New Railtype Label Scheme

Discussions about the technical aspects of graphics development, including NewGRF tools and utilities.

Moderator: Graphics Moderators

User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

New Railtype Label Scheme

Post by FooBar »

This topic discusses the Standardized Railtype Scheme.

==original post contents==

As you may know, a new generic railtype label scheme has been discussed and worked out in several topics and the devzone. For the Dutch sets we want to use this scheme, but there was no single document that explained how it works exactly. So I decided to create one.

For now it's on my user space here: http://newgrf-specs.tt-wiki.net/wiki/Us ... typelabels

Please note that I didn't invent this scheme. I just gathered the details from all different places and tried to explain how it works. Therefore this topic is not about discussing how the scheme should work. I think that has been discussed enough previously, to an extent that those who concern it agree on the details.

What this topic is about is me asking those who stood by the cradle of the scheme to check if what I've written is indeed correct and/or if there are things missing from it. If so it can be amended, with the ultimate goal of publishing this as part of the specs for everyone to use. So please check out that page and comment on it!


It must be noted that I did add monorail and maglev track classes. I didn't find anything about those, so I picked letters M and L for it. If there are previous definitions that I don't know about, please let me know. I do realize that in practice these will probably never be used, due to the defaults MONO and MGLV probably being enough. But it's good to have them predefined just in case someone wants to go berserk on a monolev set.
Also I wasn't sure about the amount of axle load classes, so I used the complete EN 15528 from A to G.
Lastly I added two-phase catenary electrification, something we'll be needing for the Dutch sets.
Last edited by FooBar on 21 Apr 2012 15:35, edited 1 time in total.
Eddi
Tycoon
Tycoon
Posts: 8254
Joined: 17 Jan 2007 00:14

Re: New Railtype Label Scheme

Post by Eddi »

i have a vague memory of "M" already being reserved for "metric gauge" (as opposed to the unspecific "narrow gauge", to differentiate some narrow gauge types). Maybe Snail knows more about that.
Michi_cc
OpenTTD Developer
OpenTTD Developer
Posts: 619
Joined: 14 Jun 2004 23:27
Location: Berlin, Germany
Contact:

Re: New Railtype Label Scheme

Post by Michi_cc »

FooBar wrote:For the Dutch sets we want to use this scheme, but there was no single document that explained how it works exactly. So I decided to create one.
Thanks, I'll have a look at it.
Eddi wrote:i have a vague memory of "M" already being reserved for "metric gauge" (as opposed to the unspecific "narrow gauge", to differentiate some narrow gauge types). Maybe Snail knows more about that.
Not sure what exactly he planned, but on the Japanese Track Set devzone ticket, he had "N" as both generic narrow gauge and wider narrow gauge and "n" for narrower narrow gauge.

-- Michael Lutz
Eddi
Tycoon
Tycoon
Posts: 8254
Joined: 17 Jan 2007 00:14

Re: New Railtype Label Scheme

Post by Eddi »

So in practice this means that you should RAIL instead of SAAN and ELRL instead of SAAE.
you are a word here.

when making the scheme this detailed, one might want to add a paragraph about "universal" engines to the power supply section, although that might get complicated when you want to provide different combinations: dual-system, quad-system, etc. however that might be a discussion for another topic.
User avatar
PikkaBird
Graphics Moderator
Graphics Moderator
Posts: 5601
Joined: 13 Sep 2004 13:21
Location: The Moon

Re: New Railtype Label Scheme

Post by PikkaBird »

So are "SAAN" "SABN" "SACN" "SADN" "SAEN" "SAFN" "SAGN" sufficient alt labels for "RAIL"?
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

If you only have one type of standard gauge unelectrified rail with no speed limits and no axle load limits: yes.

But if you want absolutely nothing fancy in your track set, you can get away with just defining RAIL. Because vehicle sets are advised to define that as fallback type anyways. But that's up to you. I'd add these labels as alternatives; it's not particularly complicated to do.
User avatar
Snail
Tycoon
Tycoon
Posts: 1283
Joined: 28 Apr 2003 18:52
Contact:

Re: New Railtype Label Scheme

Post by Snail »

Wow! Great job!
FooBar wrote:What this topic is about is me asking those who stood by the cradle of the scheme to check if what I've written is indeed correct and/or if there are things missing from it.
Yep, I'm one of them ;) The inputs I can provide are the following:

* Gauge: my original idea was actually to use "N" as narrow gauge, so "M" was indeed a free letter, and may be used for something else if needed. I don't think we need a specific letter for "metric" gauge: just two generic letters for different narrow gauges ("N" and "n"; narrow and narrower) could do the trick. I don't think any sets would be so detailed as to include, say, 1,066m, 1m, 750mm, and 600mm tracks at the same time, each as a different type... especially considering we've "merely" got 16 types to play with!

