Most newer games that utilise DX10 only do so in a fairly peripheral add-on way, like ripples in the water in Bioshock for instance. I'd be pretty surprised if the graphics were entirely rewritten for that to occur.
If we're talking about the supposed benefits of calculating physics on the GPU in dx10, then the only example I've seen is the little 'cascades' dx10 demo. I guess there are a number of avenues for physics crunching, the right one will be the one which is most compatible with the majority of systems.
physx is a complete dead end for racing sims unless youre doing a sequel to stunts where you have cars crash into walls made from cardboard boxes
i do admit that lfs plain sucks at what physx is most useful for but usually the number of rigid bodies colliding in lfs is small or very small so that code can easily be run on a sequential core
my hopes are with nvidia and amd pulling together to somehow unify cuda and streamsdk into one well rounded parallel computing extension to the x86 specs which would enable scawen to run bits of physics on massively parallel cores the way hed like to instead of the way ageia tells him to
well his statement is not that far off, but what that implys is stupid. You can't cut the support for "older" operating systems just because there is a new one... Even MS know this and so does every other proper Software company.
Yep, XP is due bug fixes through to 2014 from what I know. That said it still might be the main OS by then lol
Intel has BANNED Pista internally and as Windows 7 is just Pista core again looks like they will avoid that as well. If one of the most tech companies in the world with FULL access to the source avoids the OS what does say about it.
Unless MS gets Win7 right it could be the "end" of MS in the same way that IBM was top then died. My new lappy is Pista based and I have to say not too impressed, that is with a machine capable of running it as well (2.5Gh Intel Dual core, 2Gb Ram, ATI HD2600 with 512Mb dedicated and 250Gb drive)
PhysX can only really be used for 'bells and whistles' until *everyone* has that capability in their PC. In games like UT3 that use it, it's just for extra effects and can be disabled. If it were used for the physics in LFS it would be an integral part of the gameplay, so you couldn't just turn it off if your card couldn't do it.
It makes me laugh how people hate Vista so much. Personally I've been using it for almost a year and a half now, and I've never had any trouble with it. It's certainly a damn sight more stable than XP ever was. That said, I am using a 64 Bit version of Vista, which is more stable / reliable than the 32 Bit version. On recent hardware, Vista rules all, but if you have oldish hardware, it's a waste of time trying to use Vista, keep XP. People are just stuck in their XP ways, and fail to learn how to use / optimise Vista properly ...
More on topic, to see LFS updated to take advantage of DirectX 9 would be nice, but it's not the be all and end all. What I would like to see the most if LFS taking advantage of multiple CPU cores, as they are getting very common now a days. As for the PhysX side of things, Ageias PPU may be fairly dead, but the PhysX technology certainly isn't, especially now nVidia has taken it on. I dunno what advantage it would really bring to LFS, but it would be no harm to impliment some extra effects for those who have compatible hardware ...
You might have got lucky with your given setup. As I said earlier, one of the MOST tech companies in the world today (Intel) with FULL access to the OS source code refuse to deploy it in the company. Currently MS feel like IBM before their huge fall from grace for those old enough
As you say, back on subject...
Multi core support in LFS is a must now. In a way you would think LFS would be prime for it, even if only to split the physics cores for different cars in the race onto different CPUs to load balance.
PhysX support would probably require too much rework to make it worth while. TBH, PhysX is dead in the market until there are enough new GPUs with correct drivers and hardware to support it, something that is many years away.
At the current time the dedicated PhysX card itself is just a fancy heater than consume power and does nothing. No real software support and it suffers catch 22. You can't play online with some people on higher grade physics and others not due to all the problems that brings. Something that Voodoo cards did noy suffer from on release as they were visual only so no mismatch.
PhysX is a case of "Here is the solution, what was the problem?"
Blah blah blah...when will the "Vista Sucks" argument end? Vista sucks just as bad as XP did when it first came out, remember all the issues? Probably not...I guess it is Microsoft's fault that Vista was put on systems that didn't hardly meet the system requirements ...the rants about Vista could be said about any operating system on subpar hardware...
