DCS Features Wish List

[vrv]

I think that the C-130 would be the most superior aircraft for the next DCS. Thus I am going to start a thread to state my opinion when there are billions of posts and threads about the same subject. And I said good day sir.

[PeterP]

Tornado!! :D

 

Edit:- If you irritated of my post: I posted this before the 'just another' "next DCS Aircraft" thread has been moved into this one...;) )

Edited January 20, 2012 by PeterP

[Panthro]

I'd love to see output compatible with HID

[213]

better multicore support

 

started this thread both to suggest opening cores 3 and 4, but also to ask how come you guys don't have this yet?

[EtherealN]

213,

The reasons for this have been gone through quite a few times on this forum, actually. :)

 

1st: this is not a simple compiler flag you flip. This is a major redesign of the entire software at design level. Typically something you want to do before you even write a single line of code.

 

2nd: ED is moving towards this, but at a pace that allows it to still publish products as opposed to spending the next 5-8 years re-writing everything without a single dollar coming in. The latter approach is known as "how to go from profitability to bankruptcy in 5 years or less". ;)

 

3rd: multithreading is implemented as and when core modules of the codebase are exchanged. Like when the sound engine was replaced and given it's own thread. That is, as mentioned, the way to do it if want to stay in business.

[falcon_120]

I really believe this is not something ED wouldn't wish,but rather a to do list for the future.

Multicore implementation is not easy to obtain for a detailed simulator,with lots of calculation going by and the need from cores of getting along with each other.

Hopefully we'll get it anytime soon.

[sobek]

[...]

 

Please stick to existing threads when practical. Also, the search function is your friend, there's hours of material to read for that particular topic.

[213]

213,

The reasons for this have been gone through quite a few times on this forum, actually. :)

 

1st: this is not a simple compiler flag you flip. This is a major redesign of the entire software at design level. Typically something you want to do before you even write a single line of code.

 

2nd: ED is moving towards this, but at a pace that allows it to still publish products as opposed to spending the next 5-8 years re-writing everything without a single dollar coming in. The latter approach is known as "how to go from profitability to bankruptcy in 5 years or less". ;)

 

3rd: multithreading is implemented as and when core modules of the codebase are exchanged. Like when the sound engine was replaced and given it's own thread. That is, as mentioned, the way to do it if want to stay in business.

 

ok.

 

[Nate--IRL--]

ok.

 

 

Now that you've explained it so clearly I think ED should hire you to do the job. It looks so easy. :)

 

Nate

[joey45]

^^ Damn you beat me to it..

 

edit: still need to spread some love to rep..

Edited May 7, 2012 by joey45

[EtherealN]

213, unfortunately it seems like you did not understand what I just told you, so I'll try again with an analogy:

 

You cannot take a '91 Volvo 240, put a 1000hp engine in it, and think you just got a sports car. Once you've done this, you'll notice that your transmission dies. So now you put in a new transmission. Now your linkages are off-kilter, and you have to replace the boggies. Then you notice that your breaking system is completely inadequate and you replace that. Then you notice that handling is crap and you end up having to completely re-do the chassis to alter CG etcetera.

 

...at the end, what you have is NOT a Volvo 240. You have a complex one-off made up of random slapdash fixes.

 

You'd spend less of both time and money designing a sports car from scratch. However, designing a sports car from scratch means that until it's done, you don't have a car at all. Now what are you going to do to get yourself to work? Answer is, you won't. And if you don't get to your workplace, odds are they won't pay you. And if they don't pay you, you won't be able to pay your bills. What happens when you don't pay your bills?

 

Now go back to what I wrote to you previously. ;)

 

Seriously, programming is complex even before multithreading starts happening. Read up on concepts such as data dependecies etcetera. This is not Minesweeper we are talking about here. ;)

 

An alternative is to do all of this in the background while releasing new products - but said new products would be fixed in features. There's be no new sound engine, no new graphics effects, no nothing. It would be a fixed codebase for however much time it takes to completely rebuild everything. Not even iD software, which has tonnes of money and a major publisher at their back, thinks this is a good idea. They did end up going this way for Rage, and Carmack has labeled this one of the biggest mistakes in his professional career.

 

Now imagine what that mistake would do to a small developer that has NO big publisher at their back...

Edited May 7, 2012 by EtherealN

[mmaruda]

Still one must acknowledge that the engine we now have has serious performance issues. It may be hard, buy some news from ED that they are trying to work it out would at least come people down, since now we don't even know if they intend to work on multicore support or are just waiting for hardware that would handle the current state of affairs.

 

On the other hand, I think it's the GPU that has a lot to do with performance.

I had a Pehenom II 1090T no OC and a Radeon HD 6870. Framerates were terrible. So I went for the i5 2500K and OC'd it to 4,4Gh. There was an improvement, but still mostly around 20 FPS flying low, sometimes below that. For a game that is said to be so CPU bound something did not seem right. This week I upgraded to a GTX 680 (which was a fortune and I can hardly live with myself now) and framerates do not go below 30 even in extreme cases. Hell, I mostly get over 40 FPS in the cockpit and from 60 to 100 outside.

 

It seems to me, that the issue lies with VRAM and the fact that the eye-candy such as shadows, water and other stuff is not done using the usual tricks game developers do. It's like the game does not make use of shaders at all. Seems like everything is run equally by the CPU and GPU and if any of these two is not high-end, FPS will drop.

 

The reasonable thing to do IMHO would be to get the AI on another core and get these DirectX features to handle the eye candy. I'm not a coder, so these just assumptions, but I feel the pain of those with lesser hardware. The least ED could do is update the system requirements on the DCS page, as the ones stated as recommended are more or less minimum to play this.

