Personally I believe 3D could work. First of all, remember that the game could easily be played from the isometric perspective using a 3d engine. Indeed, many of the earliest 3d strategy games did this (total annihilation etc).
For those who argue it'd "only make prettier graphics" - not necessarily. Due to the nature of 3d, you'd be able to use views other than Just isometric, you can view from every other angle too, including directly above (satellite view!), from your trains as the driver/engineer/whatever, as the pilot of that nice snazy concorde crusing over your network etc etc. With a properly done 3d engine there many more possibilities for the view than just isometric. And with alternate views there usually come new gameplay forms.
Disclaimer - I'm not a "graphics snob". I doubt anyone who plays any current form of TT is. I mean, 256 colours!
im not personally a graphics snob.. i fully agree with some of the views here, it could break the gameplay a lot. what i had in mind was to keep it pretty much like it is now, not even use jazzy high poly count models just things that look like they do now to try and keep the integrity of the game, but the addition of the camera movement for me would be an incredible addition.
I dont for one minute believe this would ever look as good as say RailRoads did which IMHO totally blew me away with its graphics and gameplay, but i do believe a 3d version of transport tycoon is very feasable to do and in a relativly short amount of time, since the days this was written development tools have come on so much to aid the speed of development, knocking this up in C# would be a breeze, getting the game working smoothly would take a lot of tweaking and testing but i believe with a few community koders this could become a reality very quickly.
I dont expect any of the devs here to help out and i understand the reason why most probably wouldnt be interested, but it would be good to try and convince some of you guys that 3d is maybe a posibility if done correctly.
I'm all for 3D (provided it's done correctly), but could you please forget about doing this in C# (as it is interpreted bytecode (like java is), which is too slow for a processor/memory intensive (manipulates a lot of data per second) game like openttd); and while you're at it, could you please make sure it remains cross-platform. However, if you intend to stick to C# (I hope you chose that language with some decent arguments (no, "because that's the language I know" doesn't fly here)), how do you intend to keep the performance impact (due to C# being interpreted bytecode) to a minimum?
While I see the potential of 3D, I think you're in for a lot of work (even with all those fantastic development tools). Are you prepared for that? Are you really dedicated to doing this, or were you expecting to complete this project in a few days or weeks time?
How do you intend to make the camera controls non-intrusive, yet intuitive? Have you thought about situations where the player rotated the camera around to some weird angle and all of a sudden wants to pan? How do you want to handle zooming gracefully?
As for detail: If you use multiple models for the same object (each with a different level of detail) you can have a lot of detail when an object is near the camera, less detail when it's further away and nearly no detail when it's far away (this taxes the rendering engine less, while the player can have a lot of detail configured).
Expresso wrote:[...]please forget about doing this in C# (as it is interpreted bytecode (like java is), which is too slow for a processor/memory intensive (manipulates a lot of data per second) game like openttd[...]
We're currently rewriting our 3D CAD software from C++/OpenGL into C#, using Managed Direct3D (XNA in fact) - and believe it or not, the performance is completely comparable - if not faster in C#. Not a totally comparable application, I agree, but I was positively surprised by it.
I dont think your comments on the programming language are constructive at all, if you did any research at all into c# and the .net language you would know that the speed barrier is no longer as high as it used to be. C# could manage TT 3D and more with no problems whatso ever. I am capable of doing this in C, C++ c# java VB and i still believe C# is my best option for this. No i do not believe it will take a few days or weeks or months for that matter.
Thanks for the comments so far but i dont think this is the right place to discuss TT3D. its quite clearly a community who feel threatened when they really shouldnt be. i honestly thought there might be some support and backing for keeping the transport tycoon name alive but clearly not. Ill take the project elsewhere thanks.
conkerjoe wrote:Ikeeping the transport tycoon name alive
Note that OpenTTD does not use the transport tycoon name at all - it's a registered trademark of Microprose. It's called OpenTTD, not OpenTransportTycoon!
@conkerjoe:
first of all: if you can't take criticism, then you shouldn't be here at all. this is a forum with loads of people with different points of view. A few didn't like the idea and you'll have to accept this.
Woking in a forum requires you to be able to not let get people at you.
I write this here because I've witnessed the almost downfall of another forum because people got on ewach other's throats constantly so the admin finally left becaus he couldn't stand it any more.
"Ignorance is bliss"
Second: you'll have to use a language (and libraries) that can compile on mac and linux as well. Otherwise you'll never ever get it in trunk.
Now keep up the thought about 3D. I'm all for it.
Aydan
Aydan wrote:Second: you'll have to use a language (and libraries) that can compile on mac and linux as well. Otherwise you'll never ever get it in trunk.
Not to forget the other supported platforms like OS/2 and MorphOS.
Expresso wrote: forget about doing this in C# (as it is interpreted bytecode (like java is),
*This* is patently wrong - all .NET languages, among them C#, are compiled into byte code (IL), which is then compiled into machine code by the JIT compiler when the program is started (or ngen'ed). After that, it's full speed.
and while you're at it, could you please make sure it remains cross-platform.
The language does not matter, because by using Direct X a game is not even unix compatible any more. Neither MAc, to mention another platform.
