:18:30 <TrueBrain> the polling of the status of the import, is included in that number
17:20:00 <frosch123> which of the 3 requests per issue defines the number?
17:20:21 <TrueBrain> when the import is done, it is assigned an issue number
17:20:24 <TrueBrain> up till then, it can be what-ever
17:20:49 <frosch123> hmm, so you cannot run ahead with the next one
17:21:15 <TrueBrain> so maybe I can poll the site via another way that doesnt count towards my request per hour :D
17:22:52 <LordAro> i feel like i've seen github projects with issues disabled
17:22:59 <LordAro> but maybe whatever that is disables import as well
17:22:59 <TrueBrain> you can disable it, sure
17:23:03 <TrueBrain> but .. how do I import ? :D
17:23:06 *** ToBeFree has joined #openttd
17:23:09 <TrueBrain> (I tried, of course)
17:23:40 <frosch123> we could rebrand the project secretly, so noone would find it
17:23:52 <TrueBrain> that would work fine, yes :)
17:23:57 <TrueBrain> but I still dont like it .. 5 hours is a long time
17:25:59 <TrueBrain> forgot to remove 1 timestamp
17:26:03 <LordAro> could get rid of- yes
17:28:53 <supermop> andythenorth: they are blue of course
17:28:59 <_dp_> does it insert issue text as a markdown? could break something
17:29:07 <supermop> as subarus should be
17:29:07 <andythenorth> supermop: so that is harder :)
17:30:56 <supermop> could make them purple
17:31:51 <supermop> but still would like stacked sprites to get differing recoloring but doesn't look possible
17:32:17 <andythenorth> I'm not sure it's easy
17:35:33 <_dp_> that "comment imported" remark is unnecessary imo, one in issue is enough
17:36:09 <_dp_> And it's pretty clear which ones are imported anyway
17:36:14 <TrueBrain> I think it is very important over time to have that
17:36:24 <TrueBrain> as now everyone remembers
17:36:27 <TrueBrain> but in 1 year, looking back
17:36:36 <TrueBrain> it is pretty nice to see which comments were imported, and which were not
17:37:28 <TrueBrain> lol @ import API .. I just made a ticket 1 hour in the future :D
17:39:22 <_dp_> just having the link in the first comment is enough imo
17:39:41 <TrueBrain> there will be tickets which have both worlds: imported comments and GitHub comments
17:39:45 <TrueBrain> so I disagree that it is enough :)
17:40:59 <_dp_> you mean your bot will import add github ones as well?
17:41:37 <TrueBrain> but what I mean: issue is imported, and someone comments on it in GitHub
17:41:45 <TrueBrain> so much more clear if every comment states what its origin is
17:41:51 <TrueBrain> removes any form of confusing
17:41:54 <TrueBrain> not now, but in 1 year
17:42:17 <_dp_> well, if someone comments it will look like a normal comment, not DorpsGek one
17:42:18 <frosch123> _dp_: it's not about creating an archive mirror on github
17:42:31 <frosch123> in future people will report on github, no longer on flyspray
17:42:53 <andythenorth> flyspraty will die :)
17:42:57 <_dp_> you can also make smth like "matthijs wrote" a link to that comment
17:43:08 <frosch123> anyway, since the links are specific to comments it is clearly more useful
17:43:15 <TrueBrain> this is a lot more clear
17:43:55 <_dp_> yeah, but makes it much harder to read the thread
17:44:35 <TrueBrain> its difficult no matter what, as extra lines are inserted that dont add to the conversation itself (the whole "blabla wrote: " for example)
17:46:23 <_dp_> nah, that hr and link really stands out
17:46:32 <_dp_> at least make it small or smth
17:47:34 <_dp_> it's just I can't even imagine why would someone want to know where every comment came from
17:47:56 <TrueBrain> I think this is the point where the only thing we can do is to conclude: lets agree to disagree :)
17:47:58 <_dp_> and even if for some weird reason he does it will still be pretty clear
17:48:04 <TrueBrain> and I cannot make it smaller; at least, I cannot find a markdown way of doing that
17:48:11 <_dp_> but doesn't justify adding huge remark to every comment
17:50:34 <_dp_> use <sub></sub> or ######
17:50:36 <TrueBrain> okay, I tuned it the best I could towards 5000 request/h .. still don't fancy how long this is going to take ..
