[icecoldduke]

Trooper Dan, on 20 July 2016 - 06:09 PM, said:

Even in the base game, there's lots of examples of sprites that spawn lots of other sprites (jibs, smoke, scraps, etc.) All it takes is a lot of this happening at once (e.g. caused by a chain reaction of explosions, very common in the base game) and there is your 100's of new sprites.

For that I plan on doing a server to client RMI(remote method invocation) that tells the client to spawn jobs/smoke/scraps/explosions/etc. When the client receives the RMI, it would create the needed sprites inside of the client side sprite array; the client is solely responsible for those sprites. Problem solved. .

// Replicated
EXTERN spritetype *sprite;
EXTERN tspritetype *tsprite;

// Not replicated
EXTERN spritetype *clientsprite;
EXTERN tspritetype *clienttsprite;

[TerminX]

icecoldduke, on 20 July 2016 - 06:14 PM, said:

For that I plan on doing a server to client RMI(remote method invocation) that tells the client to spawn jobs/smoke/scraps/explosions/etc. When the client receives the RMI, it would create the needed sprites inside of the client side sprite array; the client is solely responsible for those sprites. Problem solved. .

Yes, that's how the implementation in EDuke32 works. You don't need to have them in a completely different array--that's a ridiculous level of added complication. You just pre-allocate a bunch of sprites into a reserved statnum and pull from that pool when the client needs to spawn something. The server allocates a pool of the same size but never touches them. That gives you the required separation without much added complexity, and ensures that the sprites the server does care about have the same IDs on both ends.

[icecoldduke]

TerminX, on 20 July 2016 - 06:28 PM, said:

Yes, that's how the implementation in EDuke32 works. You don't need to have them in a completely different array--that's a ridiculous level of added complication. You just pre-allocate a bunch of sprites into a reserved statnum and pull from that pool when the client needs to spawn something. The server allocates a pool of the same size but never touches them. That gives you the required separation without much added complexity, and ensures that the sprites the server does care about have the same IDs on both ends.

That would work too. Either way its a very solvable problem. . Back to the other issue. The server always needs to be authoritative. In my model I plan to keep the server always being authoritative. One way to run the clients, is to run the game simulation(like monster logic) on the clients, and the server syncs up the clients periodically. The other way is to take two snapshots and predict the next few frames until a server updates comes. Either method can result in a snap if there is latency between the client and server. I'm choosing the latter(taking two snapshots and predicting whats next). The only other methods I know of, are client authoritative models and I would never make a game that uses any of those methods. Hackers provide a even worse game play experience, then snapping

[Mblackwell]

icecoldduke, on 20 July 2016 - 05:56 PM, said:

If you have a effect that would spawn 100's of new sprites, then you would be correct you would need a different approach.

If a group of players are in E1L2, firing RPG rounds, and someone triggers the building explosion... you've got a bandwidth problem!

icecoldduke, on 20 July 2016 - 06:14 PM, said:

For that I plan on doing a server to client RMI(remote method invocation) that tells the client to spawn jobs/smoke/scraps/explosions/etc. When the client receives the RMI, it would create the needed sprites inside of the client side sprite array; the client is solely responsible for those sprites. Problem solved. .

// Replicated
EXTERN spritetype *sprite;
EXTERN tspritetype *tsprite;

// Not replicated
EXTERN spritetype *clientsprite;
EXTERN tspritetype *clienttsprite;

In EDuke32 this could be any sprite btw. I hope you have a pretty big etc.

[icecoldduke]

Mblackwell, on 20 July 2016 - 06:40 PM, said:

In EDuke32 this could be any sprite btw. I hope you have a pretty big etc.

The fact is not all sprites need to get replicated. There needs to be a system in place to tell the engine what sprites need to get replicate and which ones are client only. Right now I plan on manually specifying which sprites need to get replicated the rest are client side only(if they get loaded in during level load they get synced once, and that's it).

Also were is a good place to upload the SW notes too? How many people here are actually interested in Shadow Warrior modding? And if not what is your reason for not wanting to mod SW?

