Hardware, software and the future of flight combat sims.

[JG52Uther]

Well if I were them I wouldn't do it.There must be easier ways to make a living, without being attacked everytime you turn the computer on!

[SlipBall]

Originally Posted by JG52Uther

Well if I were them I wouldn't do it.There must be easier ways to make a living, without being attacked everytime you turn the computer on!

Yea no fun...isn't that why Oleg became a recluse, no longer one of the boys, hyperlobby and here. People can only take so much abuse

[Skoshi Tiger]

Damn! I just spent half an hour crafting a beautifully worded response to this topic - distilling my thirty years experiences in computing, software development and flight simming and - my bloody computer turned itself off!!! **&^%$* %%^&&% $#^%%^ - Heres the short version

One of the problems is that although our hardware is still improving and processor power is increasing and following the old moores law, we haven't developed ways of harnessing that power.

All the major CPU and GPU platforms have plateaued in terms of clock rate and that increase in power is being increased in terms of number of processors Cores or texture units or whatever you wan't to call them.

To take advantage of that power the programs need to be multi threaded and contain tasks that can be solved using parallel processing, Graphics cards work good using this approach. Other areas of programming the sims are not so good.

In a recent interview at Sim HQ Albert Zhiltsov, one of lead ROF development people, was talking about the complexity of modern fligh sims and said

I am an optimist. I believe that right now — somewhere in the doorway of a university or an office — a young man leaves with a solid idea to make a project better than Rise of Flight. But, as we said earlier — it will be harder for him to do than it was for us, because the speed of life will increase even more and financial crises will put into question all existing business models again and again. James Cameron dove to the bottom of the Mariana Trench, using his own funds. Perhaps this young man will find an opportunity to test his strength; I wish him luck and will be happy to share my experience with him. I am an optimist.

Hopefully the person Albert was talking about will have the idea about making the program scalable accross miltiple cores (to realise that improvement in performance), be able to create a demonstator as proof of concept and will then be snapped up by one of the bigger development houses rather than out lay the costs from his own funds.

The scary thing is is that any development needs to be spending part of their profits on
research and development or the life of their product will be limited. Oleg knew this and this approach showed it through out the life of the Il2 series up until present. Unfortunately he forgot that they do have to make an actual profit and they couldn't get it together before funds ran out and COD development came to a crashing halt.

My biggest fear is that the new il2 sim will go the other way and will not get that fundimental development to the core of the product and the development will be focused on content. The problem with this approach is that when something does arive that is fundimentally better it will be too late to catch up.

To fund this development a steady stream of income is required. It shows how distructive the whinning, and snide remarks can be to these products.

[MadBlaster]

in theory, one of the benefits of object oriented code is re-usability. not sure how much of old IL-2 they were able to carry forward into CLoD. they also farmed out the trees to SpeedTree and there was a post from Liz Lemon about "borrowing" some code. so, it wasn't all on them.

then, you think what happened to hardware and software over the last seven years? dual core, multi-thread, dx10/11...etc. its all constantly changing. so the developers have to learn this on the fly because the customer wants to use his expensive hardware to the fullest. lol, maybe in few years 580 card goes for $50. anyway, it might be good idea for 1C to just keep fine tuning this code on the down low. eventually (hopefully) there will be better times.

[SlipBall]

They say that the memristor will have the same leap forward for mankind this century, as the resistor did in the last century.
http://www.wired.com/wiredenterprise...hp-memristors/

[startrekmike]

I would say that at this point, the DCS world modules (other than FC3) are pretty much the end all and be all of flight sims in the market.

While I like CloD, it's realism level pales in comparison.

[Skoshi Tiger]

Originally Posted by startrekmike

I would say that at this point, the DCS world modules (other than FC3) are pretty much the end all and be all of flight sims in the market.

While I like CloD, it's realism level pales in comparison.

I've bought a few of the DCS products including the P-51 and they have some excelent systems modeling and features. Great Flight simulator. Can't wait until they release the production version.

That said as a "Combat Flight Simulator" it is lacking in that the P-51 has no contemporary targets, or maps, or adverseraries. These are not critisims, The Developers have never stated that they would have. The P-51 is just as they said it would be.

I see that some people are holding out for a contemporary environment for the P-51. That would be great. But I am not holding my breath. How long have we been waiting for the Nivarda map? It just must be very hard to create maps for DCS World.

[baronWastelan]

Also worth mentioning that the desktop PC is this decade's equivalent of the CD player, inasmuch as the under-30 consumer would rather have nothing to do with them. (yes, I still buy CD's )

For example, in my current employment, my Android phone is serving in the role a laptop would have 10 years ago.

[Igo kyu]

Originally Posted by Skoshi Tiger

Damn! I just spent half an hour crafting a beautifully worded response to this topic - distilling my thirty years experiences in computing, software development and flight simming and - my bloody computer turned itself off!!! **&^%$* %%^&&% $#^%%^ - Heres the short version

One of the problems is that although our hardware is still improving and processor power is increasing and following the old moores law, we haven't developed ways of harnessing that power.

This may be the answer:

http://developers.slashdot.org/story...ores-law-gains

Of course, it may also turn out not to work, but I don't know why it wouldn't.

[Igo kyu]

Originally Posted by MadBlaster

lol, maybe in few years 580 card goes for $50.

Man, I paid £300 (bar the odd penny) for a GeForce 2 Pro/GTS in 2000, don't joke about that, it's not funny.

<edit> It ran IL*2 fine, full details at 1600 x 1200, so there was nothing wrong with the deal, it was just expensive.