* Speed: I wouldn't define class "R" (rackrail) as just eyecandy. It's actually a tracktype with a max speed of 20 km/h for rackrail engines, and 10 km/h for non-rackrail engines (they need to slow down because of the rack's presence). Moreover, it improves the rackrail engines' adhesion coefficient, therefore boosting their TE. This is how I coded it in the French set. Therefore, I'd describe it as a special speed limit class.
(I'm not sure about "U" though, as that idea was not mine).

* Axle weight: my thoughts were to also have the "a" type in the axle weight scheme, as a "very early, very light, very cheap" track class (may be useful for sets starting around 1840).

As for the other things, the document is well written. Hope this concept can help other sets' developers out! :)
The French Narrow Gauge Train Set is now released! Get it here
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Thanks :)

Ok, then M will be fine for monorail, as nobody used it for metre gauge before. If someone really needs three types of narrow gauge, lowercase m is still available for metre gauge.

As for rack rail I will add a third group of speed classes. I wasn't really sure how you implemented it, but now that I know it is indeed much more than just eyecandy.
Eddi wrote:when making the scheme this detailed, one might want to add a paragraph about "universal" engines to the power supply section, although that might get complicated when you want to provide different combinations: dual-system, quad-system, etc.
Good idea. It will indeed get complicated, but I think it's good to make that clear to authors who are considering something like this and what the consequenses are. Like the requirement of a dedicated track type for each different type of engine (what we've discussed earlier for the Dutch trackset basically).
Snail wrote:Axle weight: my thoughts were to also have the "a" type in the axle weight scheme, as a "very early, very light, very cheap" track class (may be useful for sets starting around 1840).
So if I understand correctly, you're suggesting to add "a" as a standard axle class to the list of A through G? I'm not against that, but it's not just my call to add it. So let's see what others think of that (everyone: this is an open invitation to say yay or nay).
michael blunck
Tycoon
Tycoon
Posts: 5948
Joined: 27 Apr 2005 07:09
Contact:

Re: New Railtype Label Scheme

Post by michael blunck »

@FooBar

OK, then I´ll drop "M" for "metric". BTW, some people were asking to reserve "M" for "metro", but I don´t see much use in it.

W/r to "a", yes, this one should be sufficient to define a "very light axle weight", usually in connection with narrow-gauge, see here.

BTW, UIC "F" and "G" are not in use, but only facultative. I don´t know if such high axle loads are really being used in the US or in Russia, and thus had tio be included?

In general, the concept should in no way getting too complicated.

regards
Michael
Image
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Yes, it's best to use N for any type of narrow gauge, for compatibility reasons with other track or train sets. M for metro is conceptually wrong, as it is standard gauge rail with a different type of electrification, generally N**3. If one doesn't want regular trains on metro tracks, then don't set those as compatible with it.

About axle classes "F" and "G" I wasn't sure if we need those. However, if we don't define them now and at some point someone needs them, all previously released track sets need to be updated (or they're not fully compatible with this new train set). So in that light I'm rather safe than sorry. And in that same light it's best to add "a" to the defaults as well. The heaviest trains in the world have an axle load of over 40 t, so I guess some train sets may need these last labels (possibly deviating from the EN and using exponentially increasing axle loads for the higher classes).
michael blunck wrote:In general, the concept should in no way getting too complicated.
Agreed. But I don't think it's currently too complicated. Of course it's more complicated than just RAIL and ELRL, but that's inevitable. :)
Eddi
Tycoon
Tycoon
Posts: 8254
Joined: 17 Jan 2007 00:14

Re: New Railtype Label Scheme

Post by Eddi »

FooBar wrote:About axle classes "F" and "G" I wasn't sure if we need those. However, if we don't define them now and at some point someone needs them, all previously released track sets need to be updated (or they're not fully compatible with this new train set). So in that light I'm rather safe than sorry. And in that same light it's best to add "a" to the defaults as well. The heaviest trains in the world have an axle load of over 40 t, so I guess some train sets may need these last labels (possibly deviating from the EN and using exponentially increasing axle loads for the higher classes).
we could also make it so only weight types A through D (or E) are mandatory, and a, b, ... as well as F, G, ... optional. then it's the responsibility of the train set to ensure that vehicles have a proper fallback. this would prevent a "alternate label" list explosion, still keeping it flexible and generic enough (imho).

