IRC logs for #openttd on OFTC at 2018-12-30
⏴ go to previous day
00:14:21 *** Progman has joined #openttd
00:41:13 *** Thedarkb-T60 has joined #openttd
03:43:30 *** Progman has joined #openttd
07:11:53 *** sla_ro|master has joined #openttd
07:26:23 *** Alberth has joined #openttd
07:26:23 *** ChanServ sets mode: +o Alberth
07:43:30 *** Wacko1976 has joined #openttd
08:43:16 *** andythenorth has joined #openttd
09:02:54 <andythenorth> "Iron Horse is designed as a series of small but very playable sets "
09:03:17 <andythenorth> 65 engines isn't 'small'
09:03:23 <andythenorth> original design had 20 :P
09:03:40 <Alberth> relative to xussr, it is :D
09:03:40 <andythenorth> " it'll be easily expandable should the opportunity arise"
09:04:28 <Alberth> or world-wide train collections, etc
09:05:12 <Alberth> so who is the public of that text?
09:06:48 <Alberth> as a player, I am not sure what to make of "it's easily expandable"
09:07:28 <Alberth> it's more an authoring property
09:07:58 *** ChanServ sets mode: +v tokai
09:11:19 <andythenorth> it's too expandable :D
09:11:36 * andythenorth does the fallacy of adding 'trains I really like'
09:11:43 <andythenorth> which is messing up the tech tree :)
09:14:11 <Alberth> the difference between a hobby and a work project :)
09:14:30 <andythenorth> it's getting horribly close to 'make economic choices about engines'
09:14:39 <andythenorth> instead of 'click the obviously best engine' :P
09:15:06 <Alberth> oh dear, I have to think what engine to pick???!!!
09:15:33 <nielsm> try to avoid "why would I EVER use this one?" situations
09:15:57 <Alberth> I'll just switch on breakdowns, and make a mess of your careful calculations :p
09:16:06 <andythenorth> Horse was doing really well at 'just use this engine for situation X'
09:16:35 <andythenorth> it's interesting to see how much I can jam in without breaking that :P
09:24:05 <Alberth> mostly it depends om how accurate your X is
09:26:19 <andythenorth> it's very contingent on whether wagon speed limits are on
09:26:41 <andythenorth> if they are, the aim is a decision tree that goes like:
09:26:49 <andythenorth> "are the wagons for this cargo fast or slow?"
09:27:00 <andythenorth> "is it a big train or a small one?"
09:57:07 <Alberth> colour looks better, I like the red :)
09:57:19 <andythenorth> it fits the purchase menu better
09:57:27 <andythenorth> but I've cancelled that train :)
09:57:30 <andythenorth> it's not fitting in
10:22:49 <andythenorth> considered it :P
11:12:51 *** gelignite has joined #openttd
11:15:45 <Gabda> is it possible on the linux version to build only one cpp file instead of using make to build the whole project?
11:16:13 <Gabda> if I only changed that one file, but not the others
11:17:01 <Alberth> but make skips everything that doesn't need rebuilding
11:17:26 <Gabda> so it is enough if I don't use make clean before making?
11:17:30 <Alberth> so if you only touch one .cpp file, it will only rebuild that
11:17:54 <Gabda> ok, thx, I will try that :)
11:18:05 <Alberth> yep, make clean wipes everything, so you can rebuild almost from scratch :)
11:18:33 <Alberth> it has a few more thorough cleaning variations
11:19:03 <Alberth> you run it with a -j option ?
11:19:16 <Alberth> eg -j 2 for running 2 compiles at the same time
11:19:57 <Alberth> or rather, 2 make jobs, which can also be non-compile jobs
11:20:39 <Alberth> note that just -j (without a number) is generally a bad idea, as it starts everything at the same time
11:21:57 <Alberth> I have a MAKEFLAGS=-j4 environment variable set, so I don't need to type -j N all the time
11:22:54 <Alberth> this is somewhat hazardous, not all Makefiles are handling concurrent builds properly
11:23:04 <Alberth> ok, saves quite some time :)
11:23:55 <Gabda> I have built the game like 20 times yesterday, so I got bored with it a little for now :)
11:24:15 <Alberth> instead, play a game today :)
11:24:29 <Alberth> 20 full builds is really boring indeed
11:25:20 <Alberth> happens a lot if you make string changes, as pretty much every cpp file uses strings
11:26:32 *** Progman has joined #openttd
11:26:53 <nielsm> normally you'd only use "make clean" if for some reason regular make fails
11:27:07 <nielsm> (and if regular make fails it's often because of some wrong dependencies)
11:29:30 <Gabda> it so much better to build this way :)
11:31:38 <Alberth> that's what make was designed for :)
11:32:26 <Gabda> I think I learned it well with this
11:32:56 <Alberth> you tried writing your own Makefile? that's also an adventure :)
11:33:46 <Alberth> but yeah, there is no substitute for learning by doing
11:33:46 <Gabda> not yet, sounds scary...
11:34:24 <Gabda> and usually I use python, and there is no need for builds
11:35:01 <Alberth> nah, you just have to think from the end product back to the source, ie the other way around, but once you get that, it's not difficult
11:35:59 <Alberth> :o python, such a lovely language
11:38:33 <Gabda> is it possible that in a multiplayer game, more than one client have access to the options -> AI/Game script configuration?
