JET Engine and a good example of progress [pics]

[nicky9499]

Hello guys,

I originally tried to post this to Suggestion Boxcar but after seeing alot of one-liner "suggestion" threads pop up I assumed mine was discarded from the approval box.

In light of the recent "should we update the JET engine" discussion which has appeared once more, I thought this was something relevant and we should think about it for a bit. I will state for the record that I'm not a programmer and these are just the views of your average Joe in the best interests of the Trainz hobby. Professionals or the more technically inclined are most welcome to offer your thoughts on the issue. Let us begin.



Once upon a time there was a bunch of truck simulator games. It was pretty crappy by today's standards. For the purpose of this discussion we will only focus on the graphical aspects. These games in question had rather primitive graphics; no dynamic lighting, no shadows, even its vehicle models used ground plane shadows. So, not very unlike Trainz at present. The company behind them is a bunch of Czech folks called SCS Software, a tiny company of only 20 developers. So, quite similar to our brew crew as well. What I'd like to bring to your attention is the work they've done in the past 1 year or so.

The below images are from a forum member complaining about the excessive use of recycling in the latest iteration (ETS2). But, from those same images we can also see massive graphical improvements, possible only by updating the game engine. Note that although the assets shown are essentially the same, it was an update to the game engine that truly brought out the realism.

The current argument against an engine update is that doing so will break compatibility with all the existing assets. Well, these guys have shown not only is it possible, it doesn't take a mega corporation, millions of dollars or tens of years to achieve it. It is a most prime example of an engine update done right. In the process they have implemented an array of graphic features which rival that of big-budget blockbuster games.

I have owned both GTS and ETS2 an both have provided me many hours of fun. It is true that a large proportion of assets are reused, such as road pieces, scenery objects, industries, vehicle models and gas stations. However having played both I can tell you that ETS2 feels completely different and massively more realistic. (This is also because of rewritten physics, gameplay and interface, which make SCS' feat even more amazing but is beyond the scope of this discussion.) Trainz is blessed with many highly talented content creators who bring to us hugely realistic assets for use in the game. As technology progresses more users are also upgrading to more powerful computers. Despite this, oftentimes Trainz is prevented from achieving its full potential because of drawbacks and limitations in the JET Engine.


My point is, it is possible and probably within N3V's means to breath new life into the prehistoric JET Engine. It's called progress and we're stuck on technology from 2001. It's just a question of whether or not they are willing.


Just drop a friendly email over to the chaps at SCS and ask for some advice how they did it. They're very amicable folks. In fact, while at it, N3V could also take a few other leaves from their pages; their active community interaction, their strong support for modding, their regular development and behind-the-scenes updates, their "nightly-build" betas and most importantly, their acknowledgement of their user community and fanbase. SCS is a model game developer and N3V would do well to learn from them.

 

[oknotsen]

Good info, Nicky. I hope a nice and clearly argumented post as the one you did here will be taken serious or help in the consideration to do something about the engine.

 

[mjolnir]

Nicky9499, I would submit that we don't have enough information about the differences between the Jet engine on the one hand, and the game engine (engines?) for the truck sims on the other, to be able to make any conclusions about what differences changes in the Jet engine would have. You show that the developers of the truck sims reused content, but they might have used the same content with an entirely different game engine. And that might have been part of their plan from the beginning.

In any case, the degree to which the developers of the truck sims were able to do it, is not necessarily a good predictor as to how well the Jet engine could be upgraded.

ns

 

[nicky9499]

Hi NS,

The truck simulation titles mentioned all run on the Prism3D engine, which was based on OpenGL and has been in use since at least 2003. The discussion I wish to bring up here is not strictly "how" one should go about updating a dated engine, but of feasibility. A similarly sized company has taken a similarly old game engine and has managed to do fantastic things with it. Given the number of highly technical threads I've seen around here, we have no shortage of members who know their way around such things. Surely if N3V would just start working with the community we can get things started much more quickly.

Herein lies the biggest problem with N3V. They refuse to cooperate with their user base. We're not all a bunch of freeware-demanding kids around here, there are also many real experts who have written lengthy posts of thorough analysis and suggested solutions to solve various issues with Trainz such as database inefficiency and such. As a casual observer I can't help but feel like they're all out ignoring us and doing as they please. Service Packs have turned out to be complete fiasco. Restrictive DRM is kicking in and even that is causing issues for many who purchased DLS. Add-ons manage to misspell road names. Forums and download repository is regularly offline, and the list goes on.

