Do we need a new game engine?

Again as the outsider what is a game engine? I know what it is as regards Blender if you refer to my Blender serious discussion thread.
Again even then is it the actual coding and not just the functions. With Blender Game Engine. You make everthing, sky dome gets dragged along, roaming camera is made, lamps get set up and etc but again they are still functions.

So to talk about a game engine wouldn't you need to understand how the programming is set up and the way it reads the functions and understand coding. Can't this programming be constanly amended and updated.
Or is the way the expression used in Trainz just refers to the actual functions available at any one time which can be constantly updated and amended and change with each version.

Of course as always nowadays just a question from an outsider:) B
 
Last edited:
Great tip! I still think it should be mandatory as part of the upload process for poly and texture overheads to be explicitly stated - as a tag in the config? - so that route builders can have more accessible information about assets. Could there be a way to access an asset's 'vital statistics' in Surveyor?

Paul

Well, you should be able to see if it uses LOD by looking at the config.txt. A model using LOD could be expected to perform reasonably well. PEV's tools can provide information about poly counts. I have no idea how the author could state texture overheads.

But, I have to ask the question: even if the author provided that information, would the average user read or understand them?
 
Well, you should be able to see if it uses LOD by looking at the config.txt. A model using LOD could be expected to perform reasonably well. PEV's tools can provide information about poly counts. I have no idea how the author could state texture overheads.

Content Manager can also provide some information about poly counts and the number of textures used: right click on an asset and select "Show mesh technical information".
This tool is not perfect, as it only sees the mesh with the higher poly count and does not see meshes contained in sub-folders. Most "authors" of high-poly assets, however, don't even bother with LOD and sub folders are mostly used in locomotives or rolling stock, so this tool is somewhat useful.

The last version of Ruby TMIX automatically calculates the texture overhead and automatically adds a line in the "description" tag stating (more or less) "This model has an effective system load equivalent of XXXX polygons. YYYY polygons for the 3D model and ZZZZ polygons due to textures".

I think that a similar tool should be added to CM to allow evaluating content and tracking inefficient content. I wish I could look among the dependencies of a route and have at least a rough indication of the system load equivalent, so as to eradicate content with thousands of polys...
Unfortunately, Content Manager filters do not even allow searching the text of the description tag :(.

But, I have to ask the question: even if the author provided that information, would the average user read or understand them?

Absolutely not: it is far more easier to whine about the game engine :D
 
So to talk about a game engine wouldn't you need to understand how the programming is set up and the way it reads the functions and understand coding. Can't this programming be constantly amended and updated.

Yes, which is what happens in the real world, not that of armchair experts. Game developers or for that matter those who develop applications don't throw away what works, they keep the good bits loose the outdated bits, add the modifications for newer features not available originally as the hardware to do it didn't exist. They may well say it's a "new" engine but at basic code level it's usually the same beast.
 
In keeping this thread positive I think we should all acknowledge that Auran and subsequently N3V have done great things with a game design that was originally tasked to run a desktop model railway environment. That they have brought it to the level of TS12 is an achievement they should be proud of. However that still remains its Achilles Heel and I'm not just referring to the graphics but to the overall train driving experience. When Trainz was new we kind of forgave the fact you had to set the crudely drawn points as you drove the train, that there was no AI signaller for the player train, that it was virtually impossible to set up and have the game monitor progress against a timetable and that the physics - even in cab mode - could best be described as simplistic. However as the years moved on and other sims offered operational complexity and a better virtual environment, Trainz has singularly failed to do so. At heart, it's still the big model railway and that to me is the aspect indeed whole culture which needs to change most of all.


I see you have many versions of Trainz, but have you ever actually used them? The above seems to indicate you have not...
 
Content Manager can also provide some information about poly counts and the number of textures used: right click on an asset and select "Show mesh technical information".
This tool is not perfect, as it only sees the mesh with the higher poly count and does not see meshes contained in sub-folders. Most "authors" of high-poly assets, however, don't even bother with LOD and sub folders are mostly used in locomotives or rolling stock, so this tool is somewhat useful.

The last version of Ruby TMIX automatically calculates the texture overhead and automatically adds a line in the "description" tag stating (more or less) "This model has an effective system load equivalent of XXXX polygons. YYYY polygons for the 3D model and ZZZZ polygons due to textures".

I think that a similar tool should be added to CM to allow evaluating content and tracking inefficient content. I wish I could look among the dependencies of a route and have at least a rough indication of the system load equivalent, so as to eradicate content with thousands of polys...
Unfortunately, Content Manager filters do not even allow searching the text of the description tag :(.



Absolutely not: it is far more easier to whine about the game engine :D
If you haven't, i posted Content that i converted from Google sketchup to the DLS.
 