11:39:25 <andythenorth> you need to fork FIRS
11:39:49 <Alberth> in multiplayer, only the server can run ai/game scripts
11:39:50 <Eddi|zuHause> Gabda: well, technically, each client has a separate AI/GameScript config
11:40:22 <nielsm> give the clients a password to remote-config the server
11:40:36 <andythenorth> meh game progression vs. RL vs. buy menu spam
11:40:56 <andythenorth> vs. convergent stats as different engine types upgrade at different rates
11:41:00 <andythenorth> tech tree fun :P
11:41:28 <andythenorth> are there any other grfs besides NUTS with a tech tree concept?
11:42:11 <Alberth> very few, as most are hysterically correct rather than designed for game play
11:42:23 <Gabda> nielsm: you can add decorations as cargo
11:42:31 <Wolf01> Mmmh, the modem is crying about my router using too many NAT sessions per user, I doubled it, but I think I can raise it almost to the limit since the router should be the only user
11:43:41 <Gabda> or glitter is the decoration?
11:43:53 <Alberth> nielsm: separate production of the tree top piece
11:44:27 <nielsm> I thought about glitter->decorations->trees->shop, but think it'd be too long a chain
11:45:14 <andythenorth> yeah ok too many small engines in Horse, they contend for the same key stats of 'power, length'
11:45:17 <Gabda> i thought decoration for houses
11:45:30 <andythenorth> then player has to make 'economic' choices, which is dull :(
11:46:13 <Alberth> andy: pick the 4th one, right?
11:46:45 <Alberth> out-door decoration industries :)
11:46:57 <Alberth> would need a nice house set though :)
11:47:12 <Alberth> lights switch on after delivery :)
11:48:44 <andythenorth> meh, the problem is that I want to include my favourite UK train :P
11:49:35 <Alberth> "highly recommended by andy" sticker
11:50:43 <Eddi|zuHause> well, CETS has sort of like a tech tree, but real vehicles don't exactly follow that tree, so you have duplicates, or gaps, or niche vehicles
11:52:19 <Alberth> gives nice variations in the trains that drive around, pity you can't get a random instance (one of several equivalent vehicles)
11:53:23 <Eddi|zuHause> sometimes the differences between real world vehicles are not modelled by the game, creating more duplicates
11:56:31 <Alberth> if you fail to see the "this program is not designed for what you aim for" signs, then yeah, that's what you get
11:57:18 <nielsm> real world produces solutions like the Pacer that have no logical place in the game world
11:57:47 <andythenorth> Eddi|zuHause: acceleration and range are two of the more obvious :P
11:57:52 <nielsm> (unless you maybe make it extremely cheap to purchase)
11:58:11 <andythenorth> big trains are quite easy to balance
11:58:14 <andythenorth> but small trains...
11:58:33 <andythenorth> tend to converge on one obvious best
11:58:46 <Eddi|zuHause> the game doesn't really pose a "declining passenger numbers on branch lines" situation
11:58:49 <andythenorth> but then all small trains are a boring monoculture
11:59:14 <andythenorth> Eddi|zuHause: 'very small railcars' are like shunters, no place in the game
12:01:21 <nielsm> if you could have a kind of persistent subsidies where you're paid to maintain a connection between small towns, they could make sense
12:01:32 <nielsm> maybe that's an idea for a GS
12:03:18 <nielsm> i.e. not paid for the delivery of passengers, but paid for the vehicle making a number of stops within a time period
12:06:00 <Eddi|zuHause> nielsm: that doesn't stop the fact that the line will immediately be overcrowded
12:07:18 <Eddi|zuHause> anyway, we haven't solved game balance in the last 24 years, we won't solve it today.
12:08:10 <Eddi|zuHause> nielsm: that still makes passenger numbers grow over time
12:09:22 <Eddi|zuHause> there's no "decline age" that made up most of western europe's railways for the second half of the 20th century
12:09:48 <nielsm> cdist doesn't selct destinations based on total acceptance level, does it?
12:10:18 <nielsm> because I think it'd make sense to calculate total acceptance (even above 8/8) of each cargo on a station and rank the potential destinations by that
12:10:43 <Eddi|zuHause> i think that was tried
12:11:15 <nielsm> no effect or unintended effects?
12:11:42 <Eddi|zuHause> iirc it behaved weird for industries
12:12:03 <Eddi|zuHause> because some industries have each tile accept something, other industries only one or two tiles
12:12:07 <nielsm> eh well, special case PASS and MAIL then
12:12:34 <Eddi|zuHause> and then it makes it difficult because you have to cover the whole industry which you didn't before
12:13:09 <andythenorth> Eddi|zuHause: 'decline' doesn't make interesting gameplay :)
12:13:42 <Eddi|zuHause> andythenorth: no, the interesting gameplay would come from giving you the right tools to handle it
12:14:12 <nielsm> part of the change in rail usage since 19th century also traces to labour costs and safety requirements growing, doesn't it?
12:15:42 <andythenorth> nielsm: ranking destinations by acceptance sounds close to a demand economy
12:15:47 <andythenorth> which Railroad Tycoon 3 had
12:15:53 <andythenorth> each tile has a demand factor
12:16:13 <nielsm> andythenorth it would solve the issue of a small town being just as popular a destination as a metropolis
12:16:19 <nielsm> which it usually shouldn't
12:16:22 <andythenorth> over-supply would cause demand to fall
12:16:44 <andythenorth> we have no way to generate cargo based on demand though eh?