On the other hand we also have a small independent company, but one which demonstrates a willingness to work with its community to make the software better for the benefit of all. They have a blog, regularly updated and full of interesting snippets of upcoming features, updates, content and whatnot. They have a small forum of their own with a comparatively small membership, but the developers participate in discussion and feedback ranging from stuff like bugs and scenery discrepancies to localization and controller support. It's less important what was discussed, the point is the developers are one with the community. I am having a discussion with an actual dev instead of a mouthpiece or spokeperson. Sorry, but these days "we'll pass it on to the devs" is bureaucratic vocabulary for "suck it and get lost, we don't wanna hear it".

Then there's updates and beta testing. Referring to SP1, I came across what one user posted to the lines of "N3V, you should beta-test to within an inch of your life before releasing an update, this is wasting all our time". Beta testing is a big hassle, I'll grant you that, but you know what else is awesome? The community. SCS Software faced the same problem that N3V probably did; dammit how are we ever going to catch all them bugs? And then they thought, why not let our users do it for us! At the end of the day, Beta testing is a numbers game. There are x number of bugs and you need y number of people/hours to catch them. The company has the skill and resources to fix said bugs. The community doesn't. The company doesn't have the manpower or time to spend catching all these bugs. The community does. Do you see what I'm getting at?

Many users (especially since multiple installations of Trainz can coexist side by side) would be more than happy to help beta test an upcoming release and submit bugs and report logs for the company to fix. This is why today we see SCS having a "nightly-build" Beta release for all users who wish to help test new features and updates.


I know what some of you are thinking. This Nicky is praising SCS to the heavens, is he getting paid or what? No, I am not. I am just a casual observer who indulges in software from both companies. And as a casual observer what I observe is SCS doing everything that N3V should and can be doing as well - for the betterment of the hobby.

 

[j_maybury]

Nicky, I can see you are passionate about Trainz and want the best solution for Trainz to prosper, but things are not as clear cut as you make them out to be, and If you remember TRS2009 was a open beta project, but the problem was that they ended up with a deluge of inaccurate bug reports, but I won't go into beta testing, as all beta testers have signed an NDA so the subject is mute. but I would like to ask... In the two screen shots of the loading depot, is it the same depot station? I ask this because the background and foreground differs considerably, and dose this mean all those assets had to be scraped?

 

[nicky9499]

Hi Jonathan. If I remember correctly, SCS rebuilt the map to a much larger scale so background in the screenshot is most probably in a different location, but the many of the models used including that depot and the truck interiors, right down to the road prefabs in every city, are recycled with little or no changes.

N3V could be suntanning on Bondi Beach, sipping iced lemonade and hoping people forget about SP1 soon, or N3V could be working very hard on a new engine right now. Nobody knows. Members of the brew crew hardly ever drop by, the newsletter is basically nothing but advertising, and the "blogs" they maintain are a joke. So maybe it's not so clear cut but if that's the case, tell us, your users! We'd love to hear from you!

So updating a game engine is no small task. No sh1t. But two minds are better than one. Here on the forums we probably have more than a handful of users who know a thing or two about game or software development. Just look at some of the recent threads discussing TADDaemon and database efficiency. The information provided (by users) is mind-boggling. These people know what they're talking about. Sometimes they probably get it slightly wrong in it's implementation within Trainz, but geez I wonder why that is so. Oh, because the devs never ever share anything. It takes time and effort to write out lengthy analysis, explanation and suggested solutions to address such technical issues, and what were they met with? One-liner replies by Mr Windwalkr or outright silence.

This is sad.

 

[clam1952]

Need to consider that just because some one has apparent technical knowledge of programming does not mean necessarily that it is applicable or even possible in a gaming or simulation environment.
By all means if people think something can be done better then suggest it but why the need to get all snotty, uptight and start slinging mud around when told it's not going to happen or not just yet?
This is not an open source project, which from experience is where you often get this sort of attitude. The I know more than you do, I am right and you are wrong, I must have the last word brigade.
For all we know N3V's dev team may well be working on a next generation of Jet, however they are not likely to tell us about it until there is something to tell. Game engines take a bit longer than five minutes to develop.