17:54:02 <TrueBrain> adding some # works .. at a certain point the boldness it adds no longer stands out
17:56:16 <_dp_> yeah, h6 even has gray text
17:56:25 <_dp_> kinda questionable semantically though
17:56:38 <_dp_> I think I like <sub> more
17:57:43 <TrueBrain> I am a bit surprised that that works given it is markdown
18:00:15 <_dp_> markdown parsers usually accept some html
18:01:10 <TrueBrain> frosch123: I think what we will do before we migrate, is already import all closed tickets for more than a month ago
18:01:17 <TrueBrain> that should heavily reduce the read-only time
18:03:21 <frosch123> but increases the time where people can add stuff in github
18:03:50 <TrueBrain> I think the "hiding" of the project will work just fine
18:03:56 <LordAro> won't you only be able to add stuff up to the first open issue?
18:04:06 <TrueBrain> LordAro: good point
18:04:57 <TrueBrain> okay, so I will do a test-run than to check the real time we need
18:06:54 <TrueBrain> okay .. so I will check back in 5 hours to see how this did :)
18:07:02 <TrueBrain> pretty happy timestamps are kept now tbh :)
18:07:13 <TrueBrain> still not sure about the category label
18:07:32 <frosch123> a specific one, or in general?
18:07:52 <TrueBrain> easier to remove than to add :)
18:21:32 <andythenorth> not sure if vehicle_is_flipped works in nml
18:21:48 <andythenorth> might be why it's not documented
18:22:44 <andythenorth> it should return FD if vehicle is flipped
18:22:52 <andythenorth> I can't find any magic to handle that in nml
18:23:41 <nielsm> hmm, seems there are in fact some bugs with my midi file reading, trying it out with the scott joplin music pack on bananas right now, and several of the songs seem to break it
18:23:56 <andythenorth> how do I mask var[0xC8] to get 1 if FD otherwise 0?
18:24:22 <andythenorth> or at least something I can use in nml :P
18:24:27 *** johnwhitlow[m] has quit IRC
18:25:06 *** synchris has joined #openttd
18:25:22 *** Mahjong2 has joined #openttd
18:25:22 *** TheMask96 has joined #openttd
18:25:22 *** Hirundo has joined #openttd
18:25:22 *** Hazzard has joined #openttd
18:25:22 *** michi_cc has joined #openttd
18:25:22 *** Agiri[m] has joined #openttd
18:25:22 *** Antheus has joined #openttd
18:25:22 *** grossing has joined #openttd
18:25:22 *** TrueBrain has joined #openttd
18:25:22 *** Alkel_U3 has joined #openttd
18:25:22 *** johnwhitlow[m] has joined #openttd
18:25:22 *** Sheogorath has joined #openttd
18:25:22 *** Terkhen has joined #openttd
18:25:22 *** V453000 has joined #openttd
18:25:22 *** planetmaker has joined #openttd
18:25:22 *** liquid.oftc.net sets mode: +vvoo tokai michi_cc Terkhen planetmaker
18:25:34 *** ChanServ sets mode: +v Terkhen
18:25:34 *** ChanServ sets mode: +v planetmaker
18:27:22 <frosch123> andythenorth: pretty sure vehicle_is_flipped would return 0 or 1 in nml
18:27:37 <andythenorth> must be EBKAC then
18:28:32 <frosch123> i guess none dared to document the different between flipped and reversed
18:29:11 <andythenorth> I have the same issue in my code :P
18:29:34 <andythenorth> there is *no* way to set reversed on a vehicle in a non-articulated consist? o_O
18:29:42 <frosch123> so maybe flipping is missnig because someone thought reversed is the same
18:29:53 <andythenorth> or maybe it was the time of flipgate
18:29:57 <andythenorth> remember it was removed?
18:29:59 <frosch123> "reversed" is a consist proprty
18:30:03 <frosch123> fliipped is per vehicle
18:30:47 <andythenorth> can a single-vehicle consist be reversed?
18:31:13 <andythenorth> actually it doesn't matter :)
18:31:15 <andythenorth> I was just curious
18:31:19 <frosch123> consists reverse in terminus stations
18:32:04 <andythenorth> ok so there's the other 'reversed'
18:32:15 <andythenorth> which is adding 4000 when returning articulated cb results
18:32:28 <andythenorth> I assumed that reversed the vehicle?
18:40:45 <frosch123> i see a lot of train magic in the code
18:40:52 <frosch123> may be broken for road vehicles
18:41:40 <andythenorth> it's trains I was thinking about
18:41:48 <andythenorth> I think I'll just leave the magic in nml :P
18:41:55 <andythenorth> it's fairly horrible, but eh
18:42:02 <frosch123> vehicle_is_flipped returns 0 or 1
18:42:11 <frosch123> and yes, it is undocuented for some reason
18:44:11 <andythenorth> does flipping also flip visual effect?
