IRC logs for #openttd on OFTC at 2011-11-07
⏴ go to previous day
01:27:42 <z-MaTRiX> also, this is eternity
01:51:36 *** blotek___ has joined #openttd
02:38:03 *** glx is now known as Guest16016
02:43:23 <nicfer> can I write a GRF with NML that adds a new vehicle using default game sprites?
03:26:37 *** rhaeder1 has joined #openttd
03:56:54 <nicfer> I was trying to compile a test grf but I got this error:
03:56:56 <nicfer> nmlc: "input", line 3: Syntax error, unexpected token "name"
03:58:41 <nicfer> now I got another error:
03:58:43 <nicfer> nmlc: "input", line 35: Unrecognized identifier 'spritegroup_flatbed_truck_1_goods' encountered
05:08:20 *** supermop_ has left #openttd
05:21:08 *** supermop_ has joined #openttd
05:56:37 *** Eddi|zuHause has joined #openttd
06:21:09 *** JVassie has joined #openttd
07:00:18 *** andythenorth has joined #openttd
07:11:54 *** Qantourisc has joined #openttd
07:30:22 *** Brianetta has joined #openttd
07:38:52 *** sla_ro|master has joined #openttd
07:57:18 <Terkhen> hmm... that spritegroup nicfer mentioned is from ogfx-rv, and I remember getting that error... maybe he's compiling a broken revision
08:06:53 <planetmaker> from what I read, that error happens if one doesn't read the following two pages of the tutorial which introduce the graphics. But I didn't test
08:10:25 *** Celestar has joined #openttd
08:14:04 <Terkhen> oh, right, code for the tutorials :)
08:15:36 *** Celestar_ has joined #openttd
08:23:08 *** Progman has joined #openttd
08:24:31 *** Celestar_ is now known as Celestar
08:28:48 *** andythenorth has joined #openttd
08:28:55 *** andythenorth has left #openttd
08:32:30 *** DayDreamer has joined #openttd
08:37:15 *** Brianetta has joined #openttd
08:41:21 *** andythenorth has joined #openttd
08:48:38 * andythenorth just had the worrying thought that MB doesn't understand the cargo system :O
08:49:04 <andythenorth> I've been poking at it for 3 years, on the assurance that it's perfect, and I must be the one who's stupid
08:50:21 <andythenorth> he thinks that classes are OR when it's convenient to him for them to be OR, and AND when it's convenient for them to be AND
08:50:42 <andythenorth> and in any case he wants to use explicit labels anyway
08:53:49 <andythenorth> fibre crops are piece goods and bulk, and may travel by wagons that are refittable to bulk, wagons that refit to piece goods, or wagons that refit to piece goods AND bulk
08:54:11 <andythenorth> but food is express and refrigerated, and may only travel in vehicles that provide both express AND refrigerated
08:56:15 <andythenorth> this latter condition can only be achieved by excluding refrigerated cargo from non-refrigerated express vehicles
08:59:15 <peter1138> how about don't overcomplicate it :p
08:59:35 <andythenorth> peter1138: I'm trying to under-complicate it
08:59:55 <andythenorth> anyway I'm wrong
09:00:17 <andythenorth> fibre crops are intended to only travel in wagons that provide piece goods AND bulk
09:01:52 <andythenorth> I think the classes are using AND, which is undocumented
09:02:44 <planetmaker> uhm... piece and bulk?
09:03:11 <andythenorth> planetmaker: according to spec
09:03:25 <andythenorth> planetmaker: assume I have no opinion on any cargo classes until I understand the system
09:03:34 <andythenorth> *specific classes, especially FIRS classes
09:04:03 <andythenorth> I have been treating them as OR
09:04:14 <andythenorth> I know Pikka has treated them as OR, because he told me to
09:04:18 <planetmaker> the use is unspecified
09:04:37 <andythenorth> see train prop 28/29 docs, the implication is that it's AND
09:05:39 <andythenorth> vehicles should be providing an implicit AND by excluding all classes they don't support
09:05:47 <andythenorth> according to the spec
09:06:20 <andythenorth> it's a horrible way to set an AND, but probably the only smart way in an action 0
09:06:47 <peter1138> don't forget the XOR :D
09:06:59 <planetmaker> that's irrelevant as it's not a class thing
09:07:02 <andythenorth> the XOR is like lighting a bomb
09:07:15 <andythenorth> it ties classes and labels back together in the most horrible way :P
09:07:54 <andythenorth> it destroys the very practical abstraction that classes could have offered
09:08:17 <andythenorth> it doesn't pay any respect to 'stuff changes' :|
09:08:37 <planetmaker> it'd be easier to have classes. And labels as unconditional override
09:08:49 <andythenorth> that's what the cb route will do :)
09:09:01 <andythenorth> frosch solved this...2 years ago :P
09:09:24 <andythenorth> me and Pikka agreed with him...then did....nothing
09:24:44 *** TWerkhoven has joined #openttd
09:31:07 *** andythenorth has joined #openttd
09:33:56 * andythenorth wonders if 'may' and 'must' should be made explicit as two distinct cargo props
09:34:15 <andythenorth> (or as two return values from a 'what are my classes' cb on cargos)
09:34:47 *** Progman has joined #openttd
09:35:08 <andythenorth> for cases like refrigerated, covered, oversized, requires pressure discharge
09:35:35 <andythenorth> someone's going to invent 'must be shock proofed' at some point
09:36:08 <andythenorth> all of those are 'musts' in addition to the fundamental of whether the cargo is bulk, piece, liquid (and arguably neo-bulk)
09:36:40 <andythenorth> making it a cb would enable witty things, like milk needing refrigeration only after 1900 or whatever
09:36:46 <andythenorth> which is almost certainly a bad idea
09:36:48 <planetmaker> that sounds what I just wrote, andythenorth :-P
09:36:55 <andythenorth> sorry, was cycling :P
09:37:02 <planetmaker> no, in the FIRS issue
09:38:31 <andythenorth> if we can figure out the right answer, it could be a good time to make a 'big' change to this
09:38:45 <andythenorth> some of the older sets are quite dead anyway wrt FIRS, ECS support
09:38:49 <planetmaker> basically the conclusion of that is, that cargo classes are both, "and" and "or". Depending on which bits you look at
09:39:08 * andythenorth thinks we should non-violently deprecated the current methods
09:39:30 <andythenorth> basically I want to write 'FIRS will only work reliably with vehicle sets that implement a refitting cb'
09:39:57 <andythenorth> and then decide an interpretation of classes (AND or OR), write it firmly in the wiki and stick to it
09:40:18 <andythenorth> HEQS, FISH, OpenGFX can all be updted
09:40:40 <andythenorth> CETS, UKRS 2, egrvts 2, french set, dutch set, 2cc set are all being actively developed
09:40:46 <andythenorth> US set is being revived
09:40:52 <andythenorth> pikka may take over canset
09:41:08 <andythenorth> DB set will do whatever DB set will do, same for new ships 2
09:41:15 <andythenorth> japan set is being actively developed
09:43:26 <andythenorth> planetmaker: the excluding is a headache no?
09:44:07 <andythenorth> it makes a nonsense of the class system
09:44:51 <andythenorth> if I add 'vehicle must be red' I shouldn't rely on newgrf authors in the past excluding that bit in case I invent it
09:45:04 <planetmaker> like piece goods and not armoured makes certainly sense
09:45:38 <andythenorth> I wonder if that makes sense only because we're familiar with it
09:45:45 <planetmaker> But mostly it needs a clarification of how it's supposed to be used
09:46:29 <andythenorth> test case: if armoured didn't exist, and we added it now...
09:47:30 <andythenorth> option a) vehicles exclude that class explicitly because they don't know about it
09:47:39 <andythenorth> option b) vehicles allow that bit explicitly
09:47:50 <andythenorth> option c) vehicles neither allow nor exclude that bit
09:48:13 <andythenorth> which would have been the correct option - for a wagon that allows piece goods
09:50:35 * andythenorth suspects that (c) might be the right answer
09:50:50 <andythenorth> vehicle remains non-committal about classes the author didn't know about
09:51:28 <andythenorth> the sensible cargo author then sets piece goods and armoured
09:51:41 <andythenorth> their cargo gets transported by anything that refits piece goods
09:51:57 <andythenorth> they then ask / hope / beg for more accurate vehicle set support
09:52:19 <andythenorth> this suggests that classes are OR not AND
09:59:10 <andythenorth> if a vehicle grf declared which cargo classes it knew about...
10:01:35 <planetmaker> as the vehicle NewGRF decides about which it uses, it doesn't need to declare it
10:01:40 <planetmaker> it just allows cargos or not
10:02:11 <planetmaker> just allowing by cargo class certainly is the easiest approach. And it already works
10:03:51 <planetmaker> but a CB which allows things like allow if CC == bulk || (CC == piece,oversized) would solve all these issues of and or or
10:24:58 <andythenorth> planetmaker: so the cargo just treats all cargo as OR and the vehicle author decides? Works for me
10:25:38 <andythenorth> the proper domain of how to transport things is defined by the vehicles
10:25:57 <andythenorth> in which case we should (continue?) setting cargos as OR
10:26:18 <planetmaker> well... this has issues, too :-). But might be the best. As a vehicle set author knows best what his vehicle can transport
10:26:27 <andythenorth> what about cases like 'milk should be refrigerated after 1900' ?
10:26:36 <andythenorth> is that a bad case?
10:28:04 <planetmaker> adding a callback imho doesn't help the issue with classes and is beside the CC discussion
10:29:02 <planetmaker> even when that is a new and possibly better method we should use the most sane way of traditional CC without consideration of a CB
10:29:46 <planetmaker> i.e. let's not consider that, unless I'm given a good reason how it solves the original problem
10:30:36 <planetmaker> I'm missing a "." behind "method"
10:38:23 *** andythenorth has joined #openttd
11:07:40 *** frosch123 has joined #openttd
11:13:47 <planetmaker> hm, cargo classes are a bit. Are they "all apply" or "one or more apply" when set by a cargo?
