Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Stainless]

CryEngine
*Technologically perfectly up to the task of making a flight sim as far as world size is concerned
*free SDK
*If game is released for free, no additional licensing required; If you intend to charge money for the finished product, some sort of "attractive licensing options" are available but I can't seem to find them...
*CryTek is in some financial problems and future support may be uncertain

Where have you seen this?  All I can see is the pay version on steam.

[SAS~Storebror]

http://cryengine.com/get-cryengine/free-sdk-faq

[Stainless]

Ahhh yes looks like they didn't update the web site for that bit.

Try clicking on the links and you end up back at Steam and have to pay for it.

[SAS~Malone]

as far as maps go, i would dearly love to see a much improved 'geometry' when creating maps - currently it works off a sort of grid that only allows for limited changes within the grid 'shapes'
i would love to see it more 'free-form' to allow for better elevation changes on a more detailed scale than what is currently possible.
an example of this would be to have better riverbank details, as oppose to the current setup, where the land adjoining rivers/coastal area is completely at the same level of the water.
this sort of improvement would allow for much, much nicer looking maps, if possible

[Herra Tohtori]

Ok I've had some more time to think about it. This is what I am going to suggest.

Task	Workers	Timescale
Select Engine	Me	3 weeks
Writer mesh importer	Me	2 weeks
Convert HIM files	Modders	Open ended
Write map converter	Me	1 month
Convert maps	Mappers	Open ended
Write flight model	Me	3 weeks
Convert FM files	Java coders	Open ended

This will give us three key points in development.

• Model viewer
• Map explorer
• Simple flight sim

The model viewer will let you see what your mods look like in a modern engine with sophisticated lighting etc. This will also show you which bits of the models will have to be reworked. (probably the first will be creating a much more detailed pilot mesh.

The map explorer will let you see what the terrain and buildings would look like. Again this will produce a lot of "that looks crap, let's rework it"

The last will be the first time something actually looks like a flight sim. No combat, no weapons, no missions, but a good flight model, good aircraft, good cockpits, good terrain, etc.

Before I commit to doing this though, can someone who knows this site well put up a survey so we can find out how many people would be willing to pledge support to each section.

I would seriously suggest doing a proof of concept first: Instead of converting lots of content right away, concentrate on getting two different aircraft (or just one would work too) of roughly similar time period into the game and get the game itself to a level where you can get some basic gameplay either against AI or other players. That means a stage where you have working flight model, damage model, ballistics physics, and you can start testing whether for example the flight dynamics code is working to acceptable level, or if you need to make some additions or modifications.

Both SP and MP have their own difficulty - multiplayer means you have to get the netcode and interface working so people can host and join fights, while bots means you need to write AI that can fly aircraft... personally I think multiplayer is probably easier especially if the game engine has basic structure in place for that.

Ahhh yes looks like they didn't update the web site for that bit.

Try clicking on the links and you end up back at Steam and have to pay for it.

I'm not sure what link got you back to Steam, but CryEngine SDK is not Crysis/2/3/Warhead or any other game. It's a game engine. It also has nothing to do with Steam. You get all the things from here: http://www.crydev.net/dm_eds/download_detail.php?id=4

[Draken]

Interesting !
I would like a realistic rendering of metallic surface , so that bare metal polished planes shine like a mirror , please .

[benitomuso]

I agree with Herra. Before having something like two planes doing decently what we expect, the rest of the stuff is only more confussion.

But we have a recent ongoing example of that: Battle of Stalingrad. How easy was for skilled people having two planes doing all what they wanted? MOnths and months and months of dedication, investment and refining.

It is not an easy task, but I admire your conviction Stainless.

Regards,
                             Pablo

[SAS~HolyGrail]

Ahhh yes looks like they didn't update the web site for that bit.

Try clicking on the links and you end up back at Steam and have to pay for it.

It works 

http://www.crydev.net/dm_eds/download_detail.php?id=4

File is: CRYENGINE_Build_PC_v3_5_8_2310_freesdk.zip

Excellent download speed , it took me just 8 minutes for 1.9 GB

Version: 3.5.8
Last Updated: 12.03.2014, 13:33
Size: 1939.36 MB
Total Downloads: 7640316

~S!~

[kapitansky]

   

I think this video helps?

[Stainless]

That video is a perfect example of the problems I face.

Let's think about a typical map, say the western front map. It's about 4000 pixels wide. That represents about a 1000 miles of real world terrain giving a quarter of a mile per pixel.

Not enough resolution for anything.

Now do you want to go through the above process and generate a terrain mesh for each pixel? That's about 4,000,000 heightmaps that have to be generated by hand?

No didn't think so, we have to automate it.

[Herra Tohtori]

I would rather start with something less ambitious as far as terrain goes. The most important part would be to get the flight simulation working - to see if your aerodynamics implementation is producing acceptable results, or if the general idea is simply too ambitious to achieve. This should ideally be done with stock asset locations - possibly with an airfield modified onto the ocean for take-off and landing tests - because they offer just as good an environment for testing the fundamental gameplay potential as a more historical, complete map would.

Once you know you have a solid base for the mechanics of flight down, you can get people to work on things like more aircraft and better locations.


As for the height map thing - you don't need to create a height map for every tile, no one does that. I would recommend using a basic high resolution height map (or, alternatively, something like SRTM* data directly without conversion to a texture) and adding details on generic tiles with fractal modifiers of some kind (the biggest problem you have is making sure each tile blends well to the adjacent tiles). That way, you would only need to craft hand-made height maps for special tiles, where you want to include particular elevation formations at higher detail.


* Shuttle Radar Topography Mission. Offers elevation data for the entire globe in 500 metres/pixel resolution. With suitable interpolation, that will already provide quite a detailed terrain mesh. If you want to get fancy, NASA also has bathymetry (ocean floor depth charts) available somewhere on their site...

[Stainless]

And that is exactly what I want to do, but you can't in a lot of game engines.

For a lot of game engines you have an editor. If it's not generated in the editor, it's not in the game.

Hence the reguirement to hand generate 4 million terrain patches for a typical map.

That's why I put down such a long period of time just for working out what engine to use. Looking at a video and saying "wow that's amazing, let's use that" is about as useful as a handbrake on a canoe.

I have a copy of Unigine from a previous job and had a quick look at that. The renderer is lovely, just what we need, but it only supports it's own mesh format and the format is sealed. They supply a converter but that can only be run manually. So I would have to convert every single MSH file in the game by hand. Not going to happen.

I have compiled up JBSim as my default flight modeler, it's probably the best currently available and I should use that in the intial stages.