I've been using Vista since its release and have had 0 (ZERO) issues with it other than a few driver issues that are not even Microsoft's fault. Stop blaming Microsoft for other people's problems. Do your homework before you go blaming Vista for problems that hardware manufactures caused...Microsoft want's to avoid that with windows 7 and that's why they already rolled out the 'guidelines' for manufactures drivers and software...
If you did your homework you'd find that Vista is actually faster (it may be marginal, but it is still faster) than XP when it comes to a lot of stuff...including gaming...so there goes the argument of 'Vista is slow', right out the window...
PPU would be awesome for cones and other obsticles...and that's about it...maybe the flags...
Multi-Core would give a larger benefit over a PPU in LFS. PPU is just too restrictive as far as what can be done...it will thrive in other generes of games but I don't think it ever will take off in the simulation market...
You could say that about CPU processing power in general, but someone will always find a need to push beyond what current systems are capable of.
Atleast now, developers have a 100% free SDK that they can use in their titles (apparently Havok is also releasing its physics tools for free to developers of non-commercial titles). More games utilising these tools should quickly lead to a situation where the hardware that supports it becomes standardised and mainstream. Of course that's likely to take a little while yet. But it's interesting to think about.
Even if Physx is exclusively used in a more cosmetic way, then that should still be fun. Imagine smashing into a tire stack, and as random tires bounce away, they realistically compress slightly as they hit the ground. That would be better than what we have now.
PhysX would be useful if you wanted to simulate every component in the car down to the nuts and bolts, otherwise the CPU is doing a fine job as the game is now. Maybe one day multithreading will be useful to the game but it's a long way off from the place where hardware physics would useful.
completely wrong... every area of lfs could benefit massively from vector or stream processing
starting with the tyres which could be done more accurately with a fem model... chassis flex a la rigs of rods done properly with a parallel core thats suitable for the task and maybe at some point in the future even aero
My problem with Vista is that horizontal span feature was removed (allows to stretch desktop on two monitors, and softwares will believe you have only one very large monitor and will display on both screens).
You can not play anymore on two monitors with Vista...unless you buy a matrox box...which is stupid because it was working very well with XP.
If you google around you can find more comparisons...its more like 50% of the time...if you compare all the tests XP and Vista end up being about the same...there isn't a single test where vista is spanked by xp...
You have NEVER had to copy files over a lan then rotf.
There are well know and WELL reported issues with Vista file manipulation over a LAN and this has STILL not been resolved in SP1.
How the hell does MS expect companies to move to Vista when it will lock up explorer and climb up its own arse just trying to even look at a LAN based directory with more than a handful of files in it.
Please do not try to make out this issue does not exist. It does and if you want proof fly down to NZ and I will demo it for you or just google a little.
My XP box can read/write to the same LAN dir easy, vista locks the shell but 1 in 20 I might get a dir listing. This is what happens when people think its wise to clean room the Network stack again from scratch. It will take YEARS to get right again as is normal with complex sysstems.
Probably one of the many reasons for Intels COMPLETE rejection of Vista.
Google "vista lan hang" and you over 1 million results, do the same with "vista hand" and its over 5 million. That is a lot of results considering you are telling me Vista is 100% perfect lol.
As I said Vista has serious issues. Your use of LIMITED parts of the OS might not have shown them but do not pretend they are not there
Oh come on now. You're among friends. Cut back on the hyperbole. Intel did a normal cost/benefit analysis and found there were no benefits from upgrading at this time. Typical business desicion. (How long did Win2k stay around in the business world?) Parts of Intel are still using it if it makes sense for them. You make it sound like Intel completely rejected it on its techical merits, not allowing it on the premises, which is pushing the truth quite a bit.
That said, Vista certainly has its issues. The most annoying one I've encountered is the indexer hogging the disk even though it's set to a low IO priority. This means that if I've landed a lot of new files in my user directory (something I often do when updating some codebase from a Subversion repository for instance), I can't play a video off the same disk until the indexer is done indexing the new files. Hugely annoying, and something I can't fathom how got past QA; much like the network/media playback issue.