11:13:58 <planetmaker> both makes somewhat sense
11:16:29 <andythenorth> think about it long enough...we'll find the answer :P
11:16:43 <andythenorth> I think I can summarise
11:17:09 <andythenorth> 1. if the new proposed cb is used by the vehicle, the final decision rests with the vehicle set author
11:17:34 <andythenorth> 2. we need to agree a convention for the 'traditional' method and try to enforce it
11:17:58 <andythenorth> if we ask Pikka, I'm 99% certain he's working to 'one or more apply'
11:18:35 <andythenorth> I am proposing for FIRS that we treat them as per your comments in the ticket, i.e. OR
11:18:56 <andythenorth> if vehicles choose to AND them, there's not much we can do about that
11:19:08 <andythenorth> "This is a bit mask of all cargo classes to which this cargo belongs, out of the following:"
11:19:28 <andythenorth> I read that as 'there are multiple sets which this cargo is a member of"
11:19:46 <andythenorth> rather than "the combination of cargo classes describes precisely one set which the cargo is a member of"
11:19:59 <planetmaker> andythenorth: the question is "is member of" or "can be member of"
11:20:22 <Eddi|zuHause> it wouldn't make any sense if a cargo is "bulk AND piece goods"... only "bulk OR piece goods"
11:20:27 <planetmaker> i.e. bulk + piece is mutually exclusive for one detailed representation. Still the cargo could be available as one OR the other
11:20:42 <andythenorth> Eddi|zuHause: it may not make sense, but that *is* the apparent spec
11:20:55 <planetmaker> Eddi|zuHause: it does make sense. Just not at the same time
11:21:13 <Celestar> Yexo: what about your fence helper? :)
11:21:14 <planetmaker> i.e. water could be piece or liquid (water in boxed bottles or in a tanker)
11:21:15 <Eddi|zuHause> planetmaker: hence "OR"
11:21:43 <andythenorth> but the box car should be excluding liquid
11:21:50 <andythenorth> according to the spec
11:21:54 <Yexo> Celestar: I hid some trouble, wasn't able to fix it soon enough and put it away for a bit. Than yesterday I accidentally deleted part of that code
11:22:00 <andythenorth> so water should not travel by box car. according to the spec
11:22:26 <andythenorth> if the box car doesn't exclude liquid, it becomes refittable for all liquids, which is odd
11:22:29 <Celestar> Yexo: you need your diff back? :P
11:22:38 <planetmaker> andythenorth: bottled it makes sense :-)
11:22:52 <andythenorth> planetmaker: you make sense. But you're not compliant with the spec :P
11:22:53 <Eddi|zuHause> andythenorth: no, "not exclude" is not the same as "include"
11:23:09 * andythenorth has logical flaw
11:23:17 <planetmaker> tri-state logic :-)
11:23:27 <planetmaker> yes / no / not-defined
11:23:28 <Eddi|zuHause> andythenorth: a box car would be refittable to cargos that are "piece goods OR liquid"
11:23:44 <Eddi|zuHause> andythenorth: but not to cargos that are purely "liquid"
11:23:51 <andythenorth> Eddi|zuHause: but cargos AND
11:23:56 <andythenorth> the spec is quite clear about it
11:24:06 <andythenorth> a cargo isn't piece goods OR liquid, it's piece goods AND liquid
11:24:11 <planetmaker> Eddi|zuHause: which means that water would need to get cargo class liquid and piece goods. In principle
11:24:13 <andythenorth> so you need a vehicle that supports both
11:24:22 <planetmaker> which means changing default cargos
11:24:38 <Eddi|zuHause> andythenorth: logical AND is a "common" OR, while logical OR is a "common" AND
11:24:40 <andythenorth> for a vehicle to support both, it must allow piece goods and liquid. This 'undefined' route doesn't work
11:24:53 <Eddi|zuHause> andythenorth: it's quite likely that the person who wrote the specs messed that up
11:25:09 <andythenorth> Eddi|zuHause: this is why my first suggestion this morning is that MB doesn't understand the cargo scheme
11:25:20 <andythenorth> for years we've been told it's us who have the problem
11:25:27 <andythenorth> and that the specs are unambiguous
11:25:43 <andythenorth> not a bit of doubt
11:26:24 <andythenorth> to transport food (which is express, refrigerated) you need a vehicle that is express AND refrigerated
11:26:45 <andythenorth> and you're non-refrigerated express vehicles must exclude refrigerated
11:27:03 <andythenorth> you're / your /s
11:27:20 <andythenorth> otherwise you fail spec
11:28:18 <Eddi|zuHause> andythenorth: proof by example?
11:28:56 <Eddi|zuHause> andythenorth: an example doesn't make a generic spec.
11:29:24 <andythenorth> Eddi|zuHause: shall we suggest that I think the spec is either (a) wrong or (b) incomplete or (c) just badly written?
11:29:27 <Eddi|zuHause> andythenorth: this paragraph allows no conclusion how to interpret "bitmask(CC_BULK, CC_PIECE_GOODS)"
11:29:39 <Eddi|zuHause> andythenorth: (b)
11:30:33 <Eddi|zuHause> my answer is complete.
11:30:42 <andythenorth> b XOR c, and if you know about specific sentences in the spec and put them in a bit mask...
11:32:39 *** TWerkhoven2 has joined #openttd
11:32:48 *** HerzogDeXtEr has joined #openttd
11:33:33 * andythenorth proposes that classes are OR, and it's up to vehicle owners what they want to do about that
11:33:51 <andythenorth> which will lead to inconsistency within a game between different sets (RVs, ships, trains etc).
11:33:54 <andythenorth> but at least is sane
11:34:14 <Yexo> classes that are OR lead to very unexpected results
11:34:16 *** Toshiba has joined #openttd
11:34:46 <Yexo> for example milk would be CC_LIQUED OR CC_REFRIGERATED
11:35:01 *** George|2 has joined #openttd
11:35:25 *** yorick_ has joined #openttd
11:35:39 <Yexo> which means you'd be able to transport it in a non-refrigerated tanker but also in a refrigerated trailer
11:36:16 <Yexo> milk in non-refrigerated tanker seems a very bad idea
11:36:50 <Yexo> it also breaks the new class "clean" completely
11:36:56 <Yexo> how can a cargo be either bulk or clean
11:37:00 <andythenorth> clean might yet be a bad idea
11:37:12 *** b_jonas_ has joined #openttd
11:37:14 *** __ln___ has joined #openttd
11:37:20 <andythenorth> Yexo: it helps if we forget the current methods and think only about the new cb proposed by frosch
11:37:32 <Yexo> andythenorth: what about tourist: are they either passengers or express?
11:37:35 <andythenorth> I am quite prepared to say "FIRS doesn't work well with older sets that don't use the cb"
11:38:14 <andythenorth> Yexo: tourists? I don't know.
11:38:14 <Yexo> the spec says: "This is a bit mask of all cargo classes to which this cargo belongs, out of the following: " which to me clearly indicates it's AND, not OR
11:38:27 <andythenorth> I was going to ask you what the spec said :P
11:38:34 <andythenorth> I'm bored of having to try and figure this out :D
11:38:37 <peter1138> sorry i was on the vehicle properties
11:38:48 <Yexo> for the vehilce properties it's indeed OR
11:38:54 <peter1138> the cargo belongs to all
11:39:02 <planetmaker> so... we always stick to all
11:39:04 <andythenorth> one set, or n sets?
11:39:07 <peter1138> that doesn't specify AND or OR
11:39:41 <Eddi|zuHause> i think there need to be two sets of cargo classes. "basic" categorization that is ORed (bulk, piece, liquid) and "specific" categorization that is ANDed (refg, express, armored, ...)
11:39:55 <peter1138> i think you need to just stop making everything overly complex
11:40:11 <planetmaker> though the specification would then need be specific to the "base" class
11:40:16 <planetmaker> At which point we're back to the CB
11:40:23 <andythenorth> I think we're always back to the cb
11:40:34 <peter1138> if (_gted[engine].cargo_allowed & cs->classes) SetBit(mask, cs->Index());
11:40:40 <peter1138> do you want that to read
11:40:41 <andythenorth> fwiw, last time discussed, me pikka and frosch seemed to think the cb solved this
11:40:44 <Yexo> if you force vehicle sets to use the cb you might as well force them to add specific support for firs
11:40:52 <peter1138> if ((_gted[engine].cargo_allowed & cs->classes) == cs->classes) SetBit(mask, cs->Index());
11:41:09 <peter1138> that would make all classes required, not just one
11:41:10 *** ChanServ sets mode: +o SmatZ
11:41:12 *** dihedral has joined #openttd
11:41:20 <Yexo> peter1138: that breaks backwards and ttdpatch compatibility
11:41:22 <peter1138> and probably break some existing stuff :)
11:41:24 <Yexo> so that's not really an option
11:41:45 <frosch> btw. there is one valid point about not adding a cargoclass "clean". it does not influence whether a vehicle can be refitted to it or not; only the costs. So the "clean" could as well be added to some other property. e.g. 15 "freight status"
11:41:54 <Eddi|zuHause> maybe we should completely revamp the system for grfv8?
11:42:03 <frosch> then there would also be room for "need cleaning before loading" and "needs cleaning after unloading"
11:43:04 <peter1138> Yexo, also making it and would make it useless for new stuff too
11:43:12 <peter1138> so what's the problem again? lol
11:45:02 <planetmaker> hm, valid point, frosch
11:45:20 <planetmaker> though it could be considered that.
11:45:22 <Eddi|zuHause> there are two separate problems: the existing classes cause too many exceptions (e.g. wood, steel, livestock), and trying to resolve these exceptions opened concerns about incompleteness of the specs
11:45:38 <planetmaker> like food wagons always contain a sign like "don't fill in chemicals" or similar
11:46:17 <Eddi|zuHause> DRG had two kinds of refrigerated wagons, one for general food, and one for fish
11:46:32 <Eddi|zuHause> they exclude each other
11:47:33 <andythenorth> frosch: new property for things that are 'must' rather than 'can'
11:47:50 <andythenorth> e.g. refrigerated, clean, covered etc
11:48:21 <andythenorth> I dislike a new prop for this
11:48:41 <Celestar> to me. "bulk" "liquid" "boxed" etc might be another property than, say, "refigerated"
11:49:08 <andythenorth> but adding more props might just make it worse not better
11:49:27 <peter1138> seems to me that the spec caters for "can't be done"
11:49:33 <andythenorth> refrigerated is a bad case
11:49:39 <peter1138> whereas you want something to cater for "isn't done"
11:49:44 <Celestar> andythenorth: probably ÖP
11:50:25 <peter1138> milk & oil could both go in the same tanker
11:50:46 <peter1138> why would you need "clean"?
11:50:55 <andythenorth> affects refit cost
11:50:58 <andythenorth> not refittability
11:51:01 <Celestar> IRL, not even sequentially
11:51:07 <peter1138> you have to refit to go from oil to milk and vice versa
11:51:21 <Celestar> food tankers may not carry anything else than food.
11:51:25 <andythenorth> clean might be a case better solved another way
11:51:36 <andythenorth> 'clean' started out as 'foodstuff'
11:51:46 <peter1138> Celestar, as i said, catering for "isn't done" rather than "can't be done"
11:52:00 <Celestar> Jet A-1 tankers may not carry anything else than Jet A-1.
11:52:20 <peter1138> disallow refit after it's been used :D
11:52:48 <peter1138> i still think it's being made overly complex
11:53:07 * andythenorth would be a big fan of having it simpler
11:53:11 <Celestar> hey, this is openttd :P
11:53:20 <Celestar> if you want it simple, play Quake :D
11:53:29 <andythenorth> if you want it simple, play Dope Wars
11:53:40 <andythenorth> or puzzle bobble
11:53:45 <Celestar> if (see_mob) shoot(mob);
11:54:17 <Celestar> I'd love to remove refit altogether :D
11:54:23 <andythenorth> refrigerated is a bad case, because refrigerated was not really available until a certain date
11:54:41 <Celestar> but have something like shunting yards :D
11:55:24 <peter1138> why is it a bad case?
11:55:34 <andythenorth> if food is refrigerated AND express, there won't be any food transport before ~some date
11:55:34 <peter1138> if it's not available, you just can't transport stuff that needs it
11:56:04 <Celestar> something like ... Maschen :D
11:56:13 <planetmaker> Maschendrahtzaun?
11:56:22 <peter1138> food is refrigerated or express
11:56:37 <peter1138> fruit is actually bulk or refrigerated ;D
11:56:57 <andythenorth> fruit == 'most cargos'
11:57:01 <peter1138> i'm talking about the default cargos
11:57:50 <andythenorth> so the prop28/29 stuff on the wiki *is* wrong
11:57:58 <peter1138> if you're making a food-type cargo, it makes sense to use the default cargo classes for that
11:58:12 <peter1138> what's wrong about it?