[EtherealN]

since now we don't even know if they intend to work on multicore support or are just waiting for hardware that would handle the current state of affairs.

 

Actually, yes we do know. ED is working on it, but it is working on it in the fasion and pace that allows ED to also ship products, get revenues, and thus stay in business. Simple example: why do you think the new sound engine got it's own thread? ;)

 

Regarding GPU's, yes, from A-10C and onwards they also have a considerable impact on performance since a lot of new effects have been introduced that are very computation-intensive for the graphics card. HDR, TSSAA etcetera don't come "cheap".

 

You don't need a massive beast of a GPU though. I have 50-ish FPS on my 560Ti, running windowed, with most effects on and high (but water low). This is when not using my overclock.

 

As you have noticed though, vRAM definitely has a major impact. In fact, I'd rate it as even more important than the GPU itself - and not only the amount of vRAM on the card, even more important is the speed of it and the width of the memory bus. An often overlooked metric in graphics card performance. :)

 

In the specific case of the 680, it actually has a (comparatively) narrow memory bus, but counters this through running it at a higher speed.

[213]

if it's that much work, they better prioritize :)

 

no dcs or edge until multicore

[sobek]

no dcs or edge until multicore

 

You conveniently factored out that there is no more ED unless they periodically generate income through releasing products. Stopping everything until they have a fully multi-threaded architecture would be the death sentence for ED. I'm not sure that you really understand the ramifications of your proposal.

[upupandaway]

I know The canopy rain effect has been brought forward in a similar thread at least once before - how about a canopy fogging effect? Has that not been suggested before?

 

We´ve had the dynamic weather for a while now - it´s a real nice feature and certainly a step twoards creating realistic and believable weather effects for the entire DCS environment. The aircraft modules all have to hold up to a very high standard with regard to their systems and flight behaviour. I think it would add tremeandously to the whole DCS package if the weather also had the same beautiful little (visual) details like rain/fog/ice on the canopy, localized fog, rain columns, cloud shadows at night etc.

There are many (technical) additions in other areas especially now, but still I wonder if EDGE is going to implemt any such feature... (or at another time in the future - after all, this is no "real" secret like the next jet, is it?)

One may dismiss this as mere details that have been mentioned in separate threads, but on balance it adds a lot of realism and immersion.

 

Ahoy!

[Wichid]

I think that anything starting with the word 'dynamic' should be broken off and put into a new module called "DCS: Dynamic Server Monster!!".

 

Obviously it would include the long awaited dedicated server. It could manage dynamic campaigns with weather systems, resources and units that are carried over between each mission. If the weather gets too bad some air fields may be closed and the poor combined arms controller (once that is released) has only their ground units and artillery for support until the weather clears.

 

DCS: Dynamic Server Monster could run on multiplayer servers for days on end providing endless entertainment for anyone with DCS: World installed.

 

ED would then sell so many licenses of the air frames and maps that they could hire Gabe Newell to be their tea lady. But, they wouldn't because they also want to play Half Life 2: Episode 3.

[upupandaway]

That sure sounds like a tempting idea...

[Silver_Dragon]

A good addition.....

 

- Manage ammo and fuel of vehicles and ships via lua and editor to select adequate mixed ammo cargo to specific missions. Ej. Artillery fire HE ammunition over a fixed target, Racime ammunition over a tank column, and scattered mines over a road to block them.

 

For the warehouse.

- Make afloat warehouse ships: AK, AO, AOE, etc, capable to repost and resupply ships on sea only if that ship has integrate on formation or to a limited distance.

- Make ports the warehouse ability to give ammo and fuel.

- give air/ helicopters, trains and road truck vehicles the ability to move ammunition and fuel into warehouses / farps / airports / ports.

Edited May 11, 2012 by Silver_Dragon

[joey45]

^^ why not merge Simcity with DCS...;)

[upupandaway]

Add pedestrians to civ traffic...

[Silver_Dragon]

Engineer units:

 

- Mine dispersed units, by helos, airplanes, vehicles or deploy by maps to disrupt of fixed advanced units.

- Engineer repair teams to repair damage runways and warehouses.

- Pontoon and river crossing units, to building combat bridges on damage structures, pontoons over rivers and ferry vehicles to move units on course waves.

 

- Make ability of amphibious at some vehicles capable: BMPs, BRMDs, BTR, etc.

- A better logic, if a bridge has destroyed, convoy can exit to the roars and move to the nearest crossing point or ferry units to continue travel.

[luza]

I think that anything starting with the word 'dynamic' should be broken off and put into a new module called "DCS: Dynamic Server Monster!!".

 

Obviously it would include the long awaited dedicated server. It could manage dynamic campaigns with weather systems, resources and units that are carried over between each mission. If the weather gets too bad some air fields may be closed and the poor combined arms controller (once that is released) has only their ground units and artillery for support until the weather clears.

 

DCS: Dynamic Server Monster could run on multiplayer servers for days on end providing endless entertainment for anyone with DCS: World installed.

 

ED would then sell so many licenses of the air frames and maps that they could hire Gabe Newell to be their tea lady. But, they wouldn't because they also want to play Half Life 2: Episode 3.

 

Dunno if it's meant as troll. but this is a seriously cool idea ! :thumbup:

[KillarZ]

my wish list

 

1.) An outerra like map/engine optimized to perform well in DCS

 

2.) multi cpu and gpu support

 

3.) better sound engine. I havn't tried the p-51 but on the vids I've seen it does sound good. I'm just hoping future jet aircraft will sound as good. like this:

 

4.) An integrated multiplayer lobby system

 

5.) vapor effects based on the weather

Edited May 12, 2012 by KillarZ

[Cypher]

Sorry if this has been said before. But I would really like to see more realistic ground vehicle formations.