no idea what Cartest for DOS does, but...

Hmm, but one of the issues is always accuracy. Or, more exactly that you don't know whether the simulator is accurate in the conditions that you are using.


But, I'd be a lot more interested if it didn't involve Moonlight/Silverlight. I suppose its not as bad as running Windows, but I couldn't really describe it as an advantage.

CarTest describes itself as "CarTest is an automobile maximum acceleration computer simulation. The user enters the specifications for an automobile and CarTest will mathematically accelerate the car from a standing or rolling start until maximum speed is reached. The simulation is time-based and event-driven. That is, the acceleration times are not explicitly calculated but, rather, time progresses at small increments and the usual events that take place during acceleration: clutch dump, wheelspin, shifting, etc. happen in proper course. The acceleration times are arbitrary milestones that are reached during the car's response to these events."

A bit Yes and No (IMHO), perfection doesn't exist, even in real life you can not do 2 exact times two times, but the sim I am working on use environment conditions to help math real life (ex: altitude over sea, air humidity, barometric pressure, temperature)

And lots more like:
engine readline
power at crank or at wheels?
hp @ rpm
torque @ rpm
gear ratios up tu 8 (for GCC version, up tu 14 for SL version)
Final Drive Ratio #1
Final Drive Ratio #2 (for supercars that switch for Super high top speed)
torqueloss % for drivetrain
clutch slip factor
weight overall
weight distribution
gravity center
wheelbase
drag Coff
frontal area
traction tire Diameter
non traction tire diameter
intertia for tire traction
intertia for non traction tire
rolling factor
static factor
environment temperature
Barometric Pressure environment
car Roll Out
Launch at what RPM?
Manually shift points or computer calculated for efficiency
time to engage new gear?
FWD? RWD? or AWD?
car height
car width
tire width
tire profile
rim size
engine type (NA?, Forced?)
boost starts at? (for Forced)
and lots more are computer calculated, because are so hard to setup even for some advanced users)
and more to keep posting here.

Yes, I understand that, thats why I started everything at my linux box with GCC, right now, I can do something like this to show results.

./nxgtrsim 7500 0 140 6400 3.33 1.83 1.29 0.98 0.76 0.0 0.0 0.0 4.18 0.0 15 1.0 2461 40 25.9 95.7 0.29 18.18 18.52 18.52 0.33 0.33 0.00 0.013 1.0 65 29.38 0.0 3600 7470 7160 6930 6810 0 0 0 250 250 250 250 0 0 0 130 4800 0 1 51.8 66.1 195 55 14 0 0

Of course, it needs to be polished if its something Linux community needs.

Sorry you can't be serious about every car that exists unless the result will be average not precise. Take for example Live For Speed team(not one person but team!) - they have difficulties even finishing proper tyre physics not to mention it took ages to do stuff for each car. Speaking of LFS - you should try speak to them and maybe make Linux version or port for that game. Sure you can make some game that works but if you call it simulator it needs close to 1:1 to reality. If you are alone and without resources + explore fieldwork and paperwork regarding math and physics + know computer programming language it is not really possible to complete.

Maybe simulator was the wrong call? I dont know how else to call it, CarTest calls itself a "simulator" and has been arround 17 years now, the code I have been working its 8 years mature by now, and does what CarTest or more.

I know computer programming, actually, the question is not if I could make this kind of software, its actually done already and working with a +90% match to real life (error is caused for not precise parameters too, not just math).

The real question was if it was really worthy to create a Linux version of the Silverlight Win/Mac version that is working right now at web.

If there is demand for an application like yours in the Windows world I would assume that there is also demand in the Linux world, and if there currently no such application for Linux we can only benefit from a port.
Which framework you use for your application is absolutely up to you, IMHO, but if you use a framework like moonlight (the open source version of Silverlight) it may be that some people don't want to use it because it is a rebuilt of a proprietary framework.
Somewhat the same as the Mono/dotNet thing.

Yes I agree 100%, thats why I was wondering sharing the GCC command line edition, before even Silverlight exist I started all this at my linux machine with GCC and command line only.

I was thinking on fixing that version, get it up to date features, and release it, but thats tons of work and wanted to know if was worthy or maybe it was too few demand in Linux compared to Win+Mac users which are already covered.

Linux version? Definetly! Version that requires Silverlight as rule + waste of real time in years as you mentioned? No! Unless you get paid enough.. Do you have working version that runs without Silverlight or other addon?

Yes, the GCC command line one.

However it's a bit outdated (1 year behind SL version) and maybe I am the only one that can use it because bad usability, but that both problems could be fixed.