11:58:21 <andythenorth> my logic probably
11:58:30 <Eddi|zuHause> andythenorth: no, it's just _unspecified_
11:58:42 <peter1138> "match the classes" is perhaps vague
11:58:53 <peter1138> it's "match any of the classes" not "match all the classes"
11:59:11 * andythenorth is not clever enough
12:00:31 <andythenorth> english is vague
12:00:42 <andythenorth> but 'does not require refrigeration' seems quite clear to me
12:01:26 <Eddi|zuHause> andythenorth: it draws conclusions from some hypothetical "real world" that is not part of the specs, and open to interpretation
12:01:31 <peter1138> allowed AND NOT disallowed
12:01:56 <peter1138> allowed/disallowed is a list of cargos, not a list of classes
12:02:39 <peter1138> s/list/bitmask/g :)
12:02:57 <andythenorth> so for food, in the spec case. with two vehicles
12:03:13 <andythenorth> vehicle A is as per spec - it allows express, but excludes refigerated
12:03:19 <planetmaker> is over-sized and overweight = neo-bulk?
12:03:31 <andythenorth> vehicle B allows express, and does not include or exclude refrigerated
12:03:53 <andythenorth> planetmaker: maybe, maybe not. Depends on whoever invented it turning it up to clarify
12:04:09 <peter1138> vehicle A & B are what?
12:04:16 <andythenorth> whatever you like
12:04:19 <andythenorth> it's your set :)
12:04:37 <peter1138> what's the problem with them?
12:04:42 <andythenorth> (I'm just writing cargos, vehicle authors control what travels where)
12:05:09 <peter1138> vehicle allows express, but excludes refrigerated. therefore it cannot carry food.
12:05:32 <peter1138> vehicle B allows express, but does not care about refrigerated, therefore it can carry food.
12:05:44 <peter1138> (because food can be moved fast)
12:05:47 <Celestar> just if you guys think you build big stations:
12:05:59 <peter1138> never mind that it could be stuck on the map for 2 years :p
12:06:06 <andythenorth> so the industry set author should not be trying to enforce AND categories. It's none of their business
12:07:43 <andythenorth> so where's the problem? :P
12:08:52 <peter1138> that's what i keep saying ;)
12:10:01 <andythenorth> vehicle C: the author wants to only allow cargos that are express AND refrigerated
12:10:24 <andythenorth> exclude all other classes
12:12:08 <peter1138> what if it's hazardous food ? ;)
12:12:34 <andythenorth> like a giant burger?
12:12:52 <Eddi|zuHause> <frosch> btw. there is one valid point about not adding a cargoclass "clean". it does not influence whether a vehicle can be refitted to it or not; only the costs. So the "clean" could as well be added to some other property. e.g. 15 "freight status" <-- then that same should apply to "needs refrigeration" or "needs special protection"
12:12:55 <peter1138> hazardous, oversized, refrigerated, express
12:13:20 <peter1138> soylent green is people
12:13:44 <planetmaker> AND bulk. concurrently
12:13:46 <frosch> Eddi|zuHause: why? either a wagon is refridgerated or not
12:14:13 <Eddi|zuHause> frosch: but i _can_ put milk in a normal box van.
12:14:41 <Eddi|zuHause> frosch: i just can't move it for several days
12:14:52 <frosch> then milk should be classified as liquid + piece goods + refridgerated
12:15:17 <frosch> Eddi|zuHause: everything in ttd moves for several days :p
12:15:19 <planetmaker> frosch: and water as liquid + piece
12:15:47 <frosch> anyway, packaged milk and water is rather "goods"
12:15:48 <andythenorth> peter1138: you said 'no' to excluding above
12:15:50 <Eddi|zuHause> frosch: but the current interpretation is that it should be (liquid + piece)*refrigerated
12:15:55 <andythenorth> but what's the correct answer?
12:15:55 <frosch> so milk is no piece goods
12:16:24 <andythenorth> frosch: a wagon is either refrigerated, or not, or doesn't know
12:16:24 <frosch> you can package everything
12:16:35 <andythenorth> piece goods should be the default class :P
12:16:37 <frosch> so, just because something can be packaged, it is no piece goods
12:16:39 <andythenorth> everything can be packaged
12:16:51 <andythenorth> default refit all vehicles to anything that has class piece goods
12:16:59 <peter1138> andythenorth, it still allows express or refrigerated
12:17:01 <andythenorth> instant forward compatibility for all time
12:17:27 <frosch> the important fact is that you do not package milk when transporting it from the farm to the dairy
12:17:45 <frosch> while the dairy just produces food (packaged piece goods)
12:17:51 <frosch> milk (packaged) is nonsense
12:18:32 <andythenorth> I think anything that is predicated on a discussion of real cargos doesn't help
12:18:56 <andythenorth> if $someone wrote an unambiguous spec some of this would go away
12:19:30 <andythenorth> frosch: what is milk in churns if not packaged (real world examples don't help, there's always a counter-case)
12:20:14 <frosch> that is a pre-historic type of transportation that can be done for anything liquid
12:20:24 <frosch> so just allow boxvans to transport liquid before 1950
12:20:27 <andythenorth> but it's not pre-historic for ottd
12:20:38 <andythenorth> frosch: date based refits using a cb :D
12:20:44 <andythenorth> suddenly all your routes break :)
12:21:31 <frosch> that is actually a valid point about mb's scheme. early wagons can carry basically anything; while wagons introduced later are specialised
12:21:42 <Eddi|zuHause> andythenorth: the traditional way to do this is offer a new vehicle with fewer refit options, but other advantages (increased load, increased speed, etc.)
12:23:12 <andythenorth> in the case of vehicle C (author wants express AND refrigerated), they can't
12:24:18 <planetmaker> that's what you have the possibility to select individual cargos
12:25:23 <andythenorth> so the spec is firmly OR?
12:26:44 <andythenorth> when setting cargo classes on cargo prop 16?
12:26:52 <andythenorth> it's a list of sets to which the cargo belongs?
12:28:41 <frosch> cargoclasses of a cargo are OR
12:29:08 <andythenorth> so can I set 'refrigerated' as one of the classes on steel? :)
12:29:55 <andythenorth> I don't mean I propose to do this, I just mean, "it would be valid"
12:30:19 <andythenorth> but logically holds up
12:30:34 <frosch> a cargo that is defined as piecegoods or liquid can be transported by either tankers or fladbed wagons
12:31:03 <frosch> it is not the purpose of a vehicle to set the EXCLUDE-mask to the NOT-value of the INCLUDE-mask
12:32:06 <andythenorth> a cargo that is piecegoods or liquid or refrigerated can be transported by tankers, flatbed wagons, or any vehicle with refrigeration
12:32:37 <andythenorth> a cargo that is liquid or refrigerated can be transported by any tankers or any vehicle with refrigeration
12:32:41 <frosch> refridgeration is a class the is more useable for the EXCLUDE-mask, not for the INCLUDE-mask
12:32:52 <andythenorth> more usable maybe
12:33:00 <andythenorth> but to be logically strict about the spec...
12:33:10 <andythenorth> and free ourselves from 10m interpretations...
12:33:15 <frosch> not all classes can be treated the same
12:33:29 <frosch> "express" is espeically broken
12:33:30 <andythenorth> all animals are equal, but some are more equal?
12:34:08 <frosch> yes, horses and cows are equal; especially the cows
12:34:42 <andythenorth> frosch: how do you feel about a new prop?
12:34:52 <andythenorth> for 'additional cargo crap'
12:35:02 * andythenorth thinks it might be logical, but just as much of a mess
12:35:32 <frosch> cargo classes are for refittablity. period.
12:36:27 <planetmaker> so do we unify oversized and neo-bulk?
12:36:36 <andythenorth> only when we know what the intention was
12:37:01 <andythenorth> neo-bulk doesn't necessarily mean heavy or large
12:37:45 <andythenorth> not neatly packaged
12:38:09 <andythenorth> not going to flow nicely
12:38:24 <andythenorth> logs are the epitome of neo bulk
12:40:29 <peter1138> well, wood is anyway
12:40:48 <frosch> i wondered whether that "pouring" thingie is the opposite of "neo-bulk". i.e. whether every "bulk" cargo is either "pouring" or "neo-bulk"
12:41:36 <peter1138> it's stuff that's "pourable" in ttd-land
12:42:05 <peter1138> grain, wheat, maize
12:42:21 <andythenorth> frosch: every cargo is either bulk (dry, liquid) or break-bulk
12:42:29 <andythenorth> break-bulk = piece goods
12:42:30 <peter1138> sugar, toffee and cotton candy, yes
12:43:01 <andythenorth> neo-bulk is a sub-class of piece goods, for things that are basically shipped like bulk, but can't be poured
12:43:21 <andythenorth> I would leave bulk alone
12:43:31 <andythenorth> although probably lots of people will fall into the same trap as me
12:43:54 <andythenorth> they will think of 'bulk' as bulky, rather than spending an afternoon reading about the classifications of the international cargo industry
12:44:50 <frosch> well, if every bulk cargo is either pouring or neo-bulk, then one of those classes is too much
12:45:00 <planetmaker> andythenorth: it really never occured to me that bulk would mean the same as bulky...
12:45:09 <planetmaker> (yes, it doesn't)
12:45:15 <andythenorth> planetmaker: you are more germanic that mean :P
12:45:22 <andythenorth> you are probably more used to precision
12:45:26 <frosch> neo-bulk = bulk AND (NOT pouring); resp, pouring = bulk AND (NOT neo-bulk)
12:45:35 <andythenorth> and less used to spending half your life socially unpicking ambiguity :P
12:45:47 * andythenorth reads proposal by Eddi|zuHause
12:46:06 <andythenorth> I'd be quite prepared to break all previous classes, and basically go for a v2 of the schema
12:49:14 <planetmaker> Eddi|zuHause: that's a bit problematic, too: consider water: liquid, piece, clean (when transported in tanker, doesn't matter when palettized)
12:50:59 <Eddi|zuHause> planetmaker: possibly needs further refinement
12:52:59 <planetmaker> the longer I think about it, the better it seems to me to treat it as 'and'. And as such a cargo / industry set has to decide whether milk is liquid or piece goods
12:53:29 <planetmaker> vehicle sets then still can do the other thing, when they explicitly cater for milk
12:53:33 <peter1138> planetmaker, the default cargos wouldn't work
12:53:43 <andythenorth> milk is only piece goods when packed
12:53:47 <andythenorth> packing != cargo
12:53:51 <planetmaker> peter1138: how not?
12:53:52 <andythenorth> all cargos can be packed into piece goods
12:53:56 <peter1138> food would be express and refrigerated
12:54:34 <peter1138> right, now which sets define that?
12:54:40 <andythenorth> and not transportable in many sets until > 1900 or so
12:55:11 <planetmaker> the vehicle set decides itself anyway what it does with the cargo classes given for a cargo
12:55:28 <planetmaker> it's about how it most consistently can be used what we have
12:56:08 <andythenorth> when setting cargos, there's no code difference between AND or OR
12:56:21 <andythenorth> but we as industry set authors have to *make* a fricking decision
12:56:27 <andythenorth> and so there should be a convention for that
12:56:32 <andythenorth> currently it's all over the place
12:56:51 <andythenorth> but the actual refit is controlled by vehicle set, not industry
12:57:08 <peter1138> you still have some confusion?