The last version of Ruby TMIX automatically calculates the texture overhead and automatically adds a line in the "description" tag stating (more or less) "This model has an effective system load equivalent of XXXX polygons. YYYY polygons for the 3D model and ZZZZ polygons due to textures".

I think that a similar tool should be added to CM to allow evaluating content and tracking inefficient content. I wish I could look among the dependencies of a route and have at least a rough indication of the system load equivalent, so as to eradicate content with thousands of polys...

i dont want to see anything of the sort. i still think trying to compare everything to a 'polygon' is ridiculous. a polygon is not a means to measure system load. drawing a polygon will have a different effect on different systems anyway. this is completely absurd! you absolutely can NOT put every texture or material into an equivalent amount of 'polygons' and actually have this mean anything. there are so many other things at play here. i wish that whoever said this to begin with had not put it into all your heads that it means something because it does not. even if it were that easy, and it isnt, that scale would be forever changing. quite possibly the worst measurement system ever conceived.
 
Last edited:
I see you have many versions of Trainz, but have you ever actually used them? The above seems to indicate you have not...

Just shows how much you think you know, i also have every version of TZ and gave up on it over 12 months ago.

Auran\N3V don't listen and belong in the "three wise monkeys" class of business, along with RSC.
 
Welcome to the forums ReggieKray. Can I ask you a question - if you have every version as you say you have, how come they're not registered on your profile?

Shane
 
Just shows how much you think you know, i also have every version of TZ and gave up on it over 12 months ago.

Auran\N3V don't listen and belong in the "three wise monkeys" class of business, along with RSC.
Yes, why are they not on your Registered Trainz Timeline?
 
Just shows how much you think you know, i also have every version of TZ and gave up on it over 12 months ago.

Auran\N3V don't listen and belong in the "three wise monkeys" class of business, along with RSC.


if this were about 'what i think i know' then we would be discussing a different topic. what it is about, is that Vern clearly stated that xyz cannot be done but it most certainly can. i know of no other offering that can actually do those things as well as give a decent appearance.
 
I see you have many versions of Trainz, but have you ever actually used them? The above seems to indicate you have not...

Erm, yes actually I have... aka, if you're going to troll at least get your facts right matey.

Several items in the file library here and at TS.com, including the route I created last year (Mount Hood) which wouldn't upload to the DLS and N3V washed their hands of as to why.

There's also plenty of fallen flags from past route building when everything was going fine in Surveyor but load up Driver with the washed out look through the cab window, laughable physics etc. and I had no heart to carry on.

So as someone who has, over the years, persevered on and off with Trainz quite a bit.
Show me where in the menu or Surveyor you can set up a timetabled run for the player as easily as it can be done in MSTS or Railworks?
Show me where at the end user level in TS12, I can simply tick an option to have the despatcher route my train without having to unrealistically work out and throw every point/switch myself.
Show me where we can magically make all junctions and crossings (including those on gradients) appear correctly drawn with switches and checkrails. This is 2013 not 1999 and it doesn't take many minutes of running Trainz to spot that glaring issue.

I would have expected a bit more respect from someone with a Jointed Rail banner in their signature with, presumably a commercial association with N3V even if that does mean you have to defend the indefensible.
 
So as someone who has, over the years, persevered on and off with Trainz quite a bit.
Show me where in the menu or Surveyor you can set up a timetabled run for the player as easily as it can be done in MSTS or Railworks?
Show me where at the end user level in TS12, I can simply tick an option to have the despatcher route my train without having to unrealistically work out and throw every point/switch myself.
Show me where we can magically make all junctions and crossings (including those on gradients) appear correctly drawn with switches and checkrails. This is 2013 not 1999 and it doesn't take many minutes of running Trainz to spot that glaring issue.

Hi Vern,

Trainz does lack an easy way to implement timetables, but considering the flaws of RW (chief among which would be it's payware-centric environment) I think that's not a very big deal.

I would assume each route in the "other sims" would also need to have paths of some sort to be set up in the route-building stage. To draw parallels with Trainz, this is also possible if the route builder takes it upon himself to create paths with the path rule, so that the end user can easily select a "set path" command and be on his way.

Personally, I still prefer Trainz implementation of turnouts. While crude-looking in older versions, modern content have allowed turnouts to look and operate much more realistically, even added sound at the frogs, which is lacking in RW. With some effort made to line up the ties, the overall results beat any other sim out there. Sirgibby even made creative use of switch machine to animate the switch blades. Sure, building turnouts is much more convenient in RW, but at the cost of precision and detail. Very often the frogs will be drawn erroneously resulting in one rail closing the frog. Track ties, due to the dynamic tie length, appear 2D-ish and flicker in the distance. Physically, perhaps they are more accurate, but overall, a properly made Trainz switch can look better than one in RW.