12:16:58 <andythenorth> e.g. if demand was saturated, generation would plateau, or back off
12:17:17 <nielsm> if GS could control industry production, maybe you could make that :)
12:17:34 <andythenorth> it could be done via proxy
12:17:57 <andythenorth> communicate via the town
12:17:58 <nielsm> the station rating somewhat intends to model that, I think
12:18:14 <andythenorth> it is pretty close
12:18:16 <nielsm> if you leave huge amounts of cargo not picked up, production drops
12:19:09 <Alberth> you don't need to control production, just make it run losses if you over-supply
12:19:37 <Alberth> players will stop moving stuff if it's negative in payments
12:20:29 <nielsm> SPI somewhat does that, with secondary industries having a limit on how fast they can produce, regardless of how much raw material you pump in
12:20:40 <Alberth> industry may need to get improved by needing time to ramp up or down
12:21:26 <Alberth> so continued serving a factory pays
12:22:28 <Alberth> the gung-ho thing of firs is too
12:23:18 <Eddi|zuHause> <Alberth> players will stop moving stuff if it's negative in payments <-- there's so many things wrong with this statement...
12:24:59 <andythenorth> my ideal is demand-based town cargos
12:25:07 <andythenorth> but not industrial cargos
12:25:12 <andythenorth> so pax, mail, food, water, goods
12:25:25 <andythenorth> the ones that need distributed rather than concentrated
12:26:00 <nielsm> like, the water tower has a capacity, and when that's filled it can't accept more?
12:26:44 <andythenorth> not hard acceptance
12:26:50 <andythenorth> more like a weighting to cdist
12:26:55 <andythenorth> and maybe to payment
12:26:56 <Eddi|zuHause> cargodist lacks an incentive to actually distribute one-to-many cargos like goods, instead of just dumping them all in one place
12:27:18 <nielsm> actually could that be made a flag or cargoclass or something?
12:27:36 <nielsm> cdist has four classes right now, PASS, MAIL, VALU, and misc
12:27:40 <andythenorth> it could be made a flag if it could be used
12:27:54 <andythenorth> 'distribute widely'
12:28:03 <Alberth> just "town goods" imho
12:28:23 <andythenorth> we discussed simply extending cdist cargos before years ago :)
12:28:25 <Eddi|zuHause> it probably needs a newgrf property "one-to-many"
12:28:25 <nielsm> it'd also solve your Supplies problem :)
12:28:43 <andythenorth> extending the hard-coded cdist options in core game, spirals horribly
12:28:56 <Eddi|zuHause> i would not tie it to "town effect" or something like that
12:29:01 <andythenorth> unrelated, I deleted the high-power bi-mode engine, it's too weird
12:29:38 <Alberth> 2 seconds compile time saved :p
12:30:13 <Eddi|zuHause> nielsm: seems like the wrong place as well
12:31:11 <Eddi|zuHause> nielsm: trying to save a few bits here by merging unrelated things will create even more maintenance mess
12:31:42 <nielsm> Eddi|zuHause, but flags 0, 1, and 3 are exactly those used for cdist aren't they?
12:32:33 <Eddi|zuHause> nielsm: i'm not sure about that
12:43:53 <Eddi|zuHause> what i think CDIST needs is two separate flag groups: [one-to-one, one-to-many, many-to-one, many-to-many] and [unidirectional, bidirectional (unbalanced), bidirectional (balanced)]
12:46:32 <xahodo> I´m trying to get started on NML, but somehow can´t get the notepadpp.xml file to download from the openttdcoop server (don´t know whether this is the right location to ask).
12:47:04 <Eddi|zuHause> that's probably not needed to "get started"
12:47:58 <xahodo> Well, I would like to have some colors when typing. So I can actually discern what I´m doing.
12:48:42 <xahodo> nvm problem between chair and keyboard
12:48:43 <Eddi|zuHause> i'm usually picking a random syntax highlighting from the list. C/C++ probably works well enough
12:52:29 <Alberth> niels: Poultry production seems missing?
12:52:45 <nielsm> producing Wool and Poultry
12:52:51 <Alberth> ah, right, missed it, thanks
12:55:37 <Eddi|zuHause> nielsm: reindeer could be helicopter-like and heliports "chimneys"?
13:09:44 <nielsm> hmm I wonder if the IND_FLAG_CUT_TREES flag works with non-primary industries
13:10:23 <nielsm> i.e. if the tree farm given that flag will receive glitter, then cut a bunch of trees, and then produce cargo
13:11:31 *** Thedarkb-T60 has joined #openttd
13:13:07 <Eddi|zuHause> i don't think you can correllate it to receiving cargo
13:13:40 <nielsm> hmm yeah, the function that cuts trees also adds the output cargo
13:14:06 <nielsm> so it'd have to be a separate industry producing "raw" trees for that to work
13:15:01 <Eddi|zuHause> also, tree growth is usually not fast enough to sustain that flag
13:15:18 <nielsm> the player has to plant new forests
13:36:13 *** HerzogDeXtEr has joined #openttd
14:04:46 *** Wacko1976 has joined #openttd
14:46:40 <nielsm> hmm what unit is glitter measured in? :D
15:08:17 <nielsm> need to disable a bunch of the default industries and cargos
15:13:48 <Gabda> the workplace of water elves
15:26:57 *** sla_ro|master has joined #openttd
15:35:39 *** andythenorth has joined #openttd
15:41:53 <andythenorth> 'Shredder', 'Super Shredder', 'Ultra Shredder' ??