Personally I would prefer they get TS12 SP1 sorted out before doing anything else. Then dump the legacy support and update the engine.

 

[clam1952]

One-liner replies by Mr Windwalkr or outright silence.

Ever considered that he may be limited in what he is allowed to agree with or divulge regarding technical aspects of the game.

 

[j_maybury]

Clam,s right, I am sure Chris would love tell to you all the good news about what is going on, but all he can tell you is what he's allowed to tell you, and it is no different for us beta testers, he can only let us know what we need to know, and we in turn have to keep that to our selves, under the NDA. Why not sign up for the next batch of beta testers when they next advertise for them? you would understand better then.

 

[j_maybury]

By the way Hats off to SCS for making a sinking ship seaworthy again. and thanks for the post Nicky. A bit of hope is always welcome.

 

[nicky9499]

snotty, uptight and start slinging mud around

I am sorry you feel this way although I'm puzzled from where you derived this conclusion. Nowhere did I also request or demand that N3V update JET and get it out the door before the end of the week. However your point (and Malc's too) regarding an NDA is something I haven't thought of. Which begs the question, why such a restrictive NDA in the first place? Not revealing source code, I perfectly understand. However, beta testers should be free to discuss new features and such. Putting a gag order on everyone just makes the company look like a selfish kid at the corner of the playground who refuses to talk to anyone and keeps all the sweets to himself. Besides, it's not like we're beta testing a new version of NSA's surveillance software or anything.

If I am crossing the wrong lines or asking the wrong questions here, please advise.

 

[j_maybury]

No it's pretty much standard practice so that the opposition can't get a jump on what the others are doing, it,s all part of corporate spying.

 

[whitepass]

My NDA for TRS4 was up years ago so one thing that the NDA is for is a beta tester telling about thing he is testing that never get released. In TRS4 there were moving passengers that would board the cars, it never worked right and there was at lest one other loco and route that never made it in. Just think of the post if everyone had known that back in 2004!

 

[kenneth18278]

No it's pretty much standard practice so that the opposition can't get a jump on what the others are doing, it,s all part of corporate spying.

Except that Trainz doesn't really have any genuine competition. MSTS and Failworks et al aren't simulators in the sense that Trainz is, and won't attract the same customer base.

Regards,

Ken

 

[clam1952]

I am sorry you feel this way although I'm puzzled from where you derived this conclusion. Nowhere did I also request or demand that N3V update JET and get it out the door before the end of the week. However your point (and Malc's too) regarding an NDA is something I haven't thought of. Which begs the question, why such a restrictive NDA in the first place? Not revealing source code, I perfectly understand. However, beta testers should be free to discuss new features and such. Putting a gag order on everyone just makes the company look like a selfish kid at the corner of the playground who refuses to talk to anyone and keeps all the sweets to himself. Besides, it's not like we're beta testing a new version of NSA's surveillance software or anything.

If I am crossing the wrong lines or asking the wrong questions here, please advise.

I wasn't referring in any way to you but some of the somewhat personal comments made by other members of the forum which are either uncalled for or display a total lack of knowledge as to what is involved and the time taken in creating a game. I was involved in game creation and mods last century albeit open source on Linux, so tend to sympathise with developers. I gave it up as my programming skills were not good or inventive enough for games. Nothing came of the project in the end as like many open source ventures people lost interest and it fizzled out. If I could have produced anything half as good as Trainz I would have been a happy person..........

NDA love it or hate it, is a fairly common requirement for any closed alpha or beta test as opposed to an open beta test. In the past I've tested for large companies both Games and Applications and you do not disclose anything or should not when on an alpha or closed beta test, due partly I have always assumed, as it may be useful information to a competitor and the more obvious is what you are testing doesn't always make it into the final release. Plus an NDA agreement in theory should stop you uploading an un-released version on the net.
As happened with Valves original Half-Life when someone uploaded the Beta, wasn't me by the way and we all know what happened next, the creation of that infernal Steam junk they now use.

Open Beta tests used by some companies don't have NDA agreements however you will probably find that it was first tested under a closed Beta.