12:57:23 <andythenorth> just waiting for everyone to agree on something
12:57:28 <andythenorth> and change a spec somewhere
12:57:35 <andythenorth> might take a while...
12:58:33 <peter1138> and changing it breaks anything already out there
12:58:42 <planetmaker> on how it should be used
12:58:49 <andythenorth> peter1138: by 'spec' you mean wiki, or code?
13:01:38 <peter1138> ok, i changed the wording to "match any of the classes"
13:01:41 <peter1138> what's ambiguous now?
13:05:50 <planetmaker> and what's the reason that and is not applicable?
13:06:16 <planetmaker> i.e. you just change the specs
13:06:31 <planetmaker> previously both arguments were similarily valid
13:06:42 <peter1138> to state what actually happens
13:06:58 <planetmaker> 'what happens' is what vehicle sets choose to make happen
13:07:04 <planetmaker> It's not what openttd does
13:07:20 <peter1138> ok, i dunno what you're talking about now :)
13:09:00 <planetmaker> errm... don't you think it should be clarified at the *cargo* definition rather? If so?
13:09:06 <planetmaker> What about the other vehicles?
13:09:48 <peter1138> i just clarified that bit of the spec, the bit which andythenorth kept pasting around
13:10:10 <andythenorth> I made a picture of it :)
13:10:29 <andythenorth> from a cargo author's point of view, the classes *must* be treated as 'any of these'
13:10:42 <frosch> andythenorth: that depends on the class
13:10:55 <andythenorth> a class is a class
13:11:07 <frosch> it might be OR for piece and bulk; but it is AND for refridgerated and covered/sheltered
13:11:26 <andythenorth> peter1138 proved to me earlier that you can't AND
13:11:39 <andythenorth> there is no way to do it on a vehicle refit
13:11:46 <planetmaker> cargo: piece + liquid + refrigerate
13:11:55 <frosch> there is a difference when you look from the vehicle pov, compared to the cargo pov
13:12:02 <andythenorth> are we describing current spec now, or as it should be in an ideal world?
13:12:04 <planetmaker> vehicle A: refit to piece and not to refrigerate --> don't carry cargo
13:12:16 <planetmaker> vehicle B: refit to piece --> allow cargo
13:12:23 <frosch> some classes are only meaningful for the INCLUDE-mask, some only for the EXCLUDE-mask, some for both
13:12:36 <planetmaker> vehicle C: refit to liquid -> allow cargo
13:13:00 <planetmaker> vehicle D: refit to refrigerate and NOT piece --> disallow?
13:13:15 <planetmaker> thus piece makes no sense in the disallow
13:13:22 <planetmaker> nor does liquid or bulk
13:13:26 <andythenorth> simple isn't it :P
13:13:32 <planetmaker> but observing that: all is fine
13:13:40 <planetmaker> otherwise: vehicle grf error
13:14:08 <planetmaker> maybe one could make two further columns for the classes table: use in vehicle allow. use in vehicle disallow
13:15:28 <andythenorth> that's why I wondered about two properties
13:15:35 <peter1138> can you try explaining it in clear english, cos sometimes these "X A B -> Y" things don't mean anything
13:16:07 <andythenorth> peter1138: it's too easy for cargo authors to make it easy for vehicle authors to do bad things
13:17:14 <andythenorth> if I understood maths, I'd talk about this in terms of sets :P
13:19:13 <andythenorth> planetmaker: are there cases where bulk / piece / liquid should ever be excluded?
13:19:56 <planetmaker> those three should never be excluded it seems
13:20:28 <andythenorth> shuffle the bits around, and reduce the size of the exclude property :P
13:21:29 <frosch> maybe we should extent the cargoclasses to support 1024 of them
13:21:35 <frosch> then every cargo can have its own class
13:24:38 * andythenorth excludes cargos in HEQS
13:24:45 <andythenorth> possibly that is a mistake
13:24:50 <peter1138> just do what people wanted for aircraft
13:26:39 <andythenorth> why do I disallow stuff?
13:26:42 <andythenorth> I must be a dumbass :P
13:27:09 <andythenorth> why do mining trucks need to care about piece goods?
13:27:13 <andythenorth> just don't include them :P
13:34:48 <andythenorth> makes no bloody difference if I set exclude to 00
13:34:58 <andythenorth> except now they refit to FICR
13:36:45 *** Oneiric_Soul has joined #openttd
13:36:56 <andythenorth> classes are piece + bulk
13:37:01 <andythenorth> so the behaviour is correct
13:37:12 <andythenorth> if it was marked as 'clean' I could exclude it :P
13:37:16 <andythenorth> from mining trucks
13:37:21 <peter1138> but does piece & bulk make sense?
13:37:34 <andythenorth> I don't know, it's not my decision
13:37:49 <andythenorth> MB controls FICR
13:37:59 <planetmaker> peter1138: yes, it does
13:38:04 <planetmaker> wood + wood chips
13:38:22 <andythenorth> planetmaker: I have avoided discussing wood versus wdpr for three days
13:38:26 <andythenorth> now is not the time :)
13:39:13 <peter1138> but fibre crops *could* be transported in a mining truck
13:39:17 <peter1138> you just wouldn't do it
13:39:39 <supermop_> people could also sit in the back
13:39:41 <andythenorth> the solution is already known
13:39:45 <peter1138> disallow it via the label, yeah
13:39:50 <andythenorth> I use the label to exclude grain, wheat and maize
13:39:55 <andythenorth> that's what labels are for
13:40:17 <peter1138> that just means the label needs to be in the lower 32 entries of the translation table
13:40:32 <andythenorth> which is a bit of an arse
13:40:47 <andythenorth> frosch: the cb would be able to access >32 ctt entries?
13:41:00 <Eddi|zuHause> andythenorth: if you take my proposal, a mining truck would exclude "light" cargos
13:41:12 <planetmaker> yes. But feasible. ogfx+ uses 21 cargos in explicit refits currently
13:41:28 <peter1138> it's only the refit mask that's limited to 3
13:41:55 <andythenorth> so, where's the problem?
13:42:10 <andythenorth> point and laugh at people who use bulk, piece or liquid in exclude
13:42:26 <andythenorth> encourage the use of a cb to overcome CTT limits
13:42:35 <andythenorth> find out wtf oversized is supposed to be
13:42:46 <andythenorth> set cargo authors straight about AND / OR
13:42:58 <andythenorth> my work here is done :P
13:43:22 * andythenorth thinks neo-bulk should be a core class like bulk and piece, but....hysterical raisins
13:43:38 <supermop_> needs a better name
13:44:08 <Eddi|zuHause> andythenorth: i disagree there
13:44:40 <Eddi|zuHause> andythenorth: neo-bulk is about how to handle the cargo, and thus is a *specialized* property
13:45:24 <Celestar> Why is it so difficult for people to understand the causilty principle?
13:46:15 <Celestar> I just spent 90minutes in a friggen meeting to explain to some fsckwit that I cannot change the CAUSE of something when I encounter its EFFECT.
13:46:42 <Celestar> no amount of mandays they give me will enable me to break relativity :P
13:47:35 <planetmaker> I agree with Eddi|zuHause here. bulk+piece+liquid are 'core' classes. neo-bulk just describes the piece-good. It's also in the transportation stuff I found classified as a sub-category to piece goods
13:47:41 <peter1138> yeah but got anywhere with tile layers?
13:47:58 <planetmaker> and pikka also has a point: any cargo should define at least one of the 6 original classes
13:48:09 <Celestar> peter1138: played around with benchmarking/performance a bit more.
13:48:51 *** Kurimus has joined #openttd
13:48:56 <peter1138> i guess you're exclusing passengers & mail
13:49:03 *** Progman_ has joined #openttd
13:49:10 <Celestar> that's unexclusable ....
13:49:42 *** Progman_ is now known as Progman
13:51:56 <Celestar> michi_cc: does your code call GetTileSlope way more often than the original one?
13:52:48 <andythenorth> planetmaker: ok I'm convinced on neo-bulk
13:52:55 <andythenorth> it's a subclass of break-bulk (piece)
13:54:49 <Celestar> michi_cc: I'm loading the same map in trunk, and your repo, run it for 10k ticks. Your code calls GetTileSlope 153426028 times, while trunk calls it 6306034 times.
13:54:59 <Celestar> that's a factor of 25 which I don't understand
13:59:31 <Terkhen> hmmm... lots of long conversations lately
14:00:20 *** Progman_ has joined #openttd
14:01:26 <Yexo> Celestar: did you check afterload?
14:06:23 *** Progman_ is now known as Progman
14:08:50 *** DayDreamer has joined #openttd
14:11:05 *** Neosublimation has joined #openttd
14:11:37 *** Eddi|zuHause2 has joined #openttd
14:15:04 *** SirSquidness has joined #openttd
14:18:22 <frosch> Terkhen: don't judge conversations by their length; but by their results :p
14:21:24 *** SirSquidness has joined #openttd
14:22:14 <andythenorth> MB proposes moving to wagon classes
14:22:17 <andythenorth> which is not a new idea
14:22:21 <andythenorth> but he has worked it through
14:23:23 <planetmaker> for certain definitions of 'worked through'
14:23:35 *** Adambean has joined #openttd
14:27:28 <andythenorth> so...we could keep the current scheme
14:27:36 <andythenorth> adopt the scheme from Eddi|zuHause2
14:27:41 <andythenorth> or adopt the scheme from MB
14:27:54 *** Eddi|zuHause2 is now known as Eddi|zuHause
14:28:00 <andythenorth> or keep the current scheme, but enforce a split between 'core type' and 'requires'
14:28:20 <Eddi|zuHause> i spot some flaws in MBs scheme...
14:28:27 * planetmaker currently prefers 'do nothing'
14:28:39 <planetmaker> it was all clarified ;-)
14:28:52 <Eddi|zuHause> e.g. the "L" category, which covers various completely different things
14:28:52 <planetmaker> it has its limitations. But every scheme has that
14:31:24 <Eddi|zuHause> PS: in MBs scheme, mail vans should be "D"
14:34:48 <frosch> afternoon sir belugas :)
14:46:04 <andythenorth> what are the problems with 'express' as a concept? Multiple people have reservations about it
14:47:30 <frosch> it's one of those classes which are only suitable for "exclude"-classes
14:48:05 <frosch> so, if your set has two wagons carrying piece goods you can set the express-wagon to carry bulk, and the non-express wagon to carry bulk except express
14:48:16 <frosch> but which set actually has these classes?
14:48:26 <frosch> another problem is the default cargo "goods"
14:48:49 <frosch> it has only the "express class", so technically you would have to use the class in a INCLUDE-mask unless you handle it explicitly
14:48:51 <andythenorth> goods is express? :o
14:48:55 <frosch> but, INCLUDE makes no sense :p
14:49:24 <frosch> andythenorth: the default goods is some cargo, which is produced at refineries, saw mills and factories
14:49:35 <andythenorth> clearly needs express :P
14:50:14 <frosch> maybe it should have been liquid + piece goods + express
14:51:47 <planetmaker> sounds like 'yes'
14:55:09 <andythenorth> frosch: this could be changed - it's additions not removals?
14:55:55 <frosch> you can only add new defined classes
14:56:05 <andythenorth> it's only going to cause lots of horrible exclusion problems :)
14:56:10 <andythenorth> no-one will notice
14:56:28 <planetmaker> and result in all sets not carrying goods any longer?