15:42:20 <andythenorth> pulp is measured by volume eh nielsm? :)
15:42:47 <andythenorth> 'woolly' not 'wooly' :D
15:51:04 <nielsm> I need t o figure out why the butcher and knitting workshops build on water :P
15:51:42 <nielsm> they have IND_FLAG_ONLY_IN_TOWNS respectively IND_FLAG_BUILT_NEAR_TOWN
15:52:37 <andythenorth> gen 6 apparently reuses the same bodyshell for 3 trains :P
15:52:49 <andythenorth> and the paint is pretty similar too :P
16:01:34 <Eddi|zuHause> and the problem with that is?
16:01:39 <Eddi|zuHause> it's too realistic?
16:06:35 <andythenorth> is it a problem at all?
16:15:16 <nielsm> hm anywhere I can just see the DOS palette? I don't have any software ready to load any of the palette files for download
16:16:15 <andythenorth> got bored of counting numbers on my fingers
16:16:33 <nielsm> and what values are reserved/magical?
16:18:57 <andythenorth> hmm actually 232-254
16:19:29 <andythenorth> company colour is 198-205 and 80-87
16:19:42 <andythenorth> 255 is white, not used
16:20:15 <andythenorth> and I think 227-231 are animated, but I never figured out where / why
16:33:31 <andythenorth> ha ha I found a rounding quirk
16:33:57 <andythenorth> 18200hp is rendered 18198hp
16:34:41 <andythenorth> Eddi|zuHause: so this idea of adding engine power by counting coaches works brilliantly
16:34:51 <andythenorth> what about TE? o_O
16:35:10 <Eddi|zuHause> you can add TE just the same way?
16:35:22 <andythenorth> it's range is 0-1
16:35:34 <andythenorth> I could move the weight to the engine
16:35:40 <andythenorth> dunno how accurate the physics is
16:35:41 <Eddi|zuHause> well, then add weight
16:35:53 <Eddi|zuHause> the physics is totally wrong either way
16:35:58 <andythenorth> that's a good answer
16:36:33 <Eddi|zuHause> you also need to add running costs
16:36:45 <nielsm> as long as coupling between cars isn't modeled (infinitely strong couplers with no slack) individual car weight shouldn't really matter
16:36:47 <andythenorth> I could just put the power on the coaches
16:36:58 <LordAro> Borg: precompiled headers
16:37:31 <Eddi|zuHause> the game adds all these stats up to the front engine anyway
16:39:31 <Borg> LordAro: oh damn... im glad im in C world :D
16:39:44 <andythenorth> so the physics of the train stretched over hills....?
16:40:08 <Eddi|zuHause> andythenorth: irrelevant
16:40:09 <nielsm> hmm, would Glitter be CC_ARMOURED ? :D
16:40:13 <Borg> LordAro: I found other nice PDF while ago... Title: Pitfalls of C++ programming... nice reading too
16:40:29 <andythenorth> nielsm: hazardous
16:41:07 <Eddi|zuHause> powderized, pourable
16:41:20 <nielsm> nah it's measured in Canisters
16:41:26 <Eddi|zuHause> transport it in hopper cars
16:42:34 <andythenorth> hmm, I can just add the run costs to the coaches directly no?
16:42:42 <LordAro> Borg: in general, i agree with the author of the original article. C++ has many advantages over C, but it's all too easy to write unintelligible nonsense
16:43:00 <LordAro> C++14 is peak C++, imo
16:43:22 <LordAro> (with a few things cut out/discouraged)
16:47:31 <Borg> LordAro: yeah, you are probably speaking out of game programming world.. I then agree..
16:49:35 <TrueBrain> whoho, I finally have a self-maintaining k8s cluster :D
16:49:48 <TrueBrain> OpenTTD-IaC defines everything, and can be redeployed without 'administrator' stuff
16:49:59 <TrueBrain> took me only a day longer than I expected :P
16:55:04 <LordAro> Borg: not exclusively
16:56:44 <TrueBrain> it is really nice; I now commit to OpenTTD-IaC, it creates a docker, and if my k8s was connected to the internet, it would auto-deploy the new version :D
16:57:06 <TrueBrain> means I can focus on things that are interesting .. not stupid SysOp shit :D
16:59:33 <TrueBrain> now I can finally setup staging for OpenTTD
17:01:03 <LordAro> TrueBrain: now the trick is seeing if anyone else can understand/deploy/modify it :p
17:01:27 <TrueBrain> LordAro: yeah, but that is not really an issue tbh
17:01:37 <TrueBrain> more important: once I setup, for example, www.openttd.org repository
17:01:43 <TrueBrain> anyone with commit rights, can get a new version in production
17:01:57 <TrueBrain> that I have to do initial work to get it setup, fine by me
17:02:05 <TrueBrain> after that ... I have nothing to worry about :P
17:02:34 <TrueBrain> hmm, that makes me wonder about eints ..
17:02:52 <TrueBrain> could we make it authenticate against github, and have a file in the repo stating who has access to what language
17:03:01 <TrueBrain> and adding a 'translator' is just make a PR requesting for it
17:03:06 <TrueBrain> or would that be weird?
17:03:44 <TrueBrain> I believe frosch was working on an adapter to support both OpenTTD LDAP and GitHub SSO for login .. guess that would work too
17:03:56 <TrueBrain> still .. I think the list of translators should be in git :)
17:04:11 <TrueBrain> (now it is so hidden who translators are, and how many, etc etc)
17:05:19 <nielsm> andythenorth, teach me how production callbacks work
17:05:38 <andythenorth> what do you want to do?