Cheerio,
Nicholas
 
Gadgets and tablets are not a lot of use for those of us who build routes and assets which is probably what 50 % of users do. We are not all obsessed with giant screens, in fact I find they are totally unrealistic. PC sales are only down due to the "Email and browse, need to announce every visit to the toilet on farcebook or tweet it to to the Twits, using a simpler gadget option that they can carry around. It's a fad, much as the CB radio obsession became in the UK a while back and will vanish when some new toy aimed at the brainless comes along. I reckon it will be some form of holographic communication gizmo. :hehe:

Those who actually use PC's to their full extent will not be dumping them any time soon and will still be purchasing them, all that's happened is the market has lost the portion of the market that just play on social media. That's of course if you accept market figures as being accurate and not fiddled by the likes of Apple Microsoft and the other gadget makers by their attempts to push people in the direction they think is beneficial to their profit margins.

Spot on post Sir!

I've noticed a lot of anti-P.C. news on the totally un-Bias, Bolshevik, Clart news (B.B.C News in case you don't know what I mean). This comes as no surprise as just lately I have notice a lot of fruit logo computers being used on a lot of B.B.C. T.V programs (a lot of those poor, no money east enders types seem to own them, Money saving program reporters seem to use them, lots of detectives and/or spies use them as well!).

Regards.
CaptEngland (Pro-P.C, and non-P.C. if it can upset lefty type people).
 
Erm, yes actually I have... aka, if you're going to troll at least get your facts right matey.

a troll? i simply asked if you really had the trainz experience to call it a simple model railway simulator and say that it cant do anything. hardly trolling, unless you are one of those that consider opposing opinions trolling.

the only simulator i dont have any experience with is Run 8, just havent made it there yet. the others fall sadly short of anything useful. MSTS is a wreck, and RS is a scenery sim with NO operating ability at all. nearly EVERYTHING is wrong with how the trains work in it.

Several items in the file library here and at TS.com, including the route I created last year (Mount Hood) which wouldn't upload to the DLS and N3V washed their hands of as to why.

There's also plenty of fallen flags from past route building when everything was going fine in Surveyor but load up Driver with the washed out look through the cab window, laughable physics etc. and I had no heart to carry on.

that is very sad, perhaps the wrong content was used? given the wide range of content available with only a small amount of it being of any quality level one has to wonder...

So as someone who has, over the years, persevered on and off with Trainz quite a bit.
Show me where in the menu or Surveyor you can set up a timetabled run for the player as easily as it can be done in MSTS or Railworks?
there are assets that allow this with much more depth and detail than even thought of in the trainwatching sims.
Show me where at the end user level in TS12, I can simply tick an option to have the despatcher route my train without having to unrealistically work out and throw every point/switch myself.
if the tracks and signals are properly constructed this isnt a problem. i regularly run upwards of 20 trains in large routes with no problems with origin or destination of those trains. if there does happen to be a hiccup, i address the issue rather than say the single train control minimal interactive sims obviously handle it so much better, when they do not even offer that.
Show me where we can magically make all junctions and crossings (including those on gradients) appear correctly drawn with switches and checkrails. This is 2013 not 1999 and it doesn't take many minutes of running Trainz to spot that glaring issue.
again use the right content. you cant just complain that nothing has improved in over x amount of years if you still use the items from x amount of years ago... try opening MSTS from the box and using it and see what the great features and operations are.

I would have expected a bit more respect from someone with a Jointed Rail banner in their signature with, presumably a commercial association with N3V even if that does mean you have to defend the indefensible.

respect should be earned not expected. i owe you no more than you show to me. starting off by calling me a troll is not getting you off on the right foot.

:)
 
a troll? i simply asked if you really had the trainz experience to call it a simple model railway simulator

Perhaps the word "troll" was a bit excessive then, but if you're going to make a statement at least do some research first, i.e. check what's on the DLS.

As to the rest of your dissertation, well I'm not going to get into a tit for tat point by point argument, this is a forum about a train game not a court of law. You're entitled to your view as I am to mine. If my perception of what Trainz can deliver is wrong or outdated then maybe it's time N3V started documenting some of the features - starting with what all these commands actually do and how to use them.

I understand there are Path and Timetable commands buried away in the session editor but they're not much use there and certainly not without instructions for anyone less than a power user as to how they work. Yes it's easy to set up a non-player set of commands in Trainz and have the AI follow it, what N3V need to be doing is making it easy for players to do with their own train. Forget MSTS and TS2013 for a minute - look at Zusi. Graphics in the stone age admittedly but here you can set up a comprehensive timetable of many services for the route, pick whichever one you want to run and have the AI signaller dynamically path it for you, including a rating based on adherence to timetable and safe/economical driving.
 
but here you can set up a comprehensive timetable of many services for the route, pick whichever one you want to run and have the AI signaller dynamically path it for you, including a rating based on adherence to timetable and safe/economical driving.
And is there and option for dynamic change of timetable according to payer's speed?

