Porting Xonotic to idTech 4

Xonotic is one of few games which can run 100 fps on integrated GPU which is probably related to use of Darkplaces.
Other games are made with assumption that everyone has latest power hungry vacuum cleaner with discrete GPU in order to Nvidia was profitable.
I and people with old PC could probably play only games like Quake 3 if Xonotic would switch to different engine.

Porting Xonotic to a better engine would be a god send IF it's possible to do it without changing the minimum requirements of the game. Currently as is, Xonotic can be a monster if you crank the settings up (probably because the engine doesn't know how to properly use every resource of a gaming rig), but playing with default or only slightly higher will enable to run on just about anything, hell I could probably make it run on a Pentium 4 and Geforce 7600 if I turned some stuff down. Also, with the way it's development has been moving, it has a high chance of being a game that everyone could just max out and run over 100fps by the time it finally gets a 1.0 release.

Different engine doesn't mean that the framerate will go down and Unvanquished runs smoothly.
Btw, people have problems running Xonotic with stable framerate.

I tried latest Unvanquished and couldn't exceed 40-50 fps with bots at worst quality.

Unvanquished has very high quality maps and art, I have tested a Xonotic map in that game, it runs at almost the same FPS.
Performance and requirements most likely won't change much.

It seems Sauerbraten also has good engine. I have higher fps with probably higher image quality than in Xonotic.

Red Eclipse (a fork of Sauerbraten) uses the same textures as Xonotic (the Trak texture set seems to be a commonly used set).

Not really feasible to move to Sauerbraten, we would lose client side VM code, and would have to recreate all our maps (probably from scratch).

There's a distinct feel in how a game plays in different engines, changing the engine is basically changing the entire game and there's just too many configs, maps and mods that won't necessarily have the same level of support on other engines.

How about we do what Nexuiz failed to do and not switch to a new engine with entirely new weapon/movement physics and unoptimized performance that more or less changed the game experience from head to toe, outraging the playerbase in the process and ultimately going bankrupt.

In my opinion, supporting of external scripting language is *VARY IMPORTANT*. It makes the development much easier and saves a lot of time for the developers. Almost all of the *MODERN* game engines support scripting language:
- Crysis Engine 1,2,3 supports LUA,
- Unreal Engine 4 supports UnrealScript,
- Unity3d supports Javascript,
- Torque3D supports TorqueScript,
- Source Engine supports VScript [layer which supports Python, Lua, Squirrel, Gamemonkey] [also not that modern],
- ShiVa supports Lua,
- Id Engine 4 supports Lua,
- Blender Game Engine supports Python (And it looks awesome),
- And so on...
(This was a quick research so the information may not be 100% correct).

And when we are talking about MODERN game engines, i am thinking about something like this:
https://www.youtube.com/watch?v=I4t43RM5aVc
https://www.youtube.com/watch?v=rCcEXQuH6Fg
I know that i wouldn't be able to run most of this games, but this is considered modern now. (I know that this are high budget games, don't kill me.)

However the point of this post is that the scripting languages are MUST HAVE and saves a lot of time for development. Something you can do with 20 lines of C code, you can do it with 5 lines of script code.
And Lua especially is designed to be integrated into any kind of software. And it has small memory overload (400kb ?).
Link: http://www.lua.org/uses.html

So integrating something like this could save a lot of work.

'Scripting languages' can be useful to allow servers to modify aspects of the game both server side AND client side, but we already have QuakeC for this (which is compiled to platform independent bytecode)

Almost all of the *MODERN* game engines support scripting language:
...

All but one of these languages (UnrealScript) is dynamically typed.
Dynamically typed scripting is well suited for small higher level functions such as gameplay rules like our mutators because it is easier for the non-programmer to use, but not the entire 100kloc codebase.
Even though QuakeC provides little type safety because it only has 3 or so types (a problem we can solve with compiler improvements - entity classes), going fully dynamic would be a step backwards in my opinion.

And when we are talking about MODERN game engines, i am thinking about something like this: https://www.youtube.com/watch?v=I4t43RM5aVc
I know that i wouldn't be able to run most of this game, but this is considered modern now.