17:06:58 <TrueBrain> "simple" and NML .. riiigggghhhtttttt
17:07:23 <nielsm> make 1*GIFT from 1*TOYS+1*PAPR, or 1*GIFT from 1*SWTR+1*PAPR
17:08:13 <andythenorth> do you want them made when cargo arrives?
17:08:31 <andythenorth> hmm you want to combine
17:08:32 <nielsm> when cargo arrives, but with stockpiling
17:09:06 <andythenorth> ok you have a few options
17:09:07 <nielsm> so unused inputs stockpile at the industry until there's a usable combination
17:09:15 <andythenorth> you can use the weird built-in stockpile thing
17:09:24 <andythenorth> probably best for this idea
17:09:44 <andythenorth> or you can do your own stockpile with storage
17:10:26 <andythenorth> so when cargo arrives you'll need to check what's on the stockpile
17:10:48 <andythenorth> if you can combine things for output, then produce and consume appropriately
17:10:56 <andythenorth> otherwise just leave it on the stockpile
17:11:19 <nielsm> also, displaying the current stockpile is via cargo_subtype_display right?
17:11:24 <Borg> callllllbacks :D yeah.. took me a while to use them
17:16:54 <andythenorth> nielsm: current stockpile is displayed automatically
17:17:44 <andythenorth> we changed the options for that, trying to find the docs
17:18:24 <andythenorth> yeah looks like cargo_subtype_display
17:19:42 <nielsm> oh, stockpile is automatic? nice
17:20:59 <Borg> yeah, you just need to enable it. for industry..
17:22:35 <nielsm> and uh, how do you do that then...
17:22:57 <andythenorth> enable production cb, iirc
17:23:06 * andythenorth never codes nml, can't remember :)
17:23:14 <andythenorth> I live in my own weird world of abstracted python
17:23:21 <andythenorth> or python abstractions :P
17:23:34 <Borg> yeah. I dont remember either :D
17:24:23 <Borg> lets see my industry definitions
17:26:06 <Borg> seems it enough do do: 21 04 on action 0
17:26:45 <nielsm> oh, maybe I could also add random events like, "Hidden cache of wrapping paper discovered at Tartbottom Warehouse!" (adds 500 units of wrapping paper to stockpile)
17:27:45 <nielsm> is that actually possible?
17:28:26 <andythenorth> you can just negatively consume
17:28:47 <Borg> but that would be uber complicated to code
17:29:31 <Borg> my single production CB on factory is like 3 screens long :D
17:31:31 <Borg> its doable.. like catch random montly production update CB... write do persistent storage about event..
17:31:55 <Borg> and then.. on production CB (called once every 256 ticks) handle that.. yeah
17:34:41 <andythenorth> I could have shortened it with loops :P
17:35:30 <Borg> anyone chacked how NML generates NFO? :) is it efficient?
17:36:14 <Borg> for now I use NFO.. but.. it hard to maintain :)
17:36:19 <Borg> and I forgot it quickly ;D
17:40:31 <andythenorth> it's not efficient
17:40:40 <andythenorth> I am fairly certain
17:41:16 <Borg> I should put more comments.. :D
17:41:28 <Borg> its part of my NewGRF w/ stockpile industries + electricity
17:42:15 *** Wacko1976 has joined #openttd
17:45:24 <Alberth> Depends on how you define efficient, I tend to measure it in programmer/maintainer time. CPU time is very very cheap.
17:46:19 <andythenorth> except when compiling :P
17:46:25 <andythenorth> NML is hideous to compile
17:46:40 <andythenorth> the cycle time on small tweaks is horrid
17:48:28 <andythenorth> I should optimise the grf design to suit the compiler :D
17:51:42 <Borg> Alberth: oh yeah.. I forgot everyone these days run those Intel Extreme edition CPUs.. clocked to 3GHz and 8+ cores :P
17:51:56 <andythenorth> it's single threaded :P
18:00:26 <nielsm> tha paper handling isn't quite right though
18:04:21 <nielsm> hm toys should be in boxes shouldn't it
18:04:42 <andythenorth> cargo units are weird :P
18:04:46 <andythenorth> how many boxes per ton? o_O
18:05:11 <nielsm> okay items is the toyland default for TOYS cargo
18:05:31 <nielsm> that's the one cargo I'm not overriding a whole lot in
18:10:04 <nielsm> ah I've forgotten to modify the industry tile accepts, so it's not playable with or without the cb fixed
18:11:07 <andythenorth> eventually one templates everything :P
18:11:18 <andythenorth> and writes an industry class in python
18:11:50 <andythenorth> I have .pynml :P
18:11:57 <andythenorth> we did have CPPNML
18:12:02 <andythenorth> which was....best forgotten
18:27:42 *** gelignite has joined #openttd
18:40:41 <nielsm> hmm even with "again" as a fixed zero register, I get the production loop?
19:03:43 <nielsm> okay I removed the again arg entirely from the produce() statement and now it's not looping
19:03:49 <nielsm> but it's also just not producing anything
19:05:58 <andythenorth> what happens if you just dump values into produce, instead of using the registers?
19:06:31 <andythenorth> e.g. produce(warehouse_prod_prod, [TOYS: 255; SWTR: 255; PAPR: 255;], [GIFT: 255;])
19:08:54 <nielsm> it starts producing all the time :)
19:09:09 <andythenorth> ok so produce works
19:09:30 <andythenorth> are the register expressions the right way round?