14:56:55 <frosch> andythenorth: just leave it as an exception
14:57:04 <andythenorth> tidy mind problem
14:57:06 <frosch> in fact we might need to do that anyway for all existing cargos
14:57:35 <frosch> cargo classes are for unknown cargos after all
14:57:54 <frosch> so, classes of known cargos do not matter actually :p
14:57:55 <andythenorth> goods doesn't practically cause problems
14:58:26 <planetmaker> frosch: ehm... don't matter?
14:59:12 <frosch> planetmaker: yes. you can always get away with not changing the classes
14:59:16 <andythenorth> I think he means, 'issues regarding goods can be handled via the label' ?
14:59:23 <planetmaker> but it might be nicer to have not a xor existing cargo label but two properties like "allow label" and "disallow label"
14:59:56 <frosch> the cb is already decided, isn't it?
14:59:56 <peter1138> instead of a refit mask, a list of cargos from the translation table ;)
15:00:09 <peter1138> cb makes more sense :)
15:00:13 <andythenorth> frosch: I'm banking on the cb
15:00:45 <frosch> currently the topic is only to reduce "oversized", "neo bulk" and "pourable" to a single class
15:00:52 <andythenorth> cb relies on vehicle authors to Do The Right Thing
15:01:07 <frosch> and whether to use "clean" as class, or whether to put it in a different property similar to "freight"
15:01:22 <andythenorth> frosch: it smells wrong as a class
15:01:31 <andythenorth> it's a nice hack, but...
15:01:45 <planetmaker> well... it's a cargo property
15:01:57 <planetmaker> I still think it's useful to refit a wagon from coal to grain
15:02:30 <andythenorth> is clean an attribute or a requirement?
15:02:39 <andythenorth> does coal need clean?
15:02:56 <andythenorth> if the previous cargo was something like mud
15:03:09 <planetmaker> dirt + dirt = still dirt
15:03:37 <planetmaker> grain is clean. fruit are clean. grain and fruit give some kind of food. But it's not clean anymore
15:04:00 <planetmaker> or tell the grain mill that processing sugar beet is about the same as oat
15:04:18 <andythenorth> it was intended to show 'needs clean vehicle' rather than 'is clean cargo'
15:04:18 <frosch> maybe "clean" is the wrong property
15:04:29 <andythenorth> foodstuffs was less ambiguous initially
15:04:32 <andythenorth> but the scope crept
15:04:33 <frosch> maybe it should be "easily cleanable" vs. "hard cleanable"
15:04:51 <frosch> you also have to clean the vehicle when refitting from oil to chemical
15:05:00 <frosch> though maybe that holds for all liquid stuff
15:05:26 <andythenorth> set a base cost for cleaning the vehicle to the next cargo?
15:05:46 <planetmaker> that doesn't make sense
15:05:48 <Eddi|zuHause> <frosch> in fact we might need to do that anyway for all existing cargos <-- if the purpose was to reduce the number of exceptions, then we hereby failed
15:05:53 <frosch> or to word it differently: does "clean" mean the cargo needs a clean environment before loaded, or does it mean "it does not make it dirty for the next cargo"?
15:06:06 <planetmaker> or rather: we have that. It's refit cost
15:06:09 <andythenorth> needs a clean environment (was the intention)
15:06:34 <planetmaker> frosch: I interpreted it as "needs (somewhat) clean environment"
15:06:37 <frosch> Eddi|zuHause: the goal is to require no execptions for future cargos, not for existing ones
15:06:50 <Eddi|zuHause> frosch: i disagree.
15:06:57 <frosch> planetmaker: so both chemicals and milk need a clean environment
15:07:09 <frosch> so refitting from chemicals to milk is fine?
15:07:21 <Eddi|zuHause> frosch: someone writing a new set, should have to care the least bit about hysterical exceptions
15:07:24 <planetmaker> :-) Chemicals are dirty :-P
15:07:31 <andythenorth> Eddi|zuHause: they just need to use labels
15:07:43 <planetmaker> Though of course that's problematical
15:07:48 <frosch> planetmaker: so refitting from oil to chemicals is fine?
15:07:53 <andythenorth> Eddi|zuHause or create a case for cargo scheme v2, clean sheet of paper, break all old set
15:08:10 <planetmaker> frosch: yes, that's what I implemented
15:08:16 <planetmaker> oil is just a chemical
15:08:16 <andythenorth> possibly there are enough new sets being worked on, and enough exciting 1.2 stuff to get away with a v2
15:08:32 <andythenorth> or tie this to ottd 2
15:08:40 <planetmaker> With the distinction: it needs some cleaning. But not a totally thorough
15:08:53 <planetmaker> (i.e. refit is not free, but feasible in a station)
15:08:54 <peter1138> water is a chemical
15:08:57 * andythenorth liked the idea of reimplementing the entire 'default' economy in newgrf :P
15:09:00 <Eddi|zuHause> andythenorth: there should never be an ottd2, unless it's a complete rewrite...
15:09:23 <andythenorth> Eddi|zuHause: well you're proposing to rewrite cargo schema.. just start there and work outwards
15:09:46 <Eddi|zuHause> andythenorth: everything that is an incremental edit, will just result in another 1.x
15:11:24 <andythenorth> water is a chemical :o
15:11:29 * andythenorth is not drinking water any more
15:11:37 <andythenorth> it's full of chemicals :o
15:11:49 <Eddi|zuHause> andythenorth: air is a chemical
15:11:51 <planetmaker> dihydregenmonoxide is the chemical which kills most people on the world actually
15:12:10 <andythenorth> 'may contain trace elements of mineral compounds'
15:12:27 <andythenorth> now write it on the side of a bottle of spring water as 'contains trace elements of mineral compounds'
15:12:39 <andythenorth> one will stop people drinking water, the other will make you a lot of money
15:12:55 <Eddi|zuHause> "berlusconi expected to resign within the next few hours"
15:13:23 <Eddi|zuHause> "mineral water: may contain traces of minerals"?
15:13:43 *** Celestar has joined #openttd
15:13:48 <planetmaker> may contain traces of sodium chloride and fluoride
15:14:20 <Eddi|zuHause> Celestar: i thought we discussed that the last time? :p
15:14:24 <frosch> [16:11] * andythenorth is not drinking water any more <- you know what fishes do in water?
15:14:47 <andythenorth> I've heard about that
15:15:04 <Celestar> Eddi|zuHause: discussed what?
15:15:19 <Celestar> Yexo: no, the call tree sais different.
15:15:25 <Eddi|zuHause> Celestar: the effect of the savegame conversion on profiling
15:15:42 <Celestar> Yexo: 150M out of the 153M calls are from GetFoundationSlope
15:16:28 <Celestar> Yexo: which itself, is called 50 times as often ...
15:17:31 <Celestar> Yexo: aha. in trunk, GetFoundationSlope is not called from TileLoop_Clear. In your code, it is. 150M times.
15:20:50 <planetmaker> must be the similarity of nick names :-P
15:22:31 <Eddi|zuHause> the keys are like right next to each other :p
15:22:59 <peter1138> can clear tiles even have a foundation? o_O
15:23:06 <Celestar> can MP_CLEAR have foundations anyway?
15:23:52 <Eddi|zuHause> well, if you turn MP_CLEAR into "everything underneath a rail/road tile", then yes
15:24:01 <frosch> yes, if the clear tiles are sloped and have a tunnel directly under them
15:24:27 <b_jonas_> I mean, when you replace rail tracks with monorail with the rail upgrade tool, there must be an in between state when the foundations aren't changed but there's no rail on the square
15:24:32 *** b_jonas_ is now known as b_jonas
15:25:08 <b_jonas> frosch: hmm, how does that work?
15:25:44 <Eddi|zuHause> Celestar: it probably wouldn't make a lot of sense to encode the foundation separately into each stacked MP_RAIL/MP_ROAD tile, only into the basic MP_CLEAR tile
15:26:31 <Eddi|zuHause> (and possibly MP_BRIDGE/MP_TUNNEL)
15:26:49 <Celestar> Eddi|zuHause: that's a performance hit.
15:27:20 <Eddi|zuHause> Celestar: then maybe a new tile class MP_OCCUPIED is necessary?
15:27:25 <peter1138> and 150M calls of GetFoundationSlope isn't? :D
15:27:52 <Celestar> peter1138: that's what I mean :P
15:27:58 <Eddi|zuHause> Celestar: but then again you scrap the possibility of having cliffs
15:28:12 *** Biolunar has joined #openttd
15:28:34 <Celestar> Eddi|zuHause: what peter1138 said :P
15:29:15 <Eddi|zuHause> andythenorth: how is that clarifying anything?
15:29:32 <andythenorth> improve it then ;)
15:29:37 <andythenorth> let's assume it's GPL
15:29:40 <Celestar> but assuming that about 95% of the clear tiles do not have foundations, we either make the call to it MUCH cheaper (buffering) or skip it.
15:29:51 <Eddi|zuHause> andythenorth: what exactly is it supposed to clarify?
15:29:52 <andythenorth> Eddi|zuHause: you can even have the source, although it's a mac app :P
15:30:05 <Celestar> Eddi|zuHause: and if you store the slope (or all 4 points) directly in the struct, there's no need for it anyway.
15:30:19 <andythenorth> Eddi|zuHause when you set a class on a cargo, which set(s) it's going into
15:32:03 <andythenorth> not sure my diagram is mathematically correct :P
15:32:11 <andythenorth> it's a long time since I did venn diagrams
15:38:50 <michi_cc> Celestar: Did I tell you already that it is non-optimized code? :)
15:39:44 <Celestar> michi_cc: yeah. I'm helping :P or attempting to :D
15:40:26 <Eddi|zuHause> Celestar: storing the foundation explicitly is probably the most sensible solution
15:40:46 <Celestar> it also gives the highest flexibility.
15:40:57 <Celestar> and the least hassle when attempting to retrieve it :P
15:41:17 <Celestar> michi_cc: I just really like that idea ^^
15:56:58 *** TWerkhoven has joined #openttd
16:06:31 <Eddi|zuHause> man that was DECADES ago!
16:12:06 <Elukka> really, really low orbit
16:13:08 <peter1138> i need to update...
16:13:14 <MNIM> yeah, that game evolves fast, last time I played it it had no moon yet
16:13:14 <Elukka> you need the experimental version
16:13:30 <Elukka> kerbin rotates and orbits the sun now, and there's the Mun
16:13:40 <MNIM> since the game runs in windows, and I haven't booted windows in ~2months...
16:14:04 <Elukka> it's awesome to do a translunar injection burn, hit 50x time compression and watch the planet get smaller
16:14:42 <Elukka> there's a trajectory map now which makes orbital transfers a lot easier
16:15:42 <peter1138> celestar & michi_cc to finish ;)
16:15:50 <MNIM> when I last played it you still had to eyeball orbit and compare it to orbital charts, lol
16:15:55 <Elukka> the rocket that took me there
16:16:42 <MNIM> yeah, this version has rcs already instead of the SAS
16:17:10 <MNIM> sickness avoidance system... except that it caused more motion sickness than it prevented, lol
16:17:11 <peter1138> shame it's windows only :(
16:17:27 <Elukka> it lifts about 3,5 full fuel tanks to low orbit
16:17:32 <Elukka> well there's a mac version :P
16:17:40 <peter1138> who the hell uses them?
