Is open sourcing UT a good thing? Why or why not?

[TankBeef]

Hi all. Just want to know what you think. Me personally, I am a bit on the fence, because of certain things that have happened to other games that are/ have become open source.

[Gustavo6046]

If bad things happen, the community can just not use them. With the source code available, they would be able to choose what they want to build with the source code. If an offshoot is undesirable, just go with another! That is the magic of open source - the freedom of choice!

[papercoffee]

And then those other versions wouldn't be compatible with each other.

[anth]

For offline projects, a fully open-sourced UE1 game would be great. As soon as you throw networking into the mix, things get more complicated. Even the most minor and subtle changes in the core libraries (particularly Core and Engine) or the base UnrealScript packages can irreversibly break network compatibility. Personally, I would love to include WinDrv and Editor as-is in the 469d SDK, but I certainly would not include Core and Engine without stripping them down.

[The_Cowboy]

I wonder how the model of UT4 worked (if you really want to say that). The engine and game, both had source posted online and people still got to play networked game. Who or what is stopping that model to get translated to UE1 and UT99 is a natural question!

[Feralidragon]

Well, if you want to take a look into a possible future where UT99 is open-sourced, just look at what happened with Linux itself.

In one way, it would be a great thing since it would finally allow the following:

• the game to be completely free;
• the game and engine to be expanded limitlessly by the community.

On another hand, it would also come with some severe disadvantages:

• the game would likely be forked into all sorts of different UT versions;
• mapping or modding for the game would become a nightmare for the reason above;
• the game would quickly loose its identity with each fork, and cease to be the game we know, since everyone has their own idea of "how the game should be" (as seen with UT4).

Especially if there's at least 2 forks that become popular on their own, and that's why I equated this to Linux, which has the exact same issue nowadays.

Therefore, for me, if Epic was at all interested in what is best for the game, in my opinion it should go as follows:

• as long as the OldUnreal team exists and is actively updating the game, the source would remain closed, but:
  • they would be allowed to open source some portions of the source, like the editor and such (like Anth mentioned above);
  • they would be allowed to bring more people at will, people they trust (they would need to sign an NDA with Epic, of course);
  • they would be allowed to provide a link for the complete patched game (making the game effectively free);
• the moment OldUnreal becomes inactive or decides to disband (nothing lasts forever), that's when both OldUnreal and Epic should open the entire source, and leave it up to the rest of the community to handle it from there.

This, for me, would be the ideal path to take, although it almost certainly won't happen given that Epic no longer gives a shit about these games, thus they won't do anything at all.
Hopefully one day they prove me wrong, but that's how it is for now.

[TankBeef]

the game would quickly loose its identity with each fork, and cease to be the game we know, since everyone has their own idea of "how the game should be" (as seen with UT4).

I won't add much, cause I agree with most of what you wrote, Ferali. But this paragraph right here is probably my main concern. The playerbase is not just small, but fragmented, because of this. I mentioned a game in a previous thread. I won't say the name again, cause I want to keep UT as the topic, but that game started with a different name. Trying to keep the story short, the author made the game open source, got people to work under an open source license, and then screwed everyone for personal monetary gain. He kept the name, launched a mediocre console version on steam, most of who contributed felt cheated, bailed out, and that is when the name changing and forking started. Now there is even a furry erotic fork of that game. Not that I am against that, it is not my thing, but it is not my business, I know. But clearly, some forks of that game are radically different. And strange.

[OjitroC]

I would think that any fork of UT 99 would be different - as far as I know, the terms 'Unreal', 'Unreal Engine' and 'Unreal Tournament' are trademarks owned by Epic. Unless the terms of any licence issued with the open source code provided otherwise, then none of these terms could be used to describe any 'fork' produced. Given that Epic are likely to produce a new or revised version of UT3 then I would think they would seek to guard against any unauthorised use of their trademarks (trademarked terms or names). This might preclude the production of more versions of UT99 that were explicitly stated to be versions of UT99?