19:09:55 <andythenorth> not address, value
19:12:51 <andythenorth> changing fonts in UI is a mysterious concept
19:15:42 <LordAro> nielsm: looks alright to me
19:16:04 <LordAro> how does it normally look? :p
19:16:04 <nielsm> okay now it's producing stuff even with the LOAD_TEMP'y produce, with nothing in the inputs
19:20:55 <andythenorth> no original baseset :P
19:22:02 <andythenorth> nielsm: paste again?
19:23:21 <nielsm> trying something out atm
19:29:38 <nielsm> fixed, missed a ; in three places
19:31:46 <nielsm> bah it's producing even without papr
19:34:22 *** erratic has joined #openttd
19:34:37 <andythenorth> combinatory production headaches :)
19:35:15 <andythenorth> incoming_cargo_waiting("PAPR") > 0 ??
19:35:52 <nielsm> my intention is to produce a 0 or 1 value based on whether the variable is nonzero
19:37:20 <andythenorth> yeah I figured :)
19:37:40 <andythenorth> incoming_cargo_waiting("PAPR") * ((incoming_cargo_waiting("TOYS")>0) + 2*(incoming_cargo_waiting("SWTR")>0)) won't dod that if TOYS or SWTR are present?
19:37:56 *** pete1 is now known as pete2
19:38:08 <nielsm> I changed that slightly: (incoming_cargo_waiting("PAPR")>0) * ((incoming_cargo_waiting("TOYS")>0) + 2*(incoming_cargo_waiting("SWTR")>0))
19:38:13 <nielsm> still failing the same way
19:38:18 <pete2> hey, has OpenTTD on linux any other musicdriver beside extmidi?
19:38:21 <nielsm> (producing even when PAPR is 0)
19:38:39 <nielsm> pete2, "yes", if you use a development version
19:38:40 <andythenorth> this is why I made FIRS print a lot of debug to the industry window :P
19:38:59 <andythenorth> if you use perm storage, you can read them in the built-in newgrf debug
19:39:22 <andythenorth> can be handy, even just for debugging
19:39:39 <pete2> so if i use stable my only option is to use timidity or any other as extmidi-device?
19:39:44 <nielsm> pete2, 1.8.0 and earlier only have extmidi as working midi on linux, but a few months ago we added fluidsynth support in the current development branch
19:40:24 <nielsm> or well, maybe Allegro can also be used to play midi, I don't think I tried
19:40:45 <pete2> and IIRC midi music was working on linux with 0conf before
19:40:51 <andythenorth> hmm, maybe I'm imagining the storage in the debug window :(
19:40:54 <andythenorth> I swear it was there
19:41:08 <TrueBrain> andythenorth: what is a good name for the www repository wehave been working on?
19:41:14 <TrueBrain> OpenTTD/OpenTTD-website sounds so weird
19:41:21 <TrueBrain> and 'website' .. it is only 'www'
19:41:36 <TrueBrain> OpenTTD/www.openttd.org ?
19:41:42 <andythenorth> seems reasonable
19:41:42 <TrueBrain> OpenTTD/main-website ?
19:42:10 <andythenorth> it's the website
19:42:13 <andythenorth> everything else is services
19:42:22 <andythenorth> the website is how services are discovered
19:45:34 <pete2> nielsm: OpenTTD offers me only extmidi and non as sounddriver
19:46:08 *** Eddi|zuHause has joined #openttd
19:47:12 <andythenorth> nielsm: persistent industry storage is shown in debug FYI
19:47:21 <andythenorth> just doesn't trigger display until at least one is used
19:48:34 <TrueBrain> owh joy, Ruby broke upstream,and it no longer wants to compile now :(
19:49:15 <andythenorth> pinned versions? o_O
19:49:23 <andythenorth> a rubyenv cached?
19:49:30 <TrueBrain> apk ruby is the issue
19:49:35 <andythenorth> 'upstream breaks'
19:49:35 <TrueBrain> one of the few non-pinned things
19:49:47 <TrueBrain> they have a ticket on their bugtracker, unanswered for over a month
19:49:53 <TrueBrain> owh well .. found a workaround because of that
19:50:03 <TrueBrain> Ruby depends on rdoc, rdoc depends on Ruby ... yippiiiiieeeee
19:50:08 <andythenorth> the other one should have worked nielsm
19:50:19 <andythenorth> but without testing it locally, I can't prove that :D
19:50:54 <andythenorth> branching conditions are mind-melt :P
19:51:02 <nielsm> it's still doing the same D:
19:51:24 <andythenorth> producing without input?
19:51:44 * glx is making progress in adding regression testing to MSVC
19:52:12 <nielsm> producing with only one input
19:52:16 <andythenorth> nielsm: accept_cargo("TOYS", produce_cargo("GIFT", 1))
19:52:20 <LordAro> pete2: development versions alao offer fluidsynth support, but there's no release with this yet
19:52:21 <andythenorth> accept_cargo("TOYS", produce_cargo("GIFT", 0))
19:52:32 <andythenorth> or just accept_cargo("TOYS")
19:52:39 <andythenorth> dunno the new syntax yet
19:53:06 <andythenorth> too many ways to produce
19:53:20 <andythenorth> and ultimately, one always needs the production cb anyway
19:53:30 <andythenorth> and all this legacy crap slows down things like 16 cargos in / out
19:53:44 * andythenorth starts petition for LessNewGRF
19:54:18 <andythenorth> to be achieved by simply reducing the NML interface
19:54:22 <andythenorth> leave the bytecode there
19:54:59 <TrueBrain> and the first attempt of auto-deploy fails with a ... "IndexError: list index out of range"
19:55:12 <andythenorth> sounds like every python web page I write
19:55:19 <andythenorth> that or a type error
19:57:33 <nielsm> okay NOW it's stockpiling
19:58:54 <andythenorth> it's on the 8 times / month cb
19:59:01 <andythenorth> you need produce_cargo_arrives or something
19:59:23 <andythenorth> produce_cargo_arrival
20:00:45 <nielsm> does max() not actually work at runtime, only compiletime?