Even more, if your focus is only on 3D, then the language may be not of not an issue. However, if you want things like passenger destinations, then the performance of your program code is really essential. (I would really like to know how the automatic garbage collector will cope with millions of 32 Byte structes allocated, copied and destroyed per minute.) This is another reason, why also Java would be challenged a lot with OpenTTD (at least compared to a language with pointers and machine dependent compiler omptimisations beforehand.)
OpenTTD (and related games) may appear easy and not so CPU hungry. But with large maps or many trains this changes. Only the indication "needs no 3D card" is not synonym with "does not need any processing power".
However, I think the idea of having a 3D interface nice. I admit, that the sprite calculation is deeply interwoven with the tile logics. Changing that seems to me more straight forward than starting from scratch. But I wish you good luck, and I hope you are not going the same route as zugspiel, SuS, or vlatkosim. At least you somehow passed transport empire ...
I like to look at great maps and see how things flow. A little like a finished model railway, but it is evolving and actually never finished. http://www.simutrans.com
prissi wrote:The language does not matter, because by using Direct X a game is not even unix compatible any more. Neither MAc, to mention another platform.
Who said anything about D3D?
There's openGL libraries for all platforms.
The webpage said so http://blogs.warwick.ac.uk/tt3d/, I was under the impression it is this one. However, I can be totally wrong, but this was also written in C## with instanced Direct X. I am pretty sure, that OpenGL will not be very much supported in C#, at least by Microsoft.
I like to look at great maps and see how things flow. A little like a finished model railway, but it is evolving and actually never finished. http://www.simutrans.com
conkerjoe wrote:...its quite clearly a community who feel threatened when they really shouldnt be.
I don't feel threatened, neither anyone else does.
conkerjoe wrote:I honestly thought there might be some support and backing for keeping the transport tycoon name alive but clearly not.
As I previously said most have set other priorities. But as you can see some are willing to join your project. And for sure more will get attracted in the future.
conkerjoe wrote:I dont think this is the right place to discuss TT3D.
I believe it is.
conkerjoe wrote:I 'll take the project elsewhere thanks.
Not a wise idea to my opinion. Of course you are the one to decide.
I personally think that 3D could work. Not the type of 3D you see in RRT3D or similar games, but a more gentle step would definitely add something to the game without spoiling gameplay.
Imagine that it is possible to rotate the camera by any angle, without altering the field of view itself. So the camera is only rotated around the vertical axis, making it possible to view buildings, vehicles and such from any direction. The lack of changing the viewpoint in (O)TT(D) at this moment is not really a problem, but viewing your world from different possible angles wouldn't hurt anyone.
Of course, the best way to achieve this is to use a simple 3D engine. If I would know more about 3D graphics programming I would have tried an implementation already, so maybe in the future...
Maybe a proof-of-concept could change people's minds, I would suggest: let's just try it.
Also, GNU is working on a replacement for M$'s .NET, called .GNU (dotGNU). It' pretty young right now, but I guess you could make use of it as time progresses?
Besides dotGNU, there is mono, another .NET implementation. I even didn't know about dotGNU; it seems that mono is much more popular... All I know is that dotGNU is sponsored by FSF (link) and mono is written by Mexican bisexual midget, Miguel de Icaza
from that video on the previous page 3D does not look right at all. i am FOR 3D but not when it makes the graphics worse than being in 2D therefore i support 3D if they up the graphical detail of buildings and transport to not look blocky it reminds me of a poor visual of google earth.
I've started to work on some "3d" port using opengl with a buddy of mine, not really actively, just for fun whenever we have some time and feel like it. Personally I really hate pure perspective 3D for such kind of games because it makes interacting with the world harder and it's impossible to make it look good and detailed without *severe* performance impacts. The idea we have is to go along the route of sc4 with a fixed-angle isometric camera (with a fixed camera you can embed a lot of detail into textures). Maybe trimetric like sc4, though for now I'd try to make it look as close as possible to the current graphics style. And you'd still benefit from a rotating camera (4 views), hardware-acceleration and if it's done right one would also solve the problem of needing sprites for every possible combination of things. As for models I think we could maybe use all those neat blender models which people made for the 32bpp-sprites (if allowed).
Rather than 3D, I'd like to have rotatable 2D like Chris Sawyer did in Roller coaster Tycoon. Basically the same as OpenTTD view, but rotatable by 90 degrees.
It would be very nice to rotate around a city to get a better view to build a new station, where we now have to resort to making the city transparent in order to see what we're doing.
All display code stays the same, except how you render from the internal map to the display (front/left/back/right change). All graphics that rotate by themselves (vehicles) need no change, static buildings however need to be drawn 4 times rather than 1.
With respect to graphics, I really like the hand drawn graphics (not in the last place because they are appropiate for a game from the 80's). Compared to generated 3D rendering, there really is no competition imho. Too bad it is so much work.
<wild thought>
Maybe we should raise some funds to pay somebody to make the graphics, and rid ourselves of the requirement of TTD .grf files? As time goes by, it will become harder to obtain this data for new users. Also, we would become a truly free game, which may attract new people.
</wild thougth>
Albert
Being a retired OpenTTD developer does not mean I know what I am doing.