[TerminX]

People don't want to mod SW because it's not Duke3D. Taken individually, most aspects of SW are technically superior to their Duke3D counterparts, but as a whole it falls far short of Duke3D's charm.

[Mark]

Unless something has changed, for me its the fact that most geometry can be produced in Mapster32 but then you have to go back to an ( what seems ancient ) older version of Build to do SW specific effects. That small inconvenience has stopped me. I guess the quality of Mapster has spoiled me.

EDIT: And I seem to always be involved in mods and TCs and so don't produce vanilla maps. That enters into it also. No con coding allowed IIRC

[MusicallyInspired]

I'm not at all interested in SW personally. I've owned it for years and years and have yet to really play it beyond the first level or two. Haven't even finished the demo. I'm all for a modern working source port, of course. Makes life easier for everyone. But I do hope Duke remains the priority.

[Micky C]

Shadow Warrior grew on me over the years. The main factor for this is probably that the levels are way, way too maze-like, with it being not very clear where to go at times. As you become more familiar with the levels this becomes less of a problem. Plus it's almost too hard in places (unevenly so). If these things were smoothed out it'd be a much better game IMO.

Not to turn this into a SW gameplay discussion thread. Just I can see why ICD likes it.

[icecoldduke]

Here is what I purpose. In PolymerNG, Shadow Warrior and Duke3D are merged into the same codebase. The tech I'm developing is game independent, it really applies to either game. The only delta between the interface is whether I'm using SE's or ST1's. As I'm developing tech, I need to create levels that demo proper usage of the tech. These levels I call "zoo" maps. Frankly I have a more fun time developing these tech demos in Shadow Warrior. There are many reasons for this, one big one is I don't have to develop using con script. I think it's a shame that there are so many people in the world that are scared of writing c/c++ but are alright with writing script code. I'm not talking about any one person in particular, it's just a general statement that applies to a lot of people. I think if the Blood source code ever got released, this would be less of a issue with people on this forum(since blood didn't have a scripting language).

So Duke3D is the top priority for me. That being said, I would like to continue to make these "zoo" assets in Shadow Warrior. I figure even for people that don't want to develop Shadow Warrior mods; this will show off proper usage of the technology. With Shadow Warrior I don't dread having to go in and add say lights to different monsters for say weapons fire; in Duke3D I dread making con code changes, because I hate the scripting language. In Duke3D I will have that e1l1 map, and I will ensure light spawning will get exposed to the con language. I will also work with mod guys directly on getting PolymerNG integrated into there mods. What I'm purposing only changes what technology showcase pieces come from me. Does this sound acceptable?

[FistMarine]

edit

[TON]

Months ago you said that you want to add physics to PolymerNG in the future

What type of physics?
And what code language for the physics? CON code?

[icecoldduke]

I will be adding PhysX to PolymerNG. In Shadow Warrior getting physics integrated will be easier since the game logic is all c. For con however things will be trickier, but its doable. Ill have more on this later.

[Moggimus]

I'm a big fan of Shadow Warrior so I'm definitely glad you're working on that as well. Any chance you might make an updated map editor?

[icecoldduke]

moggimus, on 21 July 2016 - 06:24 AM, said:

I'm a big fan of Shadow Warrior so I'm definitely glad you're working on that as well. Any chance you might make an updated map editor?

I was planning on taking some time this weekend and get the editor up to date. I know mapster apparently fucks up SW maps. Does anyone have any specific info on this? Like what modifications you can do safely in mapster and which ones can fuck up the data?

[Steveeeie]

Is this a shadow warrior project now?

[TON]

Steveeeie, on 21 July 2016 - 09:50 AM, said:

Is this a shadow warrior project now?

icecoldduke, on 21 July 2016 - 02:26 AM, said:

Here is what I purpose. In PolymerNG, Shadow Warrior and Duke3D are merged into the same codebase. The tech I'm developing is game independent, it really applies to either game.

So Duke3D is the top priority for me.