Modern engines aren't just about graphical effects, there are more important things such as netcode, gamelogic isolation, maintainability, and extensibility.

Something you can do with 20 lines of C code, you can do it with 5 lines of script code.

You might be able to do whatever it is in 5 lines of script code, but how many library functions are you using in comparison to C (hidden as language features or not)?
You are comparing a library of functions to basic core language features, of course the library will win
There's nothing stopping you from writing the same library features as functions, which has been done in many places.

As someone who worked a lot in the past with Id Tech 4, I can say that it's certainly not an engine that is worked easily. I prefer the editor over NetRadiant. But a lot of the tools are half baked and ruddy. And the information in the OP's post isn't really accurate.

It's a good engine. But if you were to make a port to it, you would have to rework almost everything. For instance something as simple as the materials has to be rewritten and features added since Id Tech 4 doesn't support things like Parallax out of the box.

But one question I have for the developers of Xonotic. You all lament the archaic nature of the bots waypoint system. But all the maps use Id Tech 3 map format. A format that introduced the very robust AAS (Area Awareness System) that is also used in Id Tech 4 and 5. Why isn't this system added and used for Xonotic as well?

There's a distinct feel in how a game plays in different engines, changing the engine is basically changing the entire game and there's just too many configs, maps and mods that won't necessarily have the same level of support on other engines.

How about we do what Nexuiz failed to do and not switch to a new engine with entirely new weapon/movement physics and unoptimized performance that more or less changed the game experience from head to toe, outraging the playerbase in the process and ultimately going bankrupt.

Xonotic feels perfect to me. Like a Quake-roots game whereas the actual Quake4 doesn't. What scares me is a move means there's another game a bunch of people won't migrate from, eg: right now (from serverlist anyway) nobody is playing Xonotic, there's 2 playing Vexcis, and about a dozen playing Nexuiz.

There shouldn't be any migration needed.
The port to Daemon will be made as smooth as possible, the release that includes it will feature very few differences, and will not be a completely different game.
Afterall, Daemon is still Quake based, just a bit more advanced (only slightly though, the jump from Quake 1 to Quake 3 is surprisingly small, with most of our features and maps working out of the box).
Much of what makes Xonotic special is done in the game code anyway (including physics & weapons), which is already in the process of being transpiled to work with Daemon natively.

Most of what remains of the Nexuiz community are players of a strange mod called "Rocket Minsta", which some would not even call Nexuiz, it's so different.
Nobody plays the vanilla game anymore, from what I've seen.
Vecxis is still being developed, and so may have the developers play testing on servers occasionally.

As for Xonotic... there's a time I call "dead hour", in which nobody is playing. You may have checked during this time (outside of that time, there's usually 10-16 players on some servers depending on the time zone).

Xonotic feels perfect to me. Like a Quake-roots game whereas the actual Quake4 doesn't. What scares me is a move means there's another game a bunch of people won't migrate from, eg: right now (from serverlist anyway) nobody is playing Xonotic, there's 2 playing Vexcis, and about a dozen playing Nexuiz.

261 active players and 442 games (170 ctf; 138 dm; 99 duel; 28 cts; 3 ft; 4 other) in the past 24 hours.
http://stats.xonotic.org/

Since creating this thread, I've come to realize that a port to idTech 4 would be too big to consider, apart from the fact it would result in a completely different game. I still think it would be awesome, no doubt... but not practically doable as a port.

If anything, maybe I or another interested developer might someday take all the assets and art from Xonotic and create an entirely different project with them on it or another engine (UE4 or Unity). That is a little more probable, but obviously no such plans exist on anyone's list.

There's a distinct feel in how a game plays in different engines, changing the engine is basically changing the entire game and there's just too many configs, maps and mods that won't necessarily have the same level of support on other engines.

How about we do what Nexuiz failed to do and not switch to a new engine with entirely new weapon/movement physics and unoptimized performance that more or less changed the game experience from head to toe, outraging the playerbase in the process and ultimately going bankrupt.

