Chocolate Doom C/S fork? (Just throwing ideas)

[Doomkid]

Something I've always loved the thought of is a Chocolate Doom fork that implemented an Odamex-like C/S system, and the only engine additions were increased visplane limit, increased resolution and a spectator system... Nothing more. I have absolutely zero programming knowledge (considering taking a course) but, I imagine this would be a lot of work.. But maybe the idea will appeal to some aspiring/well known Doom programmers out there?

Sorry for the slight ramble, but you know what they say, plant the idea seed and all that jazz.

[Edward850]

You just described Odamex, while incidentally mentioning it.

... I just don't know anymore.

[Doomkid]

Odamex has an array of bugs as a result of it's CSDoom heritage.

I love Odamex a great deal, but let's not fool ourselves - It's heavily ZDoom.

(Unless doom2.exe has mapinfo, slopes, deep water, literally all of boom, and I'm just missing something..)

[Edward850]

It's also heavily vanilla compliant, to the point of actual 1.9 demo compatibility. Just because it has slopes, that doesn't magically make it not vanilla comparable. You're already asking for a limit removing port, and Odamex fits that exactly. You don't want Boom features, then don't play Boom maps. It's that simple.

You are basically asking for Odamex but with less features. To which end, what's the point? Other than suggesting you don't know what you're talking about.

[Hekksy]

Actually there was a fork of chocolate doom that added more C/S support. It was discontinued in 2012 or 2013 I think and kindasortamaybenotreally worked. http://www.chocolate-doom.org/wiki/index.php/Obsidian

[Doomkid]

Odamex has an array of bugs as a result of it's CSDoom heritage. Why the heck do I have to repeat this.

Edward850 said:

To which end, what's the point? Other than suggesting you don't know what you're talking about.

Your attitude stinks like rotten meat, and always does.. I mean that not to discredit you as you always know what you're talking about re zDoom stuff, but.. Why so sour?

To be clear, I love Odamex.

Vulture said:

Actually there was a fork of chocolate doom that added more C/S support. It was discontinued in 2012 or 2013 I think and kindasortamaybenotreally worked. http://www.chocolate-doom.org/wiki/index.php/Obsidian

Interesting, thanks man!

[Edward850]

Doomkid said:

Your attitude stinks like rotten meat, and always does.. I mean that not to discredit you as you always know what you're talking about re zDoom stuff, but.. Why so sour?

Yours is worse, discrediting information based on narrow minded assumptions. You're asking people to reinvent the wheel, for what exactly? Odamex already is Doom 1.9 compliant, so what else does it need to have? If you take Chocolate Doom and expand it, you're just going to have Odamex all over again.

[Doomkid]

Edward850 said:

discrediting information based on narrow minded assumptions.

I never did that, in fact I specifically did the opposite in my last post by acknowledging your zdoom/doom engine knowledge. Christ's sake, that doesnt make your crummy attitude suddenly OK.

It is possible to provide people with information without being so snide - Crazy, I know. Your kind of attitude is literally poison to Doom newcomers.

[Edward850]

Posts: 2566

You're not exactly a newcomer.

[dg93]

I personally think Odamex is better than Chocolate Doom. I just gave it a download and it kinda runs like Doom95 to be honest. Is it even a multiplayer port? I really don't know all that much about it.

[Doomkid]

Edward850 said:

You're not exactly a newcomer.

I just meant in general. It's all fine, I'm entirely content to keep using Odamex, it just seems to bit of a point of debate whether or not Odamex is appropriate for a "true tournament-rules OS match", even after the Quakecon event. I personally think it's totally fine, but an alternative that doesn't get a lot of (unjust) hate might please many players. That's all.

Besides, if "another source port does it, don't bother" was the common opinion, we would be missing out on all the cool new little ports popping up. I meant no ill will and certainly didn't intend conflict with this thread.

[duh]

[fraggle]

Edward850 said:

You're not exactly a newcomer.

Neither are you, and you should know better than to act the way you've done in this thread.

So drop the attitude, he has a point. Odamex is not derived from Chocolate Doom in any sense.

[HumanBones]

Instead of continuing to create brand new ports, wouldn't it be a better idea to find some people willing to help ports like Odamex improve? I'm not sure that there is a need to thin the collective doom community out any more than it already is at this point.

[Gez]

Isn't ChocoDoom already using C/S netcode?

[dg93]

HumanBones said:

Instead of continuing to create brand new ports, wouldn't it be a better idea to find some people willing to help ports like Odamex improve?

I would like to see Odamex come up with a ranking system similar to the way ZDaemon has for FFA/Duel players. Example: experience points per frag or per CTF flag capture. I also do like the way Odamex is able to handle high pings unlike ZDaemon. Odamex does have an excellent netcode in my opinion.