16:17:42 <z-MaTRiX> i have just read windows will not like if you rin it off a fast ssd
16:18:11 <z-MaTRiX> there is too little io access delay
16:18:46 <MNIM> I should still get around to putting a vm on this box.
16:19:28 <Eddi|zuHause> is that like hobby-versions of pacman running too fast, because the guy programmed it to run on his 286 and didn't account for increasing CPU speeds?
16:19:53 <z-MaTRiX> Eddi|zuHause<< exactly ;)
16:19:59 <MNIM> not that Im better, but still
16:20:08 <z-MaTRiX> and guys blame it on the SSD
16:20:21 <Eddi|zuHause> even some professional games suffered from this. e.g. north&south
16:20:32 <peter1138> maybe you'd understand [citation needed] ?
16:20:39 <Eddi|zuHause> (man i'm seriously getting old)
16:20:40 <z-MaTRiX> well its on an online store as comment and in hungarian language
16:21:03 <z-MaTRiX> OCZ Agility 3 SATA III 2.5" SSD 60GB (AGT3-25SAT3-60G)
16:21:05 <peter1138> so utter tripe anyway
16:21:19 <z-MaTRiX> and its getting cheap
16:21:42 <Eddi|zuHause> wtf am i gonna do with a 60GB disk?
16:22:27 <z-MaTRiX> its SATA III 525 MB/s read/475MB/s write ;>
16:22:51 <z-MaTRiX> faster than my sd-ram was in 2001
16:23:24 <z-MaTRiX> there is 2TB one too, but it has a price
16:23:58 <z-MaTRiX> its faster, as a PCI-Express x4 card
16:25:21 <andythenorth> Eddi|zuHause: you give a 60GB disk to someone else that's what you do with it
16:25:48 <andythenorth> wait until the baby pictures kick in, then you start thinking "what is the biggest disk they make, and why don't I have it"?
16:26:34 <peter1138> that would've been about 1GB/s
16:26:46 <Eddi|zuHause> andythenorth: i have a 2TB, a 1TB, a 500GB and a 400GB disk currently
16:26:52 <peter1138> unless it was PC-66... urgh :p
16:27:18 <peter1138> that would've been 1.6GB/s
16:27:35 <peter1138> assuming crappy original
16:27:40 <z-MaTRiX> ok, low end SD-ram is about 290MB/s
16:28:14 <peter1138> that's about the speed of EDO
16:29:34 <z-MaTRiX> i didnt have EDO ram
16:29:45 <z-MaTRiX> but i do have some computers with SD-ram
16:32:34 <z-MaTRiX> single channel DDR1 400MHz goes at ~1.6GB/s with memtest
16:33:10 *** rhaeder has joined #openttd
16:36:20 <z-MaTRiX> but i dont believe you can get >600MB/s out of an SD-ram
16:38:14 <peter1138> theoretical speed, yeah
16:44:53 <Belugas> stomac growls. freaking time change :S
16:45:31 <Rubidium> Belugas: oh, I'm not over it myself yet either :(
16:45:53 <Rubidium> it's *so* annoying to wake up long before your alarm is set to go off
16:46:19 <peter1138> more time for openttd deving :D
16:47:00 <Belugas> indeed, Rubidium... indeed
16:47:56 <Belugas> i worked out late on the basement yesterday ( secretly building my speaker cabinet ;) ) so i slept late. my son jumped on the bed at 4:30. like... GET OUT!!!!
16:55:12 <Eddi|zuHause> Rubidium: but time change was over a week ago :p
16:55:30 <Rubidium> yeah... sad, ain't it?
16:56:17 *** Prof_Frink has joined #openttd
17:01:41 <Eddi|zuHause> but i must admit, my sleeping pattern also "shifted" since then
17:02:51 <frosch> yup, i get hardly out of bed before 10 am :p
17:08:05 *** |Jeroen| has joined #openttd
17:08:22 <andythenorth> clear indicator of a no-child-at-home status
17:08:39 <andythenorth> I bet Belugas has been awake at every 10am for several years
17:08:56 <andythenorth> probably maybe 1200 x10am for Beluga consecutively
17:11:06 *** TGYoshi has joined #openttd
17:12:49 <TGYoshi> just decided to try out openttd again :p
17:13:04 <TGYoshi> Lets figure out how to make opengfx work
17:13:24 <planetmaker> use the installer..
17:13:51 <planetmaker> if it works already: use online content download and select it in the game options
17:14:08 <TGYoshi> putting in data/ folder
17:14:28 <TGYoshi> Im using the nightly version
17:15:42 <TGYoshi> Except for openmsx :3
17:16:13 <planetmaker> get that via ingame content download
17:18:23 <TGYoshi> You probably know some other must-have new-gfx things ;p
17:19:46 <Belugas> i cannot remember last time i've been able to sleep past 7:00h am... my wife is the kind "get up, we have millions of thngs to do" Even before my kid... nf we switched time only this weekend, in order to match our big neighborous...
17:19:49 <planetmaker> neither is new-gfx. Both are base sets and kinda essentials
17:21:46 <TGYoshi> sure, I mean those add-on things. I remember some quite huge extension to industries etc.
17:22:47 <planetmaker> it's hard to recommend anything. Tastes vary wildly.
17:23:00 <planetmaker> Again I'd like to direct you to the online content ;-)
17:23:34 <planetmaker> Test out a few, best individually and use those which you like - in a new game :-)
17:23:57 <planetmaker> just when using an industry set, make sure you use vehicle sets, which usually support also newly introduced cargos
17:25:14 <TGYoshi> Ill first try the original game
17:25:48 <TGYoshi> Anyway, where is the inverse mouse option? XD
17:26:10 <planetmaker> somewhere in the adv. settings
17:26:19 <TGYoshi> Expectable. Somewhere
17:26:45 <planetmaker> surely not in the 'economy' branch ;-)
17:29:29 <TGYoshi> Reverse scroll direction
17:29:54 <TGYoshi> quite weird how my mouse sticks tho :P
17:34:23 *** sla_ro|master has joined #openttd
17:41:09 *** andythenorth has joined #openttd
17:46:11 * andythenorth wonders how well the wagon code system would apply to HEQS
17:47:20 <andythenorth> are mining trucks "ordinary open high- and low-sided wagons" (class E)
17:47:38 <andythenorth> or "special wagons: silo wagons, low-loaders" (class U)
17:48:02 <andythenorth> or "special open high-sided wagons" (class F)
17:48:16 *** LordAro has joined #openttd
17:48:23 <frosch> call the manufacturer and ask
17:48:38 <andythenorth> I have him on the line now
17:48:47 <andythenorth> he's wondering how the class codes apply
17:49:05 *** Brianetta has joined #openttd
17:49:08 <andythenorth> my shipping magnate wants to know too
17:49:45 <andythenorth> probably ships might be safely described by just attaching all classes
17:50:30 <andythenorth> what about planes?
17:50:45 <andythenorth> probably class G "ordinary covered wagons, box cars, vans"
18:08:45 <Elu> "It is necessary to introduce a high speed freight service on the route because the wind turbulence generated by a passing Shinkansen has the ability to derail a freight train which can be avoided if the freight train travels at a similar speed to the Shinkansen."
18:18:51 <appe> you people know this more then me, but japan really seems to be the government interested in trains. and good trains, too.
18:19:28 *** TheMask96 has joined #openttd
18:39:53 *** mahmoud has joined #openttd
18:45:15 <CIA-6> OpenTTD: translators * r23128 /trunk/src/lang/vietnamese.txt:
18:45:15 <CIA-6> OpenTTD: -Update from WebTranslator v3.0:
18:45:15 <CIA-6> OpenTTD: vietnamese - 14 changes by nglekhoi
18:53:56 *** andythenorth has joined #openttd
19:12:28 *** JVassie has joined #openttd
19:16:40 <andythenorth> planetmaker frosch I'm going to summarise some cargo class thoughts for a forum post...
19:17:28 <andythenorth> 1. there could be alternatives (eddi, MB), but the class system + cb is one option
19:17:47 <andythenorth> 2. cargo authors need to treat cargo classes as OR
19:18:20 <andythenorth> 3. cargo authors should not be concerned about what vehicles carry what, only what the intrinsic properties of the cargo are
19:20:42 <andythenorth> 4. setting lots of classes is probably unwise (I can't prove this, but I think it will hold)
19:21:45 <andythenorth> [small interruption, the toddler just weed on the floor]
19:22:16 <planetmaker> also shit happens ;-)
19:23:38 <andythenorth> 5. freight cargos should set at least one of: bulk, piece, liquid; this ensures widespread cargo support
19:24:54 <andythenorth> 6. don't use real-life transport examples as a guide to the classes, use the intrinsic properties. Milk can travel in churns, oil can travel in barrels, water can travel in bottles, but all are properly liquid not piece goods. If a vehicle set author wants to allow liquid transport by box van (liquid in containers), it's up to them
19:26:03 <andythenorth> 7. vehicle set authors should include at least one of: bulk, piece, liquid; exceptions for special vehicles or label-based support only
19:26:37 <planetmaker> I don't think a guide is needed there
19:26:38 <andythenorth> 8. vehicle set authors should not exclude: bulk, piece, liquid; this is likely to lead to poor support for future cargos
19:27:31 <andythenorth> 9. labels and explicit support are the solution to awkward cargos (highly specific cases)
19:27:53 <utn> can't you play against pc users when you got a mac?
19:29:15 <andythenorth> 10. vehicle set authors should only exclude the known classes (i.e only set the known class bits). Setting a general exclude may damage the abstraction in future
19:29:42 <Rubidium> utn: there's no such limitation
19:30:10 <andythenorth> I could add some general stuff, like 'final decision on transportation rests with vehicle author'
19:30:13 <utn> okay i will try reinstall it then, thx
19:30:24 <Rubidium> utn: except that you need to use the same version of OpenTTD, e.g all should use 1.1.3 (regardless of whether it's a Mac OS X, Windows or Linux binary)
19:30:37 <andythenorth> and 'industry authors shouldn't waste time trying to force certain refits on older vehicle sets by dicking around with classes improperly'
19:31:48 <andythenorth> and "it may be appealing to you (seem correct) to exclude refrigerated cargo from *every* non-refrigerated wagon, but that's liable to blow up in someone's face" etc
19:33:31 <appe> i think im getting the hang on track effectivity
19:33:50 <appe> adding a longer turn made my year round profit 20% higher
19:34:04 <appe> i do not want to know how people do this in real life.
19:35:08 <CIA-6> OpenTTD: michi_cc * r23129 /trunk/src/ (55 files in 4 dirs): -Add: [NewGRF] Property for the rail type name.
19:37:04 <andythenorth> planetmaker: if you agree with above, I'd like to review FIRS cargos more carefully + methodically :)
19:45:14 <planetmaker> as said, I'd skip 7
19:51:05 <andythenorth> you think it's not true, or so obvious not worth stating?
19:52:09 <frosch> i still do not know whether oversized and neo-bulk is the same :p
19:52:20 <planetmaker> it works for me well without including piece, bulk, liquid for some wagons
19:52:44 <planetmaker> frosch: I'm tempted to make it the same
19:52:56 <planetmaker> I'm not aware of any existing vehicle set which uses oversized
19:53:02 <planetmaker> nor industry set actually
19:53:19 <frosch> no idea whether george actually set them
19:53:25 <frosch> or whether mb just added them to wiki :p
19:53:57 <planetmaker> it was added in 2009
19:54:41 <frosch> we might just check the current machinery vector
19:54:49 <frosch> (i think that is the one with vehicles)
19:55:20 <frosch> anyway, what about turning the "clean" into a separate property, which is passed to the refitcost callback?