[TerminX]

icecoldduke, on 21 July 2016 - 04:54 AM, said:

I will be adding PhysX to PolymerNG. In Shadow Warrior getting physics integrated will be easier since the game logic is all c. For con however things will be trickier, but its doable. Ill have more on this later.

I'm not aware of a PhysX SDK released under a BSD or MIT-style license.

[icecoldduke]

TerminX, on 21 July 2016 - 12:09 PM, said:

I'm not aware of a PhysX SDK released under a BSD or MIT-style license.

When you get a chance, can you check your PM?

[icecoldduke]

So today I received some sad news; I'm being laid off effective today. Which is sad(trying to get out of that "just got laid off" feeling). I will pursue PolymerNG development while I look for work, but obviously I need to find a job, so that is priority . I need to agro my financies and deal with all that scary unemployment stuffs. I'll post updates when I can of PolymerNG, I just ask you guys to be patient.

[Moggimus]

Sorry to hear that, good luck in finding a new job. Obviously job searching takes precedence; do what you gotta do first and worry about the rest later.

[Danukem]

icecoldduke, on 21 July 2016 - 01:40 PM, said:

So today I received some sad news; I'm being laid off effective today. Which is sad(trying to get out of that "just got laid off" feeling). I will pursue PolymerNG development while I look for work, but obviously I need to find a job, so that is priority . I need to agro my financies and deal with all that scary unemployment stuffs. I'll post updates when I can of PolymerNG, I just ask you guys to be patient.

Well that sucks. If you qualify for unemployment checks make sure you get them! Realistically you will only be spending 2-3 hours per day looking for a job, and the rest of the time you can still work on PolymerNG. It will look good on your portfolio, too.

[icecoldduke]

Steveeeie, on 21 July 2016 - 09:50 AM, said:

Is this a shadow warrior project now?

No. As I stated before, the PolymerNG effort is still focused on Duke3D; but the "high quality content" I create for PolymerNG will mostly be in Shadow Warrior, so I don't have to touch the con code.

Trooper Dan, on 21 July 2016 - 05:08 PM, said:

Well that sucks. If you qualify for unemployment checks make sure you get them! Realistically you will only be spending 2-3 hours per day looking for a job, and the rest of the time you can still work on PolymerNG. It will look good on your portfolio, too.

That's the first thing I did was file for unemployment . I got some irons the fire, so tomorrow its time for PolymerNG work .

TerminX, on 21 July 2016 - 12:09 PM, said:

I'm not aware of a PhysX SDK released under a BSD or MIT-style license.

Since PhysX can statically link into the build engine(which is not GPL), it should be fine right? Or am I missing something?

[TerminX]

You're missing something. The licensing situation is a mess. BUILDLIC and GPL aren't compatible licenses, so the only way ports of Duke3D are even legal are because both 3D Realms and Ken Silverman granted a linking exception to allow the code to be used for anything. My understanding is that this means other code can't be included unless it's under a license that is compatible with both BUILDLIC and GPL.

[icecoldduke]

I'm moving back into rendering. I started working on water and just completed the refraction portion of the water pass. It's no where close to complete, just wanted to more or less announce I'm going back into rendering enhancements .

[Hendricks266]

icecoldduke, on 22 July 2016 - 11:41 PM, said:

I'm moving back into rendering.

Upvoted.

[Nokiaman]

So is Redneck Rampage next? hehe
Great work you're doing there. Can't wait to play both Duke3D/SW with this beauty

[Spiker]

Looking good! Tease me more!

[DavoX]

Awesome progress man!

[Kathy]

TerminX, on 21 July 2016 - 10:05 PM, said:

You're missing something. The licensing situation is a mess. BUILDLIC and GPL aren't compatible licenses, so the only way ports of Duke3D are even legal are because both 3D Realms and Ken Silverman granted a linking exception to allow the code to be used for anything. My understanding is that this means other code can't be included unless it's under a license that is compatible with both BUILDLIC and GPL.

Is there a particular reason why Ken never licenced it under GPL? Is it because of him being compensated for every build-game port?