Desired features

If you could add one feature (functionality/AI/otherwise) to the base game that cannot be accomplished via the KS mod, what would it be and why? Just out of curiosity.

1. In single player to ask for support from another AI commander and actually have a chance at getting it.

2. Automatic transfer to a voice channel in HITS when you ride within 50 yards of another player

That's 2 features Pat.

I definitely agree with Pat's second one; it seems restrictive to only communicate by courier even when two players are right next to each other.

One of my personal desires would be a "withdraw" stance similar in concept to the attack/all-out-attack stances. This would direct any commanders given the stance to conduct a gradual withdrawal of forces from combat, potentially by "leapfrogging" units to provide cover and prevent a general rout.

Draw distance. There's already some impressively large battles that can be seen in the game. Sometimes the way a battle unfolds you can sit on a hill and see nearly all of it. But only nearly. When it reaches that distance to only show the banner bearer instead of the full unit must surely be for performance gains which is understandable, but a division stretched out along some winding road and hills would look that much cooler if all of the sprites were rendered.

Real (or closer to real) physics - that would solve array of problems... line of sight, units spacing, friendly fire...

Having all the code.

That would certainly help, Kevin.
But it doesn't guarantee anything though.
A lot of coding work would still be needed.
And in some cases, possibly even more than not having the code at all (refactoring temptation would be strong)...

The voice request raised by Pat can be perfectly implemented through the (free) Discord Game SDK.
In its simplest form, just create a lobby beforehand, share its id with people joining the game (say copied&pasted into an ini file, or clipboard, or whatever) and let all the AI dlls connect to it at initialization.
Once you got that running (and updating in the main game loop or in a background thread...), when the human player officer is asked to perform is "thinking" in the AI DLL, you can manipulatie the local mute/volume settings to the other users within the Discord lobby, according to the simulation demands (like testing distances in yards from corresponding officer units, etc.).
The mapping officer unit <-> Discord lobby user id can be as simple as strcmp'ing the usernames. Or, by assuming players will log into the MP game screen with their username set to the Discord userid.
If one wants, it can be done.

About drawing distances...
I've got an experimental unit sprites renderer working lately. It does its job quite efficiently by drawing all sprites with a single draw call.
I had to batch everything(when loading the battle graphics)  into an huge volume (3d) texture atlas for that.
But it can't work with KS because the mod employs packs that are bigger (4096x4096) than the maximum extent supported by any D3D9 device for volume textures (2048x2048x2048).
Reprocessing all the sprites packs is another example of a rather time-consuming work required where having or not the source code doesn't really change the situation...

From my experience, willingness, teamwork and free time remain by far the key factors when it comes to hobby like SOW or any other deep-modding work to similar old games.

Hope you're all doing well, guys!
Ciao.

What are your thoughts on the (rumored to be in the works) new game from NSD?

I haven't been to the NSD forums in a long time. What game is this? Possibly?

I'm just as in the dark as you are...

DumpTruck wrote:Draw distance. There's already some impressively large battles that can be seen in the game. Sometimes the way a battle unfolds you can sit on a hill and see nearly all of it. But only nearly. When it reaches that distance to only show the banner bearer instead of the full unit must surely be for performance gains which is understandable, but a division stretched out along some winding road and hills would look that much cooler if all of the sprites were rendered.

Some screens of cavalry and infantry fading at distance.
Just using stupid fog plus old style cross fading / dithering.


Actually the only limiting factor I see is depth buffer precision.
This should certainly happen if you set the far clip plane too far way from the near one.
It can be tweaked to a certain degree, but a solution like this
https://developer.nvidia.com/content/depth-precision-visualized
would probably work better.

Uff, if only we could propose together a project of expansion to Norb...
Most likely it wouldn't get a positive reply.
But if a chance still exists, it might be easier to catch it together and with realistic goals laid down first.
What I'm sure nevertheless, is that we could still have some fun developing it.
AI prinz and us plebs. Lot of fun, definitely.