19:55:35 <frosch> or does anyone want to exclude clean cargos from his vehicles?
19:56:28 <peter1138> can you make the forum autoreplace mb's broken apostrophes? :p
19:56:47 <frosch> broken apo'strophe's ?
19:57:33 <peter1138> he uses U+00B4 Acute Accent
19:58:21 <peter1138> always has done o_O
19:59:09 <frosch> well, they were the same back in dos times
20:00:44 <peter1138> B4 would've been a line drawing symbol
20:01:13 <frosch> well, but the symbols which would appear when you press the keys
20:01:30 <frosch> ´ is easier to type than ' on german keyboard
20:01:34 <frosch> the latter requires shift
20:01:50 <planetmaker> ehm... frosch, not really. Yes, but...
20:01:56 <frosch> except if you use dead-keys, then you need to press ´ twice, resp use space
20:02:12 <frosch> but who uses dead keys? if you can have compose?
20:02:27 * planetmaker didn't consider shift more difficult ;-)
20:03:59 <andythenorth> frosch: wrt clean and oversized - I want to get a general approach written down as per 1-10 above (maybe excluding 7)...
20:04:13 <andythenorth> now that you've raised the problems with express, that kind of bothers me
20:04:19 <peter1138> it really messes up the text flow :(
20:04:39 <planetmaker> frosch: I'm still not convinced that clean would be a property...
20:05:14 <andythenorth> why is express a class *except* for historical reasons?
20:05:20 <andythenorth> express is a bit like clean
20:05:27 * andythenorth had some more thoughts around this
20:05:54 <andythenorth> classes are for *refitting*
20:06:14 <andythenorth> some of these classes might be more properly hints to *refit cost* and *cargo decay* callbacks
20:06:31 <andythenorth> 'express' should really mean 'enhanced decay'
20:06:39 <andythenorth> refrigerated similar
20:06:55 <andythenorth> clean absolutely
20:07:05 <andythenorth> is for refit cost only
20:08:35 <frosch> maybe "express" is the same as "clean" :p
20:09:17 <frosch> tourists are passenger+express btw
20:09:30 <andythenorth> tourists trouble me
20:09:34 <planetmaker> they're cleaner than passengers
20:09:46 <planetmaker> especially when the tourists are football fans
20:10:07 <peter1138> passenger or express?
20:10:15 <andythenorth> so tourists can travel by NARS 2 express box van?
20:10:17 <peter1138> so you can shovel them into express vans...?
20:10:29 <andythenorth> and planes without windows
20:10:36 <peter1138> so you need to exclude passengers... :p
20:10:57 <andythenorth> the madness recurs :P
20:11:16 <andythenorth> you exclude 'tourists' label
20:11:21 <andythenorth> because it's known
20:11:55 <peter1138> i think excluding the passenger class is more useful in that case
20:12:07 <andythenorth> or you exclude pax, because your vehicle definitely can't carry anything that might be PAX
20:12:23 <andythenorth> and then the unintended consequences that might occur are the fault of the cargo set
20:12:29 <andythenorth> for setting a silly class
20:12:40 <orudge> peter1138: heh, it does bug me, but the phpBB autoreplace isn't quite intelligent enough to be able to replace characters inside words
20:12:46 <orudge> at least, I don't think it is
20:12:50 <orudge> maybe you can do regexps these days
20:12:53 <peter1138> there's probably a module for it :)
20:12:59 <andythenorth> why set express? Any sane vehicle set author will just exclude tourists from express vehicles
20:13:04 <orudge> (hence people saying things like "shitty" and it appearing as "s***" with the word filter)
20:13:16 <orudge> (as oppossed to "s***ty")
20:16:59 <andythenorth> frosch: unless someone pops up and proves different, I would be happy for oversized/overweight to be merged with neo-bulk, although I don't like the 'oversized/overweight' name, it's inadequate
20:17:18 <peter1138> # funny how love is
20:17:32 <frosch> yes, i would rename it to "oversized/overweight/neo-bulk"
20:17:43 <frosch> anything transported in open wagons or flatbeds
20:17:49 <andythenorth> how about 'inconvenient' :P
20:18:05 <andythenorth> to go alongside express and clean as silly classes
20:18:34 <andythenorth> well I think we should merge in hazardous as well
20:18:43 <andythenorth> uranium is inconvenient :P
20:20:07 <andythenorth> hazardous is also a 'hint' property
20:20:34 <andythenorth> it could be used for excludes, but it's not that likely
20:21:10 <andythenorth> if you're doing a vehicle set at TTD scale, you shouldn't be adding 'tanker - non-hazardous', and 'tanker-hazardous'
20:21:14 <andythenorth> it's kind of pointless
20:21:35 <andythenorth> but you might use hazardous to lower the vehicle speed limit
20:21:52 <andythenorth> perhaps exclude from planes is valid for hazardous?
20:22:24 <andythenorth> you might use hazard to enforce using a caboose
20:24:54 <andythenorth> frosch: 'like bulk but doesn't pour'
20:25:41 <MNIM> in the future you could have town object to transporting hazardous through their town area?
20:25:47 *** mahmoud has joined #openttd
20:26:19 <MNIM> oh, and in case of a collision, bad things happen and surrounding tiles are rendered useless for some time?
20:29:36 <frosch> for some time? no. they get replace with MP_VOID for the rest of the game
20:31:43 <andythenorth> these are good ideas :P
20:32:07 <frosch> i guess you cannot even build tunnels to pass MP_VOID :p
20:32:10 <andythenorth> it would be fun - and also a huge griefing opportunity
20:34:13 <frosch> oh, but aircraft can pass it, so even if a town is enclosed by MP_VOID, you can still reach it
20:35:29 <andythenorth> frosch: where was your list of suggested changes to default cargos?
20:37:09 <frosch> i linked it in the thread
20:37:35 <andythenorth> I was looking in the thread but missed it :|
20:37:41 <frosch> but since noone commented on it, i assume noone took a look at it :p
20:38:07 <andythenorth> I just don't know what to do about it yet :)
20:38:18 <frosch> bubbles is the only cargo which is oversized but not neo-bulk
20:40:20 <andythenorth> I'm not convinced even that things like steel are either of those classes
20:40:32 <andythenorth> it could be, but is it helpful?
20:46:04 <planetmaker> why are bubbles not neo-bulk?
20:46:46 <planetmaker> we should declare steel as liquid and hazardous ;-)
20:47:49 <frosch> steel slabs look quite neo-bulk/overweight to me
20:48:40 <MNIM> it does when it belts :P
20:48:57 <MNIM> hah, that'd be a laugh, transporting molten metal
20:49:09 <Zuu> Possible in a mars scenario
20:49:38 <Zuu> Sounds like Afterlife :-)
20:49:47 <Zuu> If you ever played that game.
20:49:54 <frosch> MNIM: steel is transported via train in liquid state
20:50:04 <frosch> though only a few kilometers
20:50:45 <Zuu> MNIM: Its a sim/tycoon-ish game where you build the heven + hell.
20:53:44 <andythenorth> steel is definitely bad if it falls on you
20:53:52 <andythenorth> and I *am* planning to add torpedo cars to HEQS
20:54:08 <andythenorth> torpedo = molten metal, not weapon
20:55:55 <frosch> would make for an interesting income function
20:56:03 <frosch> if the trip take to long you have to buy a new wagon
20:57:18 <andythenorth> frosch set capacity to 0 if last trip took too long?
21:00:33 <andythenorth> frosch: wrt oversized / neo-bulk, let's plan to merge them in 7 days(?) if nobody objects?
21:00:58 <andythenorth> and lets argue away the 'oversized / overweight' aspect as not-very-important
21:01:17 <Zuu> Recently, I was thinking about a NewGRF which lowers the train speeds and increase loading time in the winter. :-)
21:01:51 <andythenorth> Zuu: you'd need hemisphere parameter
21:03:36 <Zuu> Or put it in a national train set? Though it would be interesting if you could run an australian trainset along with eg. UKRS both with this "feature" :-)
21:04:02 <Zuu> Those 500 ton wagons are nice :-)
21:05:22 <Zuu> One had a derail here in Sweden which I took quite some time to move away. :-)
21:10:53 *** TWerkhoven[l] has joined #openttd
21:11:26 <Rubidium> just use the z-level and snow line?
21:11:42 <Rubidium> or just the snow line in general
21:17:11 <andythenorth> can a train check snow line?
21:17:48 <frosch> not officially with more heightlevels
21:18:55 <frosch> but you can compare variable 9E with var 20 to draw snow on your ship :p
21:19:03 <andythenorth> ice on ships is bad
21:19:38 <andythenorth> I can do a sinking ship without NewDisasters
21:21:04 *** Hyronymus has joined #openttd
21:35:52 <Eddi|zuHause> <andythenorth> or "special open high-sided wagons" (class F) <-- mining trucks are most likely that. F are self-discharging hopper wagons.
21:36:13 <andythenorth> mining trucks definitely self-discharge
21:41:55 <Rubidium> so do box cars with gasses ;)
21:42:05 <andythenorth> babies self-discharge
21:42:19 <planetmaker> they're neo-bulk, though
21:43:19 * andythenorth ponders writing actual code to test if 'clean' is a stupid idea
21:45:54 <Rubidium> shouldn't liquid containers always be cleaned?
21:48:56 <planetmaker> there's a different between clean and clean when you refit between oil and diesel and milk
21:49:37 <planetmaker> same as there's a difference between clean hands before eating and before doing a liver transplant
21:50:06 <andythenorth> planetmaker: are you for or against? :)
21:50:41 <Rubidium> planetmaker: imo diesel and oil are the same in terms of cargo in the TTD realm
21:50:58 <planetmaker> I'm unsure. I like the idea, but of course it doesn't describe in any way a provision necessary
21:51:04 <planetmaker> as such it's a bit odd class
21:51:12 <andythenorth> I think it's odd
21:51:16 <planetmaker> Rubidium: oil and fuel oil then ;-)
21:51:20 <andythenorth> so is the livery refit class :P
21:52:25 <andythenorth> and 'regearing' is not pikka's finest moment
21:52:31 <Rubidium> planetmaker: that, maybe, goes one way (fuel oil -> oil). The other way around needs cleaning due to the crap in oil
21:53:47 <planetmaker> Rubidium: yes. But the amount of cleaning is WAY different to when you would want to put in milk after a delivery of crude oil
21:53:55 <planetmaker> (i.e. in reallife it wouldn't be allowed)
21:54:14 <andythenorth> all liquids would demand cleaning
21:54:26 <andythenorth> maybe there's a way to solve it by convention, not class
21:54:27 *** valhallasw has joined #openttd
21:54:38 <andythenorth> again, it might have to be left to the vehicle set author
21:54:41 <Eddi|zuHause> <andythenorth> hazardous is also a 'hint' property <-- "hazardous" should be the opposite of "clean", i.e. add it to "oil", "fuel" and "chemicals"
21:55:09 <planetmaker> Eddi|zuHause: andythenorth, but then it should go into the same property as clean then
21:55:15 <planetmaker> or as express possibly
21:55:17 <Rubidium> Eddi|zuHause: fluid oxygen is hazardous as well, but might not need as much cleaning
21:55:33 <andythenorth> isn't it boring when a vehicle set goes something like A, B (bigger, faster than A), C (bigger, faster than B), D (bigger, faster than C)
21:55:41 <planetmaker> fluid oxygen... much "nicer" than oil ;-)
21:56:19 <andythenorth> is it easier just to do the inverse of everything oberhumer suggests?
