-Larry
>> > Three years is actually a long time in sim development.
>> For a company like Simbin or ISI that is using a ready made engine and
>> has a team of nearly 50 people involved, yes. RL is a three man show
>> and they use no existing engines at all. Like I said, I spent a lot
>> longer than three years just on my physics engine. In fact, I'm at the
>> six and a half year mark right now and am still working on things.
>> IIRC, LFS was in development for 5 years before they put out the very
>> first demo, and it didn't go on sale for some time after that. If S3
>> is their "final, finished product" that they set out to make, well,
>> you're looking at something like 10 years total. It cracks me up when
>> folks that have never written a sim before argue with me over how long
>> development tasks should take <bg>
>> With the
>> > development of hardware (processors, gpu's, physics cards, etc) always
>> > evolving, it's very difficult to create a sim today that will still be
>> > considered "state of the art" three years from now.
>> Your own development experience backs this up?
>> Anyway, I disagree. A system I wrote recently won't be able to run on
>> PC's in a game environment for many years to come. But it's sitting
>> there ready to go when the CPU horsepower arrives. In game development
>> you ALWAYS plan things for future hardware that will be out when you
>> think the project will be finished. Just par for the course and very
>> routine.
>> Besides, RL has
>> > supposedly "been in development" for more than three years -- it's
>> > simply been three years since the transport was "showcased" on the
>> > website. I would have assumed that the sim was pretty close to
>> > completion three years ago if the developers had time for details such
>> > as that!
>> The transporter took Tony probably a week or two to model. This
>> delayed RL absolutely zero. You need to code the engine to a certain
>> point before you need to worry about artwork at all. The artist can
>> not go into any serious production mode during this period.
>> If I recall correctly, in Virtual RC Racing we were coding for four or
>> five years before Tony came along and did all our models. He did the
>> artwork so fast we were still coding when he was finished with all the
>> game content. We have ten or twelve tracks included in VRC that were
>> done in no time. By the time we released it he had another ten or so
>> ready to go for expansion packs and so on. I think he's got something
>> like 25 or 30 tracks finished for us now. He could have done 100
>> transporters in that time and it would not have effected our release
>> date (or RL's for that matter) one minute.
>> Argue with me if you want. Let's see what you've personally done in
>> sim development first though ;-)
>> I know that GTR, GTR2, rF, GTL and LFS has
>> > kept me plenty busy during that time!
>> Me too :-) Fun stuff for sure. My personal favorites are LFS and GTL
>> :-)
> Todd,
> Most of what you write is true, but I still have seen no evidence that
> RL will ever actually be completed and available to the public. I'd
> love to be proven wrong though...
> I agree that developers have to code for "future" hardware, but there
> is a practical limit to that, since it would be difficult to predict
> what hardware advances will be available 5-10 years from now. Should
> you code for faster processors or multi-multi-core? How many poly's
> will our graphics cards be pushing in 5-10 years? What new physics
> card technology will be available? etc...
> Anyway, I'm not saying a 3 man team "should" be able to create a
> state-of-the-art sim in three years or less. I believe it could take
> many more years than that. I just don't think it is a wise business
> move for the reasons I stated above. Now if they are doing it just for
> fun or a learning experience, then I say more power to em!
> Regards,
> CK