on the matter of "universal" engines: a "reasonable" approach could be:
  • E: generic catenary-powered electric engines. "universal" [4-system] if any of D/d/A/a are defined.
  • A: generic AC catenary electric engines. 25kV only if a also defined. vehicle sets should use E as fallback, if defining an engine for A
  • a: 15kV AC catenary electric engines. only defined if A also defined. vehicle sets should use A and E as fallback, if defining an engine for a.
  • D: generic DC catenary electric engines. 3kV if d also defined. vehicle sets should use E as fallback, if defining an engine for D
  • d: 1.5kV DC catenary electric engines. only defined if D also defined. vehicle sets should use D and E as fallback, if defining an engine for d.
dual-system engines cannot be modeled with this scheme. (trying to do that would lead to an explosion of combinations)
michael blunck wrote:In general, the concept should in no way getting too complicated.
Agreed. But I don't think it's currently too complicated. Of course it's more complicated than just RAIL and ELRL, but that's inevitable. :)
in general, there's always a treadoff between complexity and functionality. one should take care that you don't end up with a "Eierlegendewollmilchsau" though...
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Eddi wrote:dual-system engines cannot be modeled with this scheme.
Well, you can, but only when not using "a" and "d" for that trainset. Then your dual-system engine will be on "E", instead of the quad-system. Same applies for triple-system. How I see it you can have either dual, triple or quad system trains, but not a combination of those.

The rest of your explanation is something I can copy straight to the wiki page :)
Eddi wrote:we could also make it so only weight types A through D (or E) are mandatory, and a, b, ... as well as F, G, ... optional.
I'm just starting to think that the actual weight definitions for each class should depend on the track gauge. For narrow gauge, the A limit can be much lower than for standard gauge, because narrow gauge systems are usually lightweight when compared to standard gauge. Broad gauge is probably not much different than standard gauge.
This way, class A for standard gauge may be equal to class C for narrow gauge. And then we don't need this many classes, as for any track type 5 classes should be plenty, considering the 16 railtype limit.
User avatar
Snail
Tycoon
Tycoon
Posts: 1283
Joined: 28 Apr 2003 18:52
Contact:

Re: New Railtype Label Scheme

Post by Snail »

FooBar wrote:I'm just starting to think that the actual weight definitions for each class should depend on the track gauge. For narrow gauge, the A limit can be much lower than for standard gauge, because narrow gauge systems are usually lightweight when compared to standard gauge. Broad gauge is probably not much different than standard gauge.
This way, class A for standard gauge may be equal to class C for narrow gauge. And then we don't need this many classes, as for any track type 5 classes should be plenty, considering the 16 railtype limit.
Well, as it's well written on the Wiki page, the actual axle weights corresponding to each class should be up to the trainsets. The "A", "B" etc. letters should only be an "ordinal" number. This can also be true across the different gauges: so, S*A* could mean, say, max 16 tons, and N*A* could mean max 15 tons.
In any case, narrow gauge should always include 1, max 2, different axle-weight tracks. Usually those tracks are built to be cheap to begin with; for heavier trains, standard gauge is almost always used. So, even thinking about the 16-railtype limit, N*A* and possibly N*B* should be enough for most sets.
The French Narrow Gauge Train Set is now released! Get it here
Michi_cc
OpenTTD Developer
OpenTTD Developer
Posts: 619
Joined: 14 Jun 2004 23:27
Location: Berlin, Germany
Contact:

Re: New Railtype Label Scheme

Post by Michi_cc »

michael blunck wrote:BTW, UIC "F" and "G" are not in use, but only facultative. I don´t know if such high axle loads are really being used in the US or in Russia, and thus had tio be included?
I could find references that up to 32t are used in the US, but I have no idea how common that is.

But combined with:
FooBar wrote:I'm just starting to think that the actual weight definitions for each class should depend on the track gauge.
Treat the axle weights purely as a guideline to help with decision making, but do not in any way make them mandatory numbers. This way there's no need for any "F" or "G", because a US set could simply "shift up" all axle weights to stay in the same letter range.

-- Michael Lutz
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Yes, it may be best to let these fixed axle weights completely slide. This however does mean that track and train sets can never use any axle weight indication in tonnes, but always must the letters as reference. Otherwise you may run into the situation where a train says "I'm a label C train with axle weight 20 t", while the track set indicates for label C a maximum of 18 t. Then you get the situation that the user is lead to believe that he must build track type D, while in fact the information doesn't match and track type C will work just as well.

