It's Been Three Years Since id Software Publicly Parted Ways With Linux

[phoronix]

Phoronix: It's Been Three Years Since id Software Publicly Parted Ways With Linux

While id Software used to be the game company that was very Linux-friendly and always porting their titles over to Linux even when its gaming market was tiny and often overlooked by other game studios, today marks three years since they came out to say Linux hasn't produced positive results and since then haven't released any Linux-native titles as it doesn't "pay the bills" for the level of work involved...

http://www.phoronix.com/scan.php?page=news_item&px=id-Software-Three-Years-Linux

[s_j_newbury]

Many game makers are coming around to the idea that writing portable code is more profitable that being stuck on a single platform. This, to me is what is strange about the id policy, once you've written portable code, compiling a Linux binary, or even better releasing the source* for the engine just gives even better distribution and possible sales of your product.

I actually bought many id software games over the years, from wolf3d -> doom -> quakes mainly because they were multi-platform, port-friendly, and later released game engines source for older releases.

*Assuming you can accept the idea that the artwork, levels and game narrative are the value add where you charge for "content", if that's your preferred model.

[dimko]

ETQW was their best game, value for money.

[kalasa]

Many game makers are coming around to the idea that writing portable code is more profitable that being stuck on a single platform.

Most game developers already target multiple platforms. That has been the case since the the first Xbox since that tried to compete against Playstation. There is still a cost associated with porting your codebase to a new platform that you have to weigh against the potential sales. This is not a one-time investment, you have to constantly maintain and improve/extend that codebase and devote QA time to it. Windows almost always gets a release since most studios run windows on their desktops and you need to be able to run the game on your computer during development.

It's also not the developer that decides which platforms to release on, it's the publisher. Even if a developer would create a linux version of a game using its own money it can't release it since the publisher owns the rights. There are many people at the company I work for that would love to do an opengl version of a game we release years ago, but we can't since the publisher doesn't want to.

[kalasa]

... or even better releasing the source* for the engine just gives even better distribution and possible sales of your product.

Most games today use middlewares that don't allow you to publish the source or the binaries. There are some middlewares where even the developer doesn't have access to all the source and you have to request binaries for specific compilers/platforms. A normal list would be something like Umbra for geometry culling, Scaleform for UI and Havok for physics. Sure, you could release the source with these parts ripped out and let the community write/integrate replacements. But it's kinda hard to write a new UI system that works with Scaleform's data format (which is has to do unless you want to remake the UI). Even using the free version of Havok would be problematic since most companies have their own branches with modifications to the source that can't be published.

Switching to open-source alternatives would be awesome, but they aren't competitive enough and don't provide the support that you need as a professional developer (e.g. being able to say "this code is too slow, optimize it please and tell us what to do differently" and get response within a day at any time during the 3+ years it takes to develop the game).

[Calinou]

Switching to open-source alternatives would be awesome, but they aren't competitive enough and don't provide the support that you need as a professional developer (e.g. being able to say "this code is too slow, optimize it please and tell us what to do differently" and get response within a day at any time during the 3+ years it takes to develop the game).

Godot Engine would like to prove you wrong. Moreover, creativity is born from technical restrictions; it's ultimately how we got Doom and Quake. If PCs were powerful in 1990, we would directly have had realistic shooters.

The other problem of using those "open source" libraries is that, while it's a nice thing, there is no guarantee the end product (ie. the game) will be free/libre...

[kalasa]

Godot Engine would like to prove you wrong.

I should have been a bit more specific: open-source alternatives (complete engines or most other middlewares) aren't good enough to create AAA-titles (e.g. the next Assassin's Creed or Destiny). There are ofc lots of open-source engines that people use to create games, but since this article was about id software I assumed we talked about AAA-games. There are also other libraries that are used extensively (lua, zlib, etc).

Moreover, creativity is born from technical restrictions; it's ultimately how we got Doom and Quake.

This is the reason why most games that try to push the technical boundaries use in-house code. You need to do ugly stuff that's very specific to your problem, which is the total opposite of a good, reusable codebase. Companies don't write their own physics engines because they can't make it faster. They license an existing one because innovation in that area doesn't translate into more sales and the gains wouldn't be big enough.

The other problem of using those "open source" libraries is that, while it's a nice thing, there is no guarantee the end product (ie. the game) will be free/libre...

That's true, but I would hope that the more (paid) developers there are working with a codebase the more likely it is that it will evolve in a good direction. LLVM is a good example. I don't think that it would've been where it is today if it wasn't for the heavy investments Apple, Google etc have made.

[blackiwid]

I can't here this argument that Linux user base (steamuserbase) is 1% if microsoft would make desitions in that mindset, they would never released the xbox because a tipp before the console hit the market they had 0% market share so 1% is much compared to that.

to id soft, yes it did not work out for them because nobody installs linux on a gaming rig for one game and wine is not very compeling, but if they would not so stupid thef would see that thats now different.

We failed, so everybody else must fail, too. Else we would have made something wrong (or had not the resources to do it right) I cant be guilty so somebody else must be guilty this cheap linux users.

[gbudny]

Id software decided to abandon the idea of porting games to Linux many years ago, and I do not think about the unofficial games e.g. Doom 3, Quake 4 etc. They ported Quake: The Offering Quake II: Colossus, and Quake 3 to Linux, and these games were not successful among Linux users. They let some people in the company to port their games to Linux for free, and I do not think that ZeniMax Media will be interested to make similar experiments in the near future.

We have the similar situation in the case of Activision, and sometimes smaller companies made similar decisions e.g. Hyperion Entertainment.
I can easily point out many indie game companies that they removed games for Linux from their website.

We can observe that users who use Mac are in the worst situation because they have always paid for every game for Mac.

The new owner of this company maybe changes his mind after the releasing of the steam machines.

I have never been a big fan of the companies who published unofficial binaries for Linux because they wanted to use Linux users to pay for the games without the technical support. In this way, they could spend money of the Linux users on the production of their new games for Windows, and we can not blame many companies that the majority of Linux users tried naively accept this practice as something normal on the game market.

[ruthan]

It would be nice benchmark some game running in Wine - on both NV and AMD card and compare it to native Windows performance and Rage could be good candidate.