20:01:23 <andythenorth> I thought it was runtime
20:01:47 <andythenorth> should be runtime
20:01:50 <Eddi|zuHause> <andythenorth> sounds like every python web page I write <-- what? no syntax error?
20:02:03 <andythenorth> after the syntax error
20:02:15 <andythenorth> once it's actually compiled
20:02:43 <andythenorth> oh you found it too
20:02:53 <andythenorth> yeah the nfo ops are pretty good
20:03:01 <andythenorth> everything you need is there
20:03:04 <Eddi|zuHause> so, is ICQ finally dead?
20:08:36 <andythenorth> oh someone has owned some Tripod sites
20:08:42 <andythenorth> didn't know Tripod was still a thing
20:08:57 <andythenorth> if I click on anything, I get a plausible looking 'upgrade Flash player' message
20:09:09 <andythenorth> only Adobe don't serve that from AWS :P
20:09:31 <Eddi|zuHause> the only thing you should do with flash player is burn it with fire
20:09:55 <andythenorth> it's not installed
20:10:09 <andythenorth> so upgrading it is Definitely Not a Thing
20:14:39 <nielsm> grr something going wrong here, produced 20 sacks of presents with 14 toys and 20 rolls of paper
20:15:27 <nielsm> and then 80 presents with 5 toys and 80 rolls paper
20:15:46 <nielsm> (that's some pretty sad presents)
20:16:09 <TrueBrain> okay .. I was wondering why letsencrypt fails .. but someone added a CAA record to openttd.org :D
20:17:37 <TrueBrain> and not many people can
20:18:24 <TrueBrain> that same person wants me to switch to letsencrypt, the horror :P
20:19:40 <andythenorth> nielsm: switch to persistent storage, it's find-and-replace
20:19:46 <andythenorth> then you can watch the numbers go through
20:28:56 <nielsm> should be using min() for a bunch of this
20:29:16 <andythenorth> eh TrueBrain :) \o/
20:31:29 <TrueBrain> I am still surprised how incredible fast the page is
20:32:36 <andythenorth> that was my awesome work
20:32:42 <andythenorth> I am so optimised
20:33:55 <TrueBrain> has nothing to do that it is a static page served via 3 nginx via a loadbalancer with edge-cache enabled I assume? :P
20:34:15 <nielsm> it's responsive design!
20:35:16 <andythenorth> TrueBrain: no I have actually just hacked you and it's serving locally from your computer
20:35:35 * andythenorth will now stop talking crap
20:36:00 <andythenorth> also, why do idiots say "Can't hack a hacker"?
20:36:13 <andythenorth> cos, ya know, not true
20:36:20 <TrueBrain> because it is bullshit, they know it, and they hope you are scared now, and not try
20:39:12 <nielsm> yay now it's doing the right thing!
20:43:17 * andythenorth doesn't read *all* of it :)
20:43:59 <nielsm> but next bug, several of the industries are not built during worldgen
20:44:13 <nielsm> (it seems those based on non-toyland ones)
20:44:18 <nielsm> but can be built afterwards
20:48:45 <TrueBrain> someone is having fun :P
20:48:50 <andythenorth> is the climate set correctly?
20:50:56 <nielsm> there is no climate property for industries?!
20:51:11 <Eddi|zuHause> "item(FEAT_CARGOS, ctPASS, 0x00)" <-- can you use "PASS" instead of "0x00" there?
20:52:11 <Eddi|zuHause> i think NML replaces instances of cargo labels with the numeric position in the CTT
20:52:27 <Eddi|zuHause> so i imagine that would work
20:53:06 <nielsm> yes but it'd be incidental
20:53:18 <nielsm> since it needs to be the ID of the original PASS cargo
20:53:19 <Eddi|zuHause> well, you can reorder the CTT
20:57:24 <Eddi|zuHause> my intuition is that adding a comment to the CTT saying "original cargos must be at this postion" would be cleaner than magic numbers sprinkled throughout the code
21:01:52 <nielsm> okay something's wrong with the flags in NML
21:02:04 <TrueBrain> incredible .. this works ... I push something to master, GitHub tells Azure Pipelines that I did, AP builds a new Docker image, pushes it to Docker Hub. If it was to master, it starts a Release. The Release tells my 'deployer' on k8s that there is a new version, which tells Helm to deploy that on its own k8s :D And that really does work :o
21:02:09 <TrueBrain> I am .. shocked? surprised?