Xonotic feels perfect to me. Like a Quake-roots game whereas the actual Quake4 doesn't. What scares me is a move means there's another game a bunch of people won't migrate from, eg: right now (from serverlist anyway) nobody is playing Xonotic, there's 2 playing Vexcis, and about a dozen playing Nexuiz.

I apologize for being off topic, but can someone please explain to me what Vecxiz is? and is there anybody noteworthy developing on it?

The crazy text here should explain better than any sane person could: http://vecxis.org/about.html

If anything, maybe I or another interested developer might someday take all the assets and art from Xonotic and create an entirely different project with them on it or another engine (UE4 or Unity). That is a little more probable, but obviously no such plans exist on anyone's list.

I would quite like to see that happen in the near future to be honest. If you or someone else was going to do it, I'd like to see the game aimed at a more casual player base, given all the competitive FPS games emerging, something more towards Nexuiz style (strong weapons, high push, different style weapons) etc. Nothing worse then not having players on public servers. I know Nexuiz and Xonoti c also go down well with the open source/linux community.

If you did do it, I would hope you did it as a closed-/alpha beta style, then release public once it's at say version1.0. One of the issues these days (something Xonotic also suffers from), is released before a major version and incomplete features or lack of content, people tend to play then get bored quickly and the game never seems to quite recover.

This is why we haven't advertised Xonotic on Steam (or really anywhere officially).
Most of our current player base are Nexuiz players, with some new players from the few places the game has been announced.

Once we do reach 1.0, it'll be announced officially, and hopefully will be polished enough to hold a player base...

I have seen a lot of people surprised Xonotic is still in development, as the pace of it is so slow.

I did have hopes before but I just don't see much changing even with a 1.0 release, the game is already highly polished.

I think the lack of updates and the time it has taken to reach version milestones, has unfortunately negatively affected the project. That's not to say I for one haven't been appreciative of the devs efforts, I just think the project lost a lot of steam some time ago.

To be honest I think the games peak has been and gone already, which happens to all games, Xonotic is what, 5 years old now? That's one of the reason why i would quite like to see something new made (just my personal preference).

I do think Nexuiz for example, had a few things that made it so successful. Open source (free), well liked by Linux groups, but also that it wasn't made as a competitive game - Xonotic has gone more that route, coupled with the fact it was actually easy to get into and easy to play and frag, modding ability also. It did also help that FPS games were still played quite a bit then ofcourse.

Anyway, I feel im veering this thread off-topic, bottom line is I do like the proposol to port or make something new on a newer engine.

Switching engine seems like a lot of redundant work, but if it's not as difficult as it sounds and the devs are keen on doing so I don't see any problems with it. Even if the game "feel" changes slightly it could as well be for the better.

Well... looks like we might see a port to idTech 3.75, aka the Unvanquished engine (see comments): https://www.unvanquished.net/?p=684 A bit surprising, but overall not a bad choice if it works out

I'm eager for the Daemon (Unvanquished engine) port as well! Darkplaces is nice as well, bit I too get the impression that it's less supported. Not sure if this move might mean losing some features however... namely in the department of rendering and visual effects, although I assume Daemon should have all shaders DP does.

That seems the main point... DP is nice, but more or less unsupported by now, while Daemon has some pretty active engine developers. Shader feature wise it should have actually surpassed DP by now, but yes the Unvanquished assets often don't show this... they seem a bit on the conservative side when it comes to the use of shaders...

I wonder if all the game code will have to be rewritten or if they plan to implement a QuakeC interpretor in Deamon. All the rest should work out of the box given the engine similarities.

Darkplaces is horrible. With Intel i3 and integrated I get single digits fps, 5fps. Disable audio, disable textures leaving all surfaces solid colored, disable lights, everything that can be disabled, I join a CTF server where ppl are playing in a open floating in space map with grappling hooks, totally unplayable fps. The only thing that is left to do is play at 320 x 200, but then you can't read anything with such low resolution.

You should get about 60-100 fps at Low quality. You have something wrong on your system.
http://www.phoronix.com/scan.php?page=ar...some&num=6
http://www.phoronix.com/scan.php?page=ar...pare&num=4