May it be in England, every train (incuding cargo) is going according its own timetable, but in USA or Russia cargo trains have no timetables and travel according to line workloads (and here pre-calculated
timetables are unacceptable).

Yes it's easy to set up a non-player set of commands in Trainz and have the AI follow it
it's relatively easy to create set of rules for train "to be driven to specified marker", and with "TimeCheck" rules they can realize the same behaveor as AI. It is not so easy to search for "most optimal" path, because nearly impossible to determine which spline junction is setted forward and which deviates the path. But ! You can't pre-calculate AI motion, because it is too complicated in comparison with AI of MSTS, RW, or Zusi (which motion is predicted by timetable). So you'll have to set all timetables by yourself in any case. And to prevent all train conflicts also by yourself.


I would assume each route in the "other sims" would also need to have paths of some sort to be set up in the route-building stage. To draw parallels with Trainz, this is also possible if the route builder takes it upon himself to create paths with the path rule, so that the end user can easily select a "set path" command and be on his way.
Or to create self-generated paths between signals. Both things are required for automatic solutions of train conflicts.


Personally, I still prefer Trainz implementation of turnouts. While crude-looking in older versions, modern content have allowed turnouts to look and operate much more realistically, even added sound at the frogs, which is lacking in RW. With some effort made to line up the ties, the overall results beat any other sim out there. Sirgibby even made creative use of switch machine to animate the switch blades. Sure, building turnouts is much more convenient in RW, but at the cost of precision and detail. Very often the frogs will be drawn erroneously resulting in one rail closing the frog. Track ties, due to the dynamic tie length, appear 2D-ish and flicker in the distance. Physically, perhaps they are more accurate, but overall, a properly made Trainz switch can look better than one in RW.
Agree with you. But for signfcant facilitation of scenery-junction placement and their compatibility with old "spline junctions" already proposed to add support of scenery-junction without attached lever and opportunity to add a lever on it.
 
Last edited:
Perhaps the word "troll" was a bit excessive then, but if you're going to make a statement at least do some research first, i.e. check what's on the DLS.

As to the rest of your dissertation, well I'm not going to get into a tit for tat point by point argument, this is a forum about a train game not a court of law. You're entitled to your view as I am to mine. If my perception of what Trainz can deliver is wrong or outdated then maybe it's time N3V started documenting some of the features - starting with what all these commands actually do and how to use them.

I understand there are Path and Timetable commands buried away in the session editor but they're not much use there and certainly not without instructions for anyone less than a power user as to how they work. Yes it's easy to set up a non-player set of commands in Trainz and have the AI follow it, what N3V need to be doing is making it easy for players to do with their own train. Forget MSTS and TS2013 for a minute - look at Zusi. Graphics in the stone age admittedly but here you can set up a comprehensive timetable of many services for the route, pick whichever one you want to run and have the AI signaller dynamically path it for you, including a rating based on adherence to timetable and safe/economical driving.

right, so we are in agreement then, it is possible, and just because you dont know how doesnt mean that it doesnt exist. as for doing something like Zusi, i am sure that it is possible, but it may not be an easy one click solution. it would require very specific instrutions for a specific route. another place trainz shines, is that there are so many possibilities and route creation is hardly limited. this is good and bad. bad for the kind of thing you want there, but good for flexibility - that is not matched anywhere else.

if it hasnt been made yet, maybe you can petition some creators with interest to do it. i know of a certain complex system that is amazing in it's ability to pre-determine even weeks' worth of trains.

one of the best things about trainz is that it is very flexible, and nearly anything you want is possible, but you either have to construct it or find someone who has already. with the worldwide user base and talent that is usually obtainable. you dont really get that with any other package, you get what you get, but that is about all.

now dont get me wrong in all of this, trainz does, can, and will fall short in a lot of areas and i am in no way implying that it doesnt. but i also think the experience is what you make it.
 
If I get a chance I'll try and play around with pathing and timetable rules, but I still feel there should be an easier way of doing this for player trains.

With regard to TRam's point about appearance of points, the original question was whether Trainz needs a new graphic/game engine which by implications means probable loss of backward compatibility. A line could be drawn under TS12 for those who wish to continue using legacy items, however I would be more than happy to start over with a new Trainz with much better graphics and features as discussed.
 
I understand there are Path and Timetable commands buried away in the session editor but they're not much use there and certainly not without instructions for anyone less than a power user as to how they work. Yes it's easy to set up a non-player set of commands in Trainz and have the AI follow it, what N3V need to be doing is making it easy for players to do with their own train.

So you say Path and Timetable commands are hard to grasp. Be that it may, they are still much simpler to learn than it is trying to build anything worthy with the "other sims".
 
Back
Top