It's really cool to see an old game engine released like this for historical purposes. Not to mention, it's nice to be able to experiment and see other games become derivatives.
Still, I can't help but feel that the game industry is always a bit behind in FOSS. They don't usually open source anything until a decade or two after the tech is irrelevant. Perhaps it's to prevent competition.
Another thing that bothers me with these releases is how little documentation there is. They put up a repository without much intention of helping/building a community around it, which is a shame.
All that said, they still have the right to approach in this manner. It is their property and they are still generous when making releases like this.
Most higher level game code is actually fairly uninteresting (if not to say 'a tangled mess') and wouldn't really bring the programming world forward if open sourced. On the other hand, most of the building blocks and information to build games is in the open and can be contributed to, even if not permissively licensed. And there are a lot of lower level building blocks that are open source with a proper license (mostly MIT).
So basically, the lower you go, the more open source you will find, and the higher up you go in a specific game's code, the less useful releasing as open source would be.
Developers would find an immense amount of use in open sourcing their engines like this. It lets the community port them to every OS in existence including your toaster. A great/bad example is gemrb, which was an open source reimplementation of the Infinity Engine that almost certainly incentivized Atari / Bioware to let Beamdog re-release the games with modern engine updates across many platforms due to the free engine's popularity on mobile.
The open source nature of all the id engines is what keeps Wolf3d through Quake 2 games relevant this day. Their engines are bought to every new platform almost as a benchmark for porting software period, and you can play them effectively everywhere as a result. Compare that to abandonware titles like Blood that were their contemporaries, but never saw source releases and thus are effectively dead software that would only run on DOS.
I'd argue almost every console game release since the Xbox would benefit astoundingly from open sourcing their engine code. If we could see the APIs they wrote their renderers against we could reimplement them for modern cross platform tooling to enable native compilation and execution of games we still cannot emulate well because the hardware is similar enough to what we have now that we don't have the magnitudes of ghz to throw at it easily or often. And even older console titles would benefit, because they could be modernized with new resolution support, higher refresh rates, and modern rendering techniques.
This is all around immensely beneficial to developers because it is basically free money - people will keep buying copies for the assets even if they use much better engines you didn't have to put developer hours into making.
> Another thing that bothers me with these releases is how little documentation there is. They put up a repository without much intention of helping/building a community around it, which is a shame.
I hope there's a way to link all the people who used to make mods and maps with this repository.
Or even worse, it's not released until there's noone left to answer about the licensing of the code and assets. The community that took over Microsoft Allegiance's code has been trying for years to find someone able to make decisions about it.
The code is usually terrible in various ways (mostly due to the culture in game dev studios). Even Quake's code, which is quite good, includes various "what the fuck?"[1] comments
Unless you are a huge company that can have his own in house engine with it's own team working on it, you will basically start a game from scratch with DirectX and with an engine built only for the needs of your game. You will finish the game, release it, and now you will have to start everything again because the game you just built - with its code - is not advanced enough to ship the next game with the graphics required in 2 years. So rather to have to fight by refactoring 60% of your code you start a new engine again, or else you cut hard in it.
And secondly, a game is here to be consumed for a short period of time. After that the whole product just died. Maybe it will have a few patch, but that's it. Where for a software, it is supposed to be constantly evolving and here to stay a few years, rather to be consumed for say 6 months, and then discontinued.
And lastly because it's an entertainment industry. There is nothing huge happening if a bug comes. No fortune 500 will be down, just the computer of someone somewhere. So it's "better" to ship fast and correct a bug when you see one, rather to spend a big amount of time writing tests that you won't be able to use for your next game anyway.
I loved this game, it and games like Painkiller feel like the last of the super face paced crazy FPS that defined the genre for me.
Would like to see more older engines released.. technically they can't be that important now and with engines like Unreal being developed in the open the tech is no longer really any secret to anyone.
>There is a long history on this topic, but the short summary is: We've looked at this extensively and concluded that releasing UE2 & UE3 source would be impractical due to the large number of integrated third-party libraries which have limitations on redistribution. The public source code release only became practical with Unreal Engine 4, due to an extensive effort to clean up external dependencies.
>UE1 might be possible at some point in the longer-term future, but that would be more for nostalgia than practical utility.
Indie game developer Brendan Chung (http://www.blendogames.com) used a very lightly modified Quake Engine for his two short-form experimental narrative-based games, Gravity Bone and Thirty Flights of Loving. (The first is freely downloadable, the second you have to pay for). He's also working on a new game, Quadrilateral Cowboy, using the Doom 3 engine - apparently the gameplay might be more conventional for this one, though hopefully not too conventional...
The Quake 3 engine is used for fairly well-received open source arena shooters like Tremulous, Xonotic, and Nexuiz.
The Dark Mod is a Doom 3-based Engine for fan-made games in the style of the Thief Series
You also have a few ports of engines to cope with modern windows versions or non-Windows operating systems or just updated gameplay (Jagged Alliance 2 straciatella or 1.13 springs to mind).
Let's not forget SSam's greatest contribution - a copiously lubricated revolver that shot in 1us of clicking a button and with a very satisfying * thump * :)
To me, Serious Sam feels more like a 3D version of Robotron: 2084 than a traditional FPS. Very high numbers of simple enemies, enclosed featureless arenas and large open spaces, interactions between the various enemy types forcing improvisation instead of memorization. Robotron is a great game, so this is a clever way of designing interesting fights without needing to spend a long time on map design. The later games switched to a more traditional FPS style, but the original is something unique and I highly recommend playing it.
Typical games I can think of are Doom 1&2, Quake 1,2,3 , UT, Serious Sam and Painkiller.
Big weapon pack, Six axis free camera movement, free movement in the levels, free jump, no contextual actions,no cover mechanics and the ability to save during arbitrary points (how the games regressed from that to the current checkpoint absurdity is anyone's guess).
> how the games regressed from that to the current checkpoint absurdity is anyone's guess
Obviously because of consoles. Earlier consoles had limited save game storage cards, and the cover mechanics was introduced to give people with gamepads a chance.
Early cross platform games like GTA 3 and Splinter Cell were a lot easier on PC than on PS2/Xbox consoles (aiming with the mouse is superior). I had hopes that we get more games especially for PC, games used to be a lot better until we got all these console ports starting with Xbox 1 in 2004 and later. But the outlook with Win10 and its spy components is so bad that one can only hope Vulkan (OpenGL next), SteamOS and HumbeBundle get even more traction.
The new Wolfenstein was more like Half Life, and not a pure FPS. With restricted movement in several gameplay passages, slow narrated passages, etc. A very good game for sure.
What is this definition of "pure FPS" you're using? Isn't FPS simply First Person Shooter? If HL is not an FPS, I don't know what is. FPS article in wikipedia has a screenshot from the original Half Life.
Any future source code releases won't happen with ZeniMax Media as the owner of the IPs. They are investing all their studios in the use of the same engine.
- (Same link) https://news.ycombinator.com/item?id=11268130
- (Github) https://news.ycombinator.com/item?id=11267450