Question from the Prinz.
Very nice. How much do you decrease the graphics calls by batching all the sprites? Repackaging the sprite sheets is not a huge problem as I have all the sprites saved as single graphics. If the sprites and map terrain from SOWWL could be used with a game built on the Unity engine, that might be the easiest way to make improvements.

r59, please talk to Kevin about that dithering/fogging effect. I really like it. It might only be useful on certain maps as some of them already have some fog and colouration effects. I assume however that the max draw distance is fixed and final.

Another issue I see is with the stock Waterloo sprites where the artist made up different hats or facings or jackets that fit over a base area of another sprite. The problem is that for some daft reason the draw distances on the body parts are different to the base sprite (duh!) so at distance you get the disconcerting and almost ghoulish effect of a group of disembodied torsos or heads in view only. The QC on these products was really poor.

One hopes its possible to alter the draw distance for the base sprites so that they become visible along with the body parts.

Can we get the draw distance on ALL sprites to be the same as the draw distance on the sprite holding the unit's flag?

Thanks Digby.
More screens about that fading/fog feature (of course parameters and the "pixelation matrix" will need some better calibration by artists):
https://imgur.com/a/jWXH6DB

Prinz, just a single call. 5 or 5000 sprites doesn't change. It's 1 call instead of 10 or 10000 respectively (and I'm not considering all the per-triangle redudant settexture/setvs/setps/setrs calls involved as well).
For units there's still all the alpha polys's sorting running behind the hood at runtime. I might probably get rid off at a later stage or limit its impact exploiting temporal coherence.
And I need to profile the uploading to GPU of the giant volume texture. Details.
You are likely correct about your assumptions about porting the assets and main functions to Unity.
Sadly, I'm a noob as anybody else here could be with that engine and the C# language, so it'd take a while for sure...

Now, I've got these 3 or 4 systems, not in a 100% final implementation (especially unit's) state but quite acceptably tested, plus an handful of bugfixes, that could easily find their spot in the main codebase imho. Plus the Discord integration, if needed (I spent the last week on its SDK APIs).
I think, whatever opinion one may have about their utility, it'd be a shame to trash everything OR start fighting against Steam's protections to release it in a 100% safe manner OR discorauging users to buy it on Steam OR similar bullbusting.
You get the point.
The game is going to stay in Matrix's stores forever too.
But still I understand it's quite irrational asking NSD to let this happen in 2020.
Possibly with another title development going on...
I'm hesitating since a couple of weeks.
What has been your relationship with Norb throughout these years?
How do you think he could react to such a request or to a more "upscaled plan" like allowing a B-team to have a try on a Borodino expansion, for example?
As long as we could have some fun making it with a reasonable effort, I really wouldn't care about see it based on cutting edge tech, selling well or not.

I'm pretty sure NSD is out of the game business. Apart from Little Powell and perhaps Jim, I'm not sure there is an NSD anymore. The poor reception SOWWL received soured them on game development I fear. Napoleonic devotees are a very particular lot and the shortcomings of the game were never going to be accepted or overlooked.

I'm a little pessimistic that Norb would be interested in turning over the rest of the code in order that it be improved and used for another game. I've offered to fix some of the issues with the code so he could make another game if he chose to do so, but he was not interested at that time.

From a practical standpoint, I think it would be more productive to explore the Unity engine idea. C# isn't that different than C++. Also C++ code can be wrapped in a dll and imported into a Unity project as I recall. Things like line of sight, real physics, and pathfinding can be added from scripts that have already been written. They cost a few dollars, (Euros), but would save an immense amount of time. MP mechanics is already a part of the engine. The important piece is whether the existing resources that already exit for SOWWL, the sprites, maps, etc. can be used with that engine.

If you guys make a go of this you have my playtesting time and some $$$. I wonder if Gunship, the NSD sprite artist would be interested? His sprites are much nicer than the ones we have right now. Pom might also be good for scenery items and flags.

i can't say that I understand much of this tech speak but, like Martin, I would be willing to devote spare time and $$$ to trying to develop the game further.

I will help. Time and $$. Plus I bring the added benefit of doing really dumb things that you would want to prevent a player from doing. Vwry handy in development and test

I think an ideal addition would be increased complexity of orders; that is, being able to order an AI commander to attack up to a certain location, then hold the objective, as well as contingencies in the case of enemy contact, all in one order, rather than issuing each of these as the situation changes.