21:56:23 <andythenorth> easier / better /s
21:57:43 <andythenorth> I wondered if it was a general rule in life: !(oberhumer suggestions)
21:57:53 <andythenorth> in this case, it's about trams
21:58:29 <Eddi|zuHause> <andythenorth> and I *am* planning to add torpedo cars to HEQS <-- what UIC class are those? U? Z?
22:00:02 <andythenorth> but they will refit to anything 'clean'
22:00:19 <andythenorth> cargo might get vapourised, but won't get dirty
22:02:42 * andythenorth ponders some actual FIRS cargo classes
22:03:20 <andythenorth> planetmaker: no rush, but do you want to see if you can see any FIRS classes that look wrong - in light of the AND/OR thinking etc.
22:03:40 <andythenorth> and also the 'liquids can go in containers but that doesn't make them piece goods' type thinking
22:04:15 <planetmaker> I shall have a look
22:07:47 <peter1138> anyone heard of maverick sabre?
22:09:07 <andythenorth> planetmaker: BEER is probably wrong
22:09:11 <andythenorth> sets piece goods
22:09:46 <planetmaker> well. maybe. Though I've never seen a beer tanker ;-)
22:09:52 <planetmaker> would be good for 1st May though
22:10:01 <Eddi|zuHause> BEER should very well be piece goods
22:10:30 * andythenorth thinks this would be a good test
22:10:42 <andythenorth> what is specifically piece goods about beer?
22:10:50 <andythenorth> when I drink it, it's liquid...
22:11:08 <Eddi|zuHause> beer is delivered in crates
22:11:22 <Eddi|zuHause> or occasionally barrels
22:11:26 <andythenorth> that conflates cargo with container
22:11:31 <Eddi|zuHause> andythenorth: this is ONLY about the delivery
22:11:50 <andythenorth> this is only about the properties of the cargo
22:11:58 <Eddi|zuHause> andythenorth: everything other the (common) delivery method is COMPLETELY IRRELEVANT
22:11:59 <andythenorth> cargo set author doesn't control vehicles
22:12:39 <Eddi|zuHause> yes, the property of the cargo is that it is commonly packaged in bottles/barrels
22:13:03 <andythenorth> I don't mean you're wrong
22:13:08 <andythenorth> I just mean it's not helpful
22:13:27 <andythenorth> basing cargo classes on examples of packaging or real world vehicles isn't helpful
22:13:48 <andythenorth> every single cargo is piece goods
22:14:03 <andythenorth> I can create a case for any cargo you name that it is piece goods
22:14:58 <Eddi|zuHause> not "a case", but "a common case"
22:15:17 <Eddi|zuHause> (other than "fits into an ISO container")
22:15:24 *** supermop_ has joined #openttd
22:15:28 <andythenorth> google image search?
22:16:31 <andythenorth> Eddi|zuHause: do a nitpick reply in the thread if you like
22:16:39 <andythenorth> I'll find the link
22:17:16 <andythenorth> having a "yes!" "no!" argument here is not a good use of your time or mine
22:17:20 <Eddi|zuHause> i have that one open already :)
22:17:48 <andythenorth> also we have a new contribution
22:17:57 <andythenorth> and proof of the AND problem
22:18:13 <Eddi|zuHause> not understanding that fully, but it roughly seems to coincide with my proposal
22:18:51 <andythenorth> the upside of my solution is that it keeps the existing class structure entirely
22:19:02 <andythenorth> the downside is that nobody seems to agree wtf that is
22:19:17 <andythenorth> the upside of your solution is that it makes sense
22:19:27 <andythenorth> the downside is it's entirely new
22:19:56 <Eddi|zuHause> deprecate the existing system with GRFv8?
22:20:04 <Eddi|zuHause> replace it with a new one?
22:20:44 <andythenorth> I'm way out of my depth about how to do that
22:22:34 <andythenorth> I'm trying to figure out the newest suggestion
22:22:52 <andythenorth> doesn't cement travel in his grain car?
22:23:12 <andythenorth> and coal travels in his fish wagon
22:23:40 <peter1138> # somehow i have to make this final breakthrough... now
22:24:38 <andythenorth> or he has to have a wagon for 'covered AND bulk AND refrigerated'
22:24:56 <andythenorth> and another wagon for 'bulk AND food AND covered'
22:25:05 <andythenorth> which is going to make ships interesting at minimum :P
22:25:35 <andythenorth> if it has to be exact-match AND, then I'm going to need to invent a lot stupid ship names for FISH
22:26:03 <supermop_> thought the idea behind fish was all ships able to carry any cargo
22:26:11 <andythenorth> and the buy menu length will be (number ship models) * (number of class combinations)
22:26:30 <andythenorth> supermop_: that wouldn't be possible with exact-match AND on cargo classes
22:26:42 <andythenorth> you could only refit if all the classes matched perfectly
22:26:53 <andythenorth> so you'd need a vehicle type for every combination of cargo classes in the game
22:27:11 <DorpsGek> andythenorth: Error: invalid syntax (<string>, line 1)
22:27:18 <DorpsGek> andythenorth: Error: unexpected EOF while parsing (<string>, line 1)
22:27:39 <planetmaker> @calc factorial(11)
22:27:39 <DorpsGek> planetmaker: Error: 'factorial' is not a defined function.
22:27:48 <DorpsGek> planetmaker: Error: 'factor' is not a defined function.
22:27:50 <andythenorth> is it factorial, or nCr ?
22:28:24 <supermop_> none of you own a calculator anymore?
22:28:26 <andythenorth> if factorial, 39916800 vehicles are needed apparently
22:28:33 <Eddi|zuHause> @calc 1/sqrt(5)*(((1+sqrt(5))/2)**11-((1-sqrt(5))/2)**11)
22:28:52 <Eddi|zuHause> that can't be right
22:28:53 <planetmaker> supermop_: but in the other room ;-)
22:29:07 * andythenorth has an Apple calculator :P
22:29:28 <andythenorth> stupid thing also has the internet and a hard drive
22:29:35 <andythenorth> what's the point in those?
22:29:54 <supermop_> i want one of the rereleased hp 15Cs
22:30:05 <andythenorth> anyway, 39916800 seems to be a lot of vehicles
22:30:46 * valhallasw still works on an HP 41CV
22:31:57 <andythenorth> maybe with the callback it would be ok actually
22:32:06 <andythenorth> so if it must be AND strict-matched
22:32:14 <andythenorth> I need how many varact checks?
22:33:16 <andythenorth> but I'm writing the cb, so I can ignore all the AND stuff anyway
22:33:19 <andythenorth> and just do what I like
22:33:44 <andythenorth> I can return 'allow' if it's 1910 and disallow if the snowline is > 7
22:34:34 <andythenorth> Eddi|zuHause: do you propose moving the 'extra' classes to a new prop?
22:34:59 <Eddi|zuHause> andythenorth: implementation detail...
22:35:53 <andythenorth> but we could achieve your proposal by convention (keep existing system, argue with people who do it wrong), or by new code
22:36:15 <andythenorth> milk is clearly piece goods if I accept your beer argument
22:37:59 <Eddi|zuHause> andythenorth: the difference, like frosch123 pointed out, is that you transport milk from the farm to the dairy, while you transport beer from the brewery to the customers
22:38:17 <Eddi|zuHause> andythenorth: if you transported milk from the dairy to the customers, then the same as beer would apply
22:38:37 <Eddi|zuHause> andythenorth: but you transport food from the dairy to the customer, which already is piece goods
22:38:46 <andythenorth> you can file it under edge case, but I think you make more problems with the beer argument than you solve :P
22:38:59 <andythenorth> change it to *beer bottles* and you win
22:39:32 <andythenorth> how is beer dimensioned in FIRS?
22:39:48 <supermop_> bottles and kegs could arguable be different cargoes anyway
22:39:52 *** Neosublimation has quit IRC
22:40:08 <andythenorth> litres of alcohol
22:40:19 <andythenorth> dimension it in crates and I'd accept your case
22:40:39 <Eddi|zuHause> andythenorth: 1 crate == 12 litres :)
22:41:01 <supermop_> a hand truck would hold one keg or 6 cases
22:41:20 <supermop_> soo that doesnt really help
22:42:00 <Eddi|zuHause> andythenorth: i think you make this more of a problem than it is
22:42:45 <Eddi|zuHause> andythenorth: there really is no problem wrt newgrf development if a cargo is both "piece goods" and "liquid", as long as people stick to the "no exclude" rule
22:43:16 <supermop_> andy you've inspired me to grab a brooklyn winter ale from the fridge
22:43:21 <andythenorth> Eddi|zuHause: maybe
22:43:36 <andythenorth> it's unclear though
22:44:02 <andythenorth> now you (as cargo set author) want to control how cargo is transported in detail?
22:44:02 *** HerzogDeXtEr1 has joined #openttd
22:44:57 <andythenorth> I accept it does no harm
22:46:22 <andythenorth> cargo waiting: 99 bottles of beer
22:49:31 *** valhalla1w has joined #openttd
22:54:31 *** valhalla2w has joined #openttd
23:07:31 <michi_cc> Eddi|zuHause: Do you think it would make sense to change YAIM to show "123 €/month (456)"?
23:07:57 <Eddi|zuHause> imho just leave out the 456 completely
23:08:43 <michi_cc> The window is showing the count though if you disable maintenance cost, something some people might find useful.
23:09:00 <Eddi|zuHause> or add a details button listing "tracks" "crossings" "signals" "stations" "tunnels" etc. for each railtype
23:10:15 <Eddi|zuHause> so you can have: "tracks: 123", "crossings: 35 (x2)", "tunnels: 12 (x4)", ... "sum: 456"
23:10:52 <michi_cc> Oh, crossings are much more evil :p For overlapping track bits, its count²
23:12:44 <michi_cc> So a detail window in a detail window? I refuse to write an expandable tree or something like that.
23:14:34 <Eddi|zuHause> so it's "switches: 34 (x4, x9)"?
23:16:26 <Eddi|zuHause> maybe "switches: 15 (x4)", "crossings: 25 (x9..x36)"?
23:17:50 <Eddi|zuHause> and [?]-click on a tile could show that tile's maintenance cost
23:18:17 <Eddi|zuHause> (maybe i should actually test the patch...)
23:20:54 <peter1138> why does it need to be so detailed?
23:21:59 <Eddi|zuHause> peter1138: so people notice which sections are really evil maintenance-wise
23:22:36 <peter1138> i'm still trying to add a scrollbar to the finances window
23:22:49 <peter1138> so that no extra window is needed for the breakdown of costs, maybe ;)
23:23:09 <Eddi|zuHause> peter1138: what's the problem with that?
23:23:31 <Eddi|zuHause> with the new widget trees, that should be a really simple task
23:31:31 <Zuu> I have always missed a window with a sumary over how much infrastructure each company have. The company window have vehicle counts but if you want eg. station counts you need to open also the station list of all companies.
23:58:50 <michi_cc> Eddi|zuHause: The finance window is not much more than a single widget in which everything is hand-drawn.
23:59:29 <Eddi|zuHause> michi_cc: so just offset and crop the drawing depending on scrollbar position
continue to next day ⏵