Looking at it realistically, what is the likelihood of anyone seeking to expend time, effort and money on producing a clone of a 22 year old game (or even to do much in seeking to update UT99)? What is more likely (but not much) is the production of cheap, quick and dirty stripped down versions of a FPS shooter which has a passing but barely recognisable resemblance to UT99 or the production of other cheap games using the Unreal Engine - I say cheap in the sense of being inexpensive to purchase like DoonedFreedom, which I think is based on the Doom Engine.

[sektor2111]

Open the Source for UT and you will see a bunch or versions incompatible each-other and decades of bug-fixes. The most of players will lose patience and will move out. Is that hard to understand this ? I don't get why was even asked such thing.

#2 Bad boys figuring resources will gain an extra clean info about how to mock servers and develop cheats. Do you really understand why some resources weren't shared/opened ?

And No, its not like bad things won't be used, I think they will be used for multiple purposes - including to do damage.
If EPIC will want to eradicate UT completely, the most easy task to do at this point is exactly opening resources. The rest of destruction will be completed by community itself by messing up everything - including those who claim that they are loving UT'99.

[Dr.Flay]

Look at the DooM and Quake world to see what would happen.
At first you'd get a burst of competing rebuilds but that would settle into 3 or 4 main alternatives.
Network compatibility has always been a concern, but so far there are no big issues with using one client over another when it comes to online play.

There is also the other side to open source where downstream improvements and fixes don't have to be in their own little island branch.
Fixes and updates from branches can be accepted into the master so there can still be a consolidated and definitave version

[Gustavo6046]

And then those other versions wouldn't be compatible with each other.

I feel like the community would settle with the most popular one. There could also be community effort towards a common spec of sorts. That has happened before.

• the game would likely be forked into all sorts of different UT versions;
• the game would quickly loose its identity with each fork, and cease to be the game we know, since everyone has their own idea of "how the game should be" (as seen with UT4).

Especially if there's at least 2 forks that become popular on their own, and that's why I equated this to Linux, which has the exact same issue nowadays.

I see that as one of the advantages of Linux, the ability to choose. And the loss of identity argument seems to be a bit vague, I imagine UT would be mostly the same, with the same assets and all, and only the engine changes. For example, just because hundreds of Doom sourceports exist out there doesn't mean Doom is any less an original experience full of personality, even if there are so many ways to play it. All it means is you have choice.

What you essentially describe as an ideal path forward is OldUnreal becoming the effective lead developer of Unreal Tournament, with more open lanes of participation. Which certainly is better than what we have now, although it still feels a bit icky. I see that it seeks to solve the issues you mentioned with open source projects, but have you considered whether these issues really are as severe as they seem, or even bad?

Once again I will draw comparisons with Doom because of its rich history as an open sourced engine, it's a very good case for study. While many of its sourceports are not compatile with each other for more advanced modding features (e.g. 3D floors in ZDoom are different from 3D floors in Vavoom), almost all of them are compatible with maps and mods that target the original Doom, or the most common feature supersets thereof (Boom and MBF). Not only do they support those common denominator features, an issue with such a feature in any sourceport is regarded as a severe bug to be fixed, and not a mere incompatibility between source ports.

Now, while something like ZDoom might be able to play maps using the Boom featureset, doesn't mean it will be exactly the same in terms of determinism ("demo compatible") or internal aspects of the engine (e.g. hitscan collision detection, which in the original leads to the "blockmap bug" which is fully absenti in ZDoom). Even ZDoom's compatibility settings are not enough to retain full equity of behaviour. So case closed, forks are hell, long live the monolithic pedestal!... right?...

Well, here's the thing. Most people who use ZDoom don't care about demo compatibility. They want to play the most advanced mods there are, from Guncaster to Russian Overkill to Hideous Destructor. If they wanted demo compatibility, they would pick a port more suited for that, like DSDA-Doom orWoof.

Because, again, that's the magic of open source - it's the freedom of choice. I can't stress this enough.