[VGA]

HumanBones said:

Instead of continuing to create brand new ports, wouldn't it be a better idea to find some people willing to help ports like Odamex improve? I'm not sure that there is a need to thin the collective doom community out any more than it already is at this point.

I don't think that it's a problem to have many ports, as long as they conform to some standards.

There's the vanilla tier which all ports support, the BOOM tier which almost all ports support, the ZDoom tier for ZDoom+forks and the few rare specialized projects for 3DGE, Doomsday etc.

[Jon]

One reason I've never been happy w/ Odamex is the license. It's Zdoom derived but GPL. I don't understand that.

In terms of ports which are chocolate-like but limit removing, Crispy Doom is worth a look.

What about chocolate's network behaviour would you want changed? mid-game joins, variable player numbers, that kind of thing?

[dew]

What's the point? Even Oda comparatively struggles to attract random newcomers, because its focus is still kinda too narrow. Where are my flashy game mods dawg, etc. You haven't even mentioned features like support for 4+ players, freelook (Heretic style), jumping (Hexen style), finite actor height, console or UNLAGGED. Who's going to play this? Who's going to, uh, pay for the infrastructure?

Developing a new C/S port for maybe 10 people is a waste of time. Obsidian was laughed out cruelly. The competitive scene requires standards "vanilla with c/s" simply cannot provide, not without years of development, and casuals will be put off by the limited offer. EECS was a much brighter beacon of hope to topple the zdoom hegemony, but Ladna scrapped the project before there was even a public beta release. His goal is a C/S version of prboom+ now and while it is an admirable effort, I'm more skeptical than ever.

[Gez]

Jon said:

One reason I've never been happy w/ Odamex is the license. It's Zdoom derived but GPL. I don't understand that.

What isn't there to understand? They took an old version of ZDoom, got Randy Heit's agreement and approval for GPL relicensing, and took out like half of the code in their attempt at making sure everything they kept was either originally from Doom or from Randy's own hand.

[Hekksy]

Jon said:

It's Zdoom derived but GPL. I don't understand that.

Because when Odamex was originally created it was meant to be a free alternative to ZDaemon. However the team wanted to make sure that what happened to ZDaemon never happened to Odamex (ZDaemon being an excellent open source port that the community helped develop that went closed source so the admin team could control it more).

Since the start, Odamex's focus has remained the same: To create a free port that would be the serious competitive Doomer's port of choice. But the scope of the project has changed a little bit. If you told me 5 years ago that Odamex would have ACS in it a few years down the road I wouldn't have believed you. ZDoomisms were originally removed because of Raven license conflicts with the GPL, but they have since been re-added because they were wanted by the community and added value to CTF wads.

Also I just want to mention: Odamex is in need of programming help because the previous leaders have put the project on the back-burner due to personal reasons.

[Quasar]

Vulture said:

Since the start, Odamex's focus has remained the same: To create a free port that would be the serious competitive Doomer's port of choice. But the scope of the project has changed a little bit. If you told me 5 years ago that Odamex would have ACS in it a few years down the road I wouldn't have believed you. ZDoomisms were originally removed because of Raven license conflicts with the GPL, but they have since been re-added because they were wanted by the community and added value to CTF wads.

They were re-added at least in part because it's now possible, thanks to the relicensing of the code. Which I may have had a part in and therefore feel compelled to bring it up ;)

[Doomkid]

Really, the whole reason I want this is for deathmatch demos that can play back in vanilla doom2.exe, and are therefore "legit". Maybe some sort of "demo MP mode" so players can't join in mid game, but if not recording, they can. Again, I'm just tossing ideas around.

Jon said:

What about chocolate's network behaviour would you want changed? mid-game joins, variable player numbers, that kind of thing?

Absolutely, right on the money :)

dew said:

What's the point? Even Oda comparatively struggles to attract random newcomers, because its focus is still kinda too narrow. Where are my flashy game mods dawg, etc.

[...]

His goal is a C/S version of prboom+ now and while it is an admirable effort, I'm more skeptical than ever.

Yeah, that's true. I just feel like there are a few OS or even SP players that would go out and try some MP, but they feel that Odamex is too zDoomish.. It puts it in this unfortunate situation where half of the OS crowd we'd love to draw in is deterred. It's damn annoying.

I really hope the PrBoom C/S thing works out, interested to hear where that goes.

If Odamex found a few new helping hands, that would be amazing.

[Edward850]

Doomkid said:

Really, the whole reason I want this is for deathmatch demos that can play back in vanilla doom2.exe, and are therefore "legit". Maybe some sort of "demo MP mode" so players can't join in mid game, but if not recording, they can. Again, I'm just tossing ideas around.

