Dungeons of Dredmor Interview, v1.0.10 Patch, and Project Odin Teasers

After you’ve finished reading through Dungeons of Dredmor’s v1.0.10 patch notes and checking out the trio of Project Odin teaser images at the bottom, I’ll send you over to Gaming on Linux for a post-mortem interview with Gaslamp Games’ Nicholas Vining and Daniel Jacobsen. As the site focuses on Linux gaming, many of the questions revolve around that particular platform:

Now, many developers (and especially publishers) seem to still not want to acknowledge Linux as a viable gaming platform. What are you thoughts on this – what can be done to change their attitude?

NV: There are two major problems: toolchain* and revenue.
* toolchain – programming tools used to create a software

The toolchain situation is getting there, but the Linux distributions and vendors do not go out of their way to make it a happy playing field for anybody who wants to release commercial software for Linux. We can’t, for instance, ship an installer and have it just work. We also have a large volume of issues related to (packaging hell), where users want packages built for their distribution, and which we’re simply not set up to deal with. We don’t have the time and the energy to build separate packages for Debian, Ubuntu, Red Hat, SuSE, Gentoo, Slackware, whatever, and to test everything thoroughly. Desura makes things a lot better; it gives us a good commercial sales portal, and you can download Dredmor and have it Just Work. I would love to see the entire packaging mess just die in a fire, and for commercial software to be treated as a first-class citizen by the Linux packaging process, but… that’s just not going to happen.

I’m a little worried about the graphics driver situation for the next title, as it will be 3D. The closed-source NVIDIA driver always used to the gold standard by which everything is measured, and I’m pretty sure that it will work well. I haven’t tried the closed source Radeon driver yet. The open source drivers are amazing technical achievements – especially nouveau*. How the heck do you reverse engineer a graphics card? I have no clue. We’re happy to provide software and information for driver teams who want to get our stuff work on the open source driver stack, once we’re closer to release. I just don’t think it’s realistic, though.
* nouveau – an open source driver for nVidia cards

The revenue situation on Linux is grim. I don’t want to name numbers and percentages, but it’s a couple of orders of magnitude behind how much money we make on Windows. From a purely commercial standpoint, if not for the fact that our Linux support allowed us to release on the Humble Introversion Bundle, we would have lost money on our Linux port. Furthermore – and I have no idea why – Linux continues to occupy a disproportionate amount of our support load. So there are definitely major issues for anybody who actually wants to ship a Linux game.

We do it because it’s the right thing to do, in some sort of weird, hacker-ethos sense, and not because it’s profitable. I don’t know how to make it profitable, either.

In general, the needs of commercial software are not treated well by the Linux maintainers and development community. The best examples of this are the whole package fiasco, which I think I’ve already railed on