21:07:03 <andythenorth> nielsm: some cargos have to stay in the original position btw
21:07:14 <andythenorth> for newgrf vehicle reasons I've forgotten
21:07:33 <andythenorth> or town sets maybe
21:07:42 <andythenorth> anyway, the CTT is not a safe abstraction tool
21:07:55 <nielsm> yes PASS MAIL GOOD FOOD like things
21:08:02 <andythenorth> it's safest to assume that all of newgrf is built on sand
21:08:07 <nielsm> things town buildings might accept
21:10:07 <TrueBrain> and .. it is now live :)
21:11:59 <nielsm> okay I need to make a custom tilelayout for one of the industries, need to figure out the syntax for that then
21:15:10 <TrueBrain> I will move it to OpenTTD soonish
21:15:48 <TrueBrain> and I enforced that you need to make a PR .. no more pushing to master :P
21:16:16 <TrueBrain> still need to fix the 'latest' release/nightly indicators .. hmmmm
21:16:47 <andythenorth> ok I will look at that, but not now :P
21:17:05 <andythenorth> afaict the liquid var is set correctly
21:23:20 <Borg> guys.. instead trying to add cool stuff. maybe fix basic stuff?
21:23:36 <Borg> why OpenTTD *STILL* allows join to company.. or join & create company?
21:23:58 <Borg> that model 'connect <host>#<id>' suxx ass.. really
21:24:04 <Borg> you should always join as spectator...
21:24:18 <Borg> noone else is bothered about it?
21:26:23 <andythenorth> sounds complicated :P
21:28:04 <TrueBrain> well, I just got told we are not allowed to have fun, but we should focus on "basic stuff", so complicated or not, off you go!
21:28:20 <andythenorth> oof I can't be bothered to get that popcorn
21:28:34 <andythenorth> I am going to clone the website and see what you broke
21:29:35 <andythenorth> can you be more specific? o_O
21:30:20 <andythenorth> oof I have to build jekyll again
21:30:55 <Xaroth> Borg: nobody is stopping you from making a PR to address those issues.
21:31:36 <Xaroth> andythenorth: that's what you get when you let TrueBrain near your codebase, so you really brought it onto yourself :P :P
21:33:18 <andythenorth> why is ruby so slow :(
21:33:27 <andythenorth> wait, I fixed it
21:33:32 <andythenorth> why is $everything so slow
21:34:01 <Borg> Xaroth: yeah I know... I should do a patch.. instead of complain.. but I still didnt made Mingw env for OpenTTD building
21:34:20 <Xaroth> you can also use visual studio, if that's your fancy.
21:35:18 <Borg> nope its not.. I stopped using VS long time ago.. last version I used was 2002 NET or sth..
21:35:26 <Borg> I still have skeletonal VC5 and VC6...
21:37:01 <andythenorth> TrueBrain: you broke it in prod
21:37:09 <andythenorth> nav highlight works for me locally
21:37:28 <TrueBrain> that is not possible :D
21:37:41 <andythenorth> what, you think it's just a static site? :P
21:44:23 <andythenorth> liquid vars might not bubble up the templates
21:50:11 <andythenorth> TrueBrain: so I have to fork the whole thing and PR now?
21:50:13 <andythenorth> can't just branch?
21:50:33 <andythenorth> that's regrettable
21:50:37 <andythenorth> building Ruby takes 5 mins
21:51:09 <LordAro> ideally you shouldn't have been just branching anyway :p
21:52:34 <andythenorth> TrueBrain: are you just renaming it tomorrow? or a new repo?
21:53:14 <andythenorth> but my fork will follow?
21:53:20 <andythenorth> due to github magic?
21:53:29 <TrueBrain> no clue; but forking is easy/cheap :)
21:56:38 <andythenorth> but right now is not time to make a fork and build ruby again
21:57:01 <TrueBrain> I sometimes wonder how complex people made it theirself :D
21:57:19 <andythenorth> I made it 'working'
21:57:34 <TrueBrain> git checkout -b myfix
21:57:40 <TrueBrain> git rebase origin/master
21:58:22 <andythenorth> rebuild jekyll because the first build fails
21:58:34 <TrueBrain> my point being, if you already build it, who cares on which fork you have to push it?
21:58:43 <TrueBrain> forks are just .. a reference
21:58:56 <andythenorth> I'm assuming that I can change the remote for git
21:59:06 <TrueBrain> I just told you how, yes :P
21:59:08 <andythenorth> but I had 4 hours sleep 2 days in a row
21:59:58 <andythenorth> I think we'll stick to trash talk, not fixing my git, again
22:00:28 <andythenorth> oh alberth left :P
22:00:32 <andythenorth> NRT rebase didn't happen then :)
22:01:00 <andythenorth> he oddly volunteered to try it
22:01:32 <andythenorth> does staging pick it up?
22:01:41 <TrueBrain> not until it is reviewed and merged :)
22:01:52 <TrueBrain> PRs have nothing! :D
22:01:53 <andythenorth> I'll review it: "it's awesome"
22:02:14 <andythenorth> actually LordAro should
22:04:24 <TrueBrain> now lets see what happens :D
22:08:43 <TrueBrain> lot better andythenorth :D
22:11:25 *** sla_ro|master2 has joined #openttd
22:16:08 <LordAro> everything should have less cabbage
22:17:04 <andythenorth> the other end of my USB-C cable is not in the charger
22:17:17 <andythenorth> the electrons have not jumped the gap much either
22:55:39 <Eddi|zuHause> i managed to do that with a regular power plug charger
22:59:46 * andythenorth mumbles about tesla
22:59:48 <andythenorth> and air charging
23:04:15 <Eddi|zuHause> if you want to "air charge" a tesla, you probably better not get near that thing
23:18:35 *** sla_ro|master2 has quit IRC
23:33:33 *** andythenorth has left #openttd
23:52:57 *** Thedarkb-T60 has joined #openttd
continue to next day ⏵