Not even remotely possible. Vanilla demos have no concept of a player joining, or even a player leaving. Regardless if it's the consoleplayer or not.

Doomkid said:

I just feel like there are a few OS or even SP players that would go out and try some MP, but the feel that Odamex is too zDoomish.

It can't feel "ZDoomish" if it has vanillas exact physics to the point of 1.9 demo support, so I'm not exactly sure what you're going on about here.

[Doomkid]

I meant to say "they feel", not "the feel" - Damn typos. Some people out there, myself not included, are deterred by the ZDoom influence. There, I think that's worded far better.

Edward850 said:

Not even remotely possible. Vanilla demos have no concept of a player joining, or even a player leaving. Regardless if it's the consoleplayer or not.

Exactly, that's why I suggested some sort of "demo recording multiplayer mode" so players can't join/leave mid game, but if not in "demo recording mode", they could. I have a hunch that this may be a work around that could be used to allow vanilla deathmatch demos to be recorded. It would be nice to have this sort of option for a C/S client with a nice netcode so that players who live far away don't ruin the experience for everyone.

[Edward850]

Doomkid said:

I meant to say "they feel", not "the feel" - Damn typos. Some people out there, myself not included, are deterred by the ZDoom influence. There, I think that's worded far better.

That doesn't change the problem that any sort of "feel" seems to just be a placebo. Remember, you can't just casually have 1.9 Demo support. You actually need to have 1:1 physics support.

Doomkid said:

Exactly, that's why I suggested some sort of "demo recording multiplayer mode" so players can't join/leave mid game, but if not in "demo recording mode", they could. I have a hunch that this may be a work around that could be used to allow vanilla deathmatch demos to be recorded. It would be nice to have this sort of option for a C/S client with a nice netcode so that players who live far away don't ruin the experience for everyone.

So, basically what Chocolate Doom already does, exactly to the T? (Yes, Chocolate has a improved network model.) You can't join or leave mid game (well you can leave, but a demo is forced to end at that point), and can record 1.9 multiplayer demos.

Also there's something here that doesn't quite fit. If you use Odamex's network model, you can't actually record a 1.9 demo anymore, as none of the necessary player input data exists anymore for the demo to be function-complete. If a client couldn't join or leave, the demo still wouldn't work anyway. Among the rest of the client state which would be effectively broken due to the incomplete playsim state. Although that brings up an interesting theory about if you could match a client and a server state if provided the same input over time, however, that implies the server doesn't try to compensate for inconsistent input timings. Although considering the request at hand, you'd need to, and thus the playback won't work anyway.

[VGA]

Ahh, if people were spending as much time playing Doom in multi as OCDing over licensing issues and vanilla compatibility ... 😆

[Jon]

Gez said:

What isn't there to understand? They took an old version of ZDoom, got Randy Heit's agreement and approval for GPL relicensing, and took out like half of the code in their attempt at making sure everything they kept was either originally from Doom or from Randy's own hand.

Well that explains it, thank you. I was not aware that the code was relicensed by the author and the doom wiki article wasn't clear on that. (This might be useful data for the GLOOME people, too)

Edward850 said:

That doesn't change the problem that any sort of "feel" seems to just be a placebo. Remember, you can't just casually have 1.9 Demo support. You actually need to have 1:1 physics support.

There's more to a "feel" than the physics: there's the UI, the aesthetics, peculiarities of the sound or music code, etc. Perhaps Doomkid or the people he mentions are affected by those things (not wanting to speak for them, though).

[dew]

Edward850 said:

Although that brings up an interesting theory about if you could match a client and a server state if provided the same input over time, however, that implies the server doesn't try to compensate for inconsistent input timings. Although considering the request at hand, you'd need to, and thus the playback won't work anyway.

Doesn't that go against the core advantage of the C/S model? The game isn't supposed to be deterministic for clients anymore, because their connection isn't trustworthy. The reality is whatever the server tells you it is. In pure theory maybe a server-side demo could work for "vanilla playback"? With the slight disadvantage of having no "home" player recording their movement input directly, making the point moot, heh.

[Edward850]

Depends on how you look at it. In theory, because Doom is already deterministic, as long as the key factors of a deterministic playback are in place, the server could stream the required inputs back to the client overtime.
Of course, you do indeed eliminate the key advantage of an async framework; being able to discard old information instead of letting it hold up the pipe. It'd be the worst of both worlds (async scaled data and sync-pipe backlogs), and with Doom 1.9 it'd be worse, because all inputs must exist to have a functional demo. The last thing you'd want is two competing frameworks, uhm... competing for valuable packet space.