If that is something we can work with, then we indeed should drop all but axle weight labels A-E. Train sets that really need more than 5 different axle load classes can then define additional ones with fallback.

For the best user experience, a track set should then put the axle load class letter in the name of the railtype. A trainset should put the minimal axle load class in the additional purchase information. This is of course not mandatory, but it makes things clear for the users.

So, once more, those in favour say aye :)
User avatar
Toni Babelony
Tycoon
Tycoon
Posts: 1389
Joined: 07 Jul 2006 09:34
Skype: toni_babelony
Location: Sagamihara-shi, Japan
Contact:

Re: New Railtype Label Scheme

Post by Toni Babelony »

Sweet. This scheme fits perfectly in the Japan Set's plan! Thanks for the effort. :)
Retired JapanSet developer and creator of TIAS.
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Given that:
- nobody complained about the reduction of axle load classes;
- Snail suggested to use A and B and not a for narrow gauge (IIRC the only one initially with a use for a);
- most if not all who replied stressed that the axle load classes are relative.
I decided to reduce the amount of classes on the overview in my wiki user space to 5. Furthermore I dropped the reference to the numeric axle weights and explained how train and track sets should deal with the relativeness of the classes.

Please comment if you have any.

I also made clear that rack rail is not just eyecandy and added a note and Eddi's example on multi-current/system electrification.


To me, the scheme is now complete and everything that needs explaining about it is explained. If there are no objections in the next two weeks, I'll move it from my user space to the newgrf specs and make it an official recommendation for railtype label usage. If there are objections, then please speak now :D
User avatar
Snail
Tycoon
Tycoon
Posts: 1283
Joined: 28 Apr 2003 18:52
Contact:

Re: New Railtype Label Scheme

Post by Snail »

Nice progress!

I like the fact that you kept A-E as "primary" axle weight classes most sets should define, but you left it open for sets to define axle weights labels other than those. In fact, my plans for the French Tracks set were to start in 1840 with two types of tracks, respectively with axle weight classes "a" and "A". So, if other trainsets start defining their trains from letter "A" and not "a", they will still be able to use one track type (the most expensive one) from my trackset.

The only remark I should make is that the T electrification class is not meant to design a triple-voltage system, but the old electrification system used in Italy and Switzerland in the beginning of the XXth century (it wasn't fully abandoned in Italy until 1976). Perhaps it's better to call it "Threephase AC".
Speaking of which, there were very few trains that could powered by both "threephase AC" and normal DC catenary. Perhaps we could reserve the letter "t" for those.

So, if we need a triple-voltage system, we should find a different letter. Since "C" has been suggested for dual-voltage catenary, why not using "c" for dual-voltage and "C" for triple-voltage? And "T" for the old "threephase AC" system?
The French Narrow Gauge Train Set is now released! Get it here
User avatar
FooBar
Tycoon
Tycoon
Posts: 6553
Joined: 21 May 2007 11:47
Location: The Netherlands
Contact:

Re: New Railtype Label Scheme

Post by FooBar »

Are you going all the way to E with you axle load classes in the French set? If not, then it may be better to use "A" and "B" instead of "a" and "A". That saves you the fallback hassle. Also axle load class "A" can very well be lower for narrow gauge than class "A" for standard gauge. But that's ultimately up to you :)

Class "T" I have misunderstood, expecially when adding "C". I've changed the description of "T" and added a link to the wikipedia article.

I decided to remove "C" altogether, as most sets won't need it. You're only going to need it for a set that has both dual-voltage and triple-voltage. If a set only has one, using "E" for that will be sufficient. If we want to cater for all possible combinations, we need 8 additional labels: A+D, A+d, a+D, a+d, A+D+d, a+D+d, A+a+D, A+a+d. Let's leave defining those up to someone that is crazy enough to make a set this complicated.
michael blunck
Tycoon
Tycoon
Posts: 5948
Joined: 27 Apr 2005 07:09
Contact:

Re: New Railtype Label Scheme

Post by michael blunck »

FooBar wrote: [...] If there are no objections in the next two weeks, I'll move it from my user space to the newgrf specs and make it an official recommendation for railtype label usage.
An arrangement like this has nothing to do with the newgrf specs, especially not because those specs only require a unique label, but don´t understand concepts like "axle weight" or "electrification type".

So this would be purely a private agreement between some vehicle and track set authors, and in no way obligatory or even "recommended".


Granted, it makes more sense like that silly recommendation for object labels, but nevertheless.

regards
Michael
Image
Post Reply

Return to “NewGRF Technical Discussions”

Who is online

Users browsing this forum: No registered users and 3 guests