And, no. You don't need to religiously apply a copyleft license to everything or have a Richard Stallman statuette ominously sitting in your trophy room. In practice, copyleft licensing tends to be a thorn on the foot more than anything else, even for larger projects. In truth, all it takes to be able to have freedom of choice is for code to be open sourced, and the community be trusted with organizing itself and finding a good development scheme. Because communities do organize themselves. They figure out leadership, hold elections, whatever. That they don't earn a penny doing it is just yet another reason you should trust them - they do it out of their own volition and interest. They are given the ability to contribute to something they love. And by damn they will.

The playerbase is not just small, but fragmented, because of this.

Maybe it would be less fragmented and less small, if people had more access to its internals. I mean, that's basically why classic Doom is still talked about to this day. Good games live longer because of the community.

I mentioned a game in a previous thread. [..] Now there is even a furry erotic fork of that game.

Clearly, some forks of that game are radically different. And strange. :

Psst. Maybe people being able to choose what direction to take things in is a good thing. People don't have to participate in that fork if they don't want to. I've seen that fork. It actually regards itself as a completely different game, not a Xonotic port. So if nothing else, it certainly doesn't encroach on the identity problem. That just highlights that they have freedom of choice, and we don't.

#2 Bad boys figuring resources will gain an extra clean info about how to mock servers and develop cheats. Do you really understand why some resources weren't shared/opened ?

Who would have thought, security by obscurity is a bad thing. How come open source software aren't full of exploits then? Because a lot of people can pour eyes over the source code and fix vulnerabilities, and push hotfix updates fast, not beholden to a corporate release schedule. You can't use sneaky exploits if there are no exploits that are hidden, and all of the ones that weren't hidden are quickly found and fixed!

Open the Source for UT and you will see a bunch or versions incompatible each-other and decades of bug-fixes. The most of players will lose patience and will move out. Is that hard to understand this ? I don't get why was even asked such thing.

No, you won't. People will be able to choose what version they want to pick. The main forks will seek to be compatible with each other and probably join forces to write a new network spec. And players won't "lose patience", because "the most of players" are not you. They will settle in their favourite versions, which will share a lot in common with other versions, and most likely maintain compatibility.

Plus, most certainly some parts of UT are begging to be rewritten, including network code. Maybe such would be a change for the better, and nothing prevents other versions from adopting it too - in an open source model, they can just merge commits they like from each other, and still work towards each individual version's mission statement.

---

People here clearly have no experience with an open source model of development, and what tends to happen in one. Except Dr. Flay. It's astounding to me how ignorant people can be of such a thing in 2023, when 99% of software that is used in the backend today is driven by open source development. From databases to web servers, even the browser you're using (with massive caveats). Even OpenMPT! I mean, that's where the "open" in OpenMPT comes from!

I think it's time for you all to go to school - open source school - and get a crash course on how it works and why it works.

• "What is open source, and why does it matter today?" - Open Access Government, last accessed 2023/02/02
• I will hesitate to plug in The Cathedral and the Bazaar, given Eric S. Raymond's polemic reputation and tendency to say dumb things, but it's a good early work that delves into how the open source development model works.
• The Pros and Cons of Open Source Software Development - freeCodeCamp, last accessed 2023/02/02. Notice that, of the cons mentioned, none of them are about incompatibility. But rather, about "possible misunderstandings" (communication isn't perfect), which is an issue in all sorts of corporate environments anyways, and "potential threat to the core business value", which... maybe the solution is to know how to run a business to begin with.

I could go on and list more and more sources that emphatically glorify open source, and I could also go and look for sources that are more skeptical and critical of it. But the point is that it's just very different from how development happens in a closed environment, and if you really want to argue against open sourcing UT (and I'm sure there are valid arguments in that vein), you must at the very least know your enemy. Bla bla bla Sun Tzu bla bla.


EDIT: Sorry if I sounded condescending at any point. That is not the intention. I just feel like it would be beneficial to the conversation if we could all be in the same page, therefore I wanted to encourage everyone here to read up a bit about how open source development works in a public project, and still do. It's pretty neat and an interesting outlook into a surprisingly productive and successful alternative to traditional corporate agile development.