18:44:28 <andythenorth> I might be trying to solve the wrong thing :P
18:44:43 <frosch123> that's one of those things which may be different between trains and rv :p
18:45:17 *** SeraphAngel has joined #openttd
18:45:22 <frosch123> some places in ottd use spritenum, some other places use VRF_REVERSE_DIRECTION
18:45:28 <frosch123> so it is likely inconsistent in various cases
18:46:06 <andythenorth> seems for visual_effect_and_powered, flipped is understood
18:46:22 <andythenorth> so I can eliminate some horrible nml :P
18:46:54 <andythenorth> twice as complicated as needed :)
19:03:14 *** SeraphAngel has joined #openttd
19:15:30 <TrueBrain> we still have open bugs dating 2007? :o
19:19:23 <frosch123> the oldest valid one is from 2010
19:19:33 <LordAro> quick, set andy on it
19:20:43 <LordAro> oh, only "enhancement"
19:20:51 <andythenorth> earliest bug is 2010
19:21:50 <LordAro> can't see it on the original, nor has it been closed
19:24:07 <TrueBrain> lol .. variable was not reset
19:24:10 <TrueBrain> is value of issue before him
19:24:16 <TrueBrain> I didnt have open issues earlier
19:25:05 <TrueBrain> @calc 78 * 60 / 1677
19:25:05 <DorpsGek> TrueBrain: 2.79069767442
19:25:10 <TrueBrain> ~3 seconds per issue
19:25:19 <TrueBrain> @calc 3 * 7000 / 3600
19:25:19 <DorpsGek> TrueBrain: 5.83333333333
19:26:30 <glx> today was a little more spamy than yesterday ;)
19:27:25 <TrueBrain> creating/removing of projects, yes :)
19:28:01 <glx> but could have been worse
19:39:24 <nielsm> this is a really spooky bug I have... some songs cause my music driver to go into intensely slow mode, so I just get one new chord every 20 seconds or something
19:41:11 <nielsm> I'll have to check the tempo calculations
19:41:18 <glx> like a 16bit number becoming negative
19:41:54 <nielsm> probably an uint32 holding microseconds overflowing
19:42:48 <andythenorth> what's the antonym of 'reversed' for a vehicle?
19:43:01 <andythenorth> '', 'forwards', or 'unreserved'
19:43:15 <andythenorth> reversed / reserved /s
19:43:32 <andythenorth> 'unreversed' will do then
19:43:53 <glx> looks like a double negation
19:44:53 <andythenorth> it's clunky, which is why I wondered about ''
19:45:01 <andythenorth> but that's not explicit in places that confuse me
19:45:48 <DorpsGek> Commit by translators :: r27994 trunk/src/lang/welsh.txt (2018-03-18 19:45:38 +0100 )
19:45:50 <DorpsGek> welsh: 2 changes by kazzie
19:47:26 <glx> and it's in rail transport context
19:53:27 <nielsm> huh no doesn't look like an overflow at this point
19:53:55 <nielsm> the last few blocks of note data in the song, the realtime value is microseconds since start of song, i.e. 1 million = 1 second
19:54:19 <nielsm> and the last few have 20-40 seconds between each block
19:58:21 *** Thedarkb has joined #openttd
20:08:31 <nielsm> um but lol these scott joplin tunes are pretty crazy
20:08:41 <nielsm> this one has 320 tempo change events in it
20:10:16 <LordAro> sounds about right for Joplin :p
20:33:30 *** supermop has joined #openttd
20:36:32 <TrueBrain> 2700 issues done .. this is soooo slllooowwww :D
20:37:44 <andythenorth> faster than painting pixels
20:37:54 <andythenorth> I spent 2 days drawing 1 boat hull :P
20:39:09 <TrueBrain> it seems it no longer creates issues ...
20:40:25 <supermop> oh looks like different sprites in a stack can have different recolorings
20:40:41 <supermop> it says right there in the wiki and i never realized it
20:42:11 <andythenorth> you check them during the cb?
20:42:19 <TrueBrain> wow, it is just really really slow ..
20:42:44 <supermop> i think you enumerate for each iteration of the stack?
20:43:06 <supermop> not sure because my stacks are written a little mote simply
20:44:16 <nielsm> TrueBrain you should take the opportunity to savor this once-in-a-lifetime experience, don't hurry it
20:45:42 <LordAro> once-in-15-years experience, anyway
20:45:55 *** synchris has joined #openttd
21:02:34 *** sim-al2 has joined #openttd
21:02:51 <nielsm> yeah this seems to fix the tempo issues
21:09:04 <nielsm> hmm would be neat to add an option to music baseset definitions to change the title of the jazz jukebox :)
21:09:18 <nielsm> so it could say "Player Piano" instead for the joplin tunes
21:40:42 <Alberth> everything refactored?
21:41:17 <andythenorth> do you mean 'did I touch everything'?
21:41:25 <andythenorth> or 'is it done'? :P
21:42:39 <frosch123> the farm only builds two projects in parallel
21:42:45 <frosch123> andy is pushing to 3 at once :)
21:43:58 <andythenorth> I have got a shared library :P
21:44:06 <andythenorth> update one place, 3 projects change :P
21:44:21 <andythenorth> new ways to regret choices later
21:44:58 <andythenorth> where is Eddi to tell me I should have worked out my spec first?
21:45:23 <glx> you won't listen anyway ;)
21:46:27 <Alberth> stuck in the snow, probably
21:47:06 <andythenorth> I have no problem working out the spec first
21:47:07 <Alberth> so working on one newgrf at a time wasn't enough :p
21:47:15 <andythenorth> I just think the spec should be written in code
21:47:37 <supermop> i made a pickup truck that changes color based on cargo age
21:48:03 <supermop> seems to work, but color only currently updates if you stop the vehicle
21:48:04 <Alberth> writing a spec as code is something I try too sometimes, but it never works out
21:48:16 <supermop> not while it is moving on its own
21:48:22 <andythenorth> the default seems to be Excel documents
21:48:42 <andythenorth> "Item 3.1.2: the system must accept malformed input, except when it's a security risk"
21:48:53 <andythenorth> "Item 3.6.5: the system must work with Citrix"
21:49:33 <andythenorth> or elaborate photoshop mockups
21:49:56 <Rubidium> andythenorth: so, if it launches a Citrix server item 3.6.5 is fulfilled?
21:50:00 <frosch123> supermop: coluoring updated more often if you use the stack-recolourig instead of the usual colormap callback
21:50:24 <andythenorth> Rubidium: no at that point your lawyer has to confirm with their lawyer who is suing who
21:50:28 <nielsm> that's some requirements that just scream for malicious misinterpretation
21:51:32 <supermop> frosch123: seems i'll need a different stack for each age of cargo
21:52:27 <supermop> if its for glowing orange metal ill need to point to one stack for 'very orange', one for 'kind of orange', etc, right?
21:53:47 <supermop> this pick up truck is too small
21:59:54 <supermop> also, it seems that the the truck turns the color for 15 day old cargo at 30 days, etc
22:00:47 <supermop> and the color of 5 day old cargo at 10 and so on
22:01:05 <andythenorth> off-by-one in range checks?
22:01:09 <supermop> (this truck doesn't have any cargo age properties)
22:01:11 <frosch123> 185 ticks per cargo age, 74 ticks per day
22:01:20 <frosch123> cargo age is not in days
22:01:47 <supermop> well its fine, doesn't need to be anything exact
22:02:14 <supermop> i wonder if my pickup truck is the first rv to ever use this var
22:07:51 <supermop> also the color doesn't update at 32 days
22:09:15 <frosch123> you may set some dummy resulf to the 32day callback
22:10:54 <frosch123> "every_32_days: return bitmask(CB_RESULT_32_DAYS_COLOUR_MAPPING);" or something silly
22:14:10 <TrueBrain> whoho, first real bug is also imported :)
22:18:38 <supermop> frosch123: can i just have a random switch that gets triggered every few days?
22:20:37 <frosch123> there is a 32day rerandomisation trigger
22:23:37 *** snail_UES_ has joined #openttd
22:32:56 <andythenorth> TrueBrain: from this import bugs
22:33:54 <TrueBrain> ImportError: cannot import name bugs
22:43:25 *** SeraphAngel has joined #openttd
22:46:07 *** HeyCitizen has joined #openttd
22:49:30 <supermop> andythenorth: do you have a custom recolor sprite to use that purple?
22:49:44 <andythenorth> that purple is recoloured in pixa
22:49:52 <andythenorth> but it could be done with a recolor sprite
22:50:07 <andythenorth> the main thing is to split from CC
22:53:38 *** HeyCitizen has joined #openttd
23:09:40 *** Biolunar has joined #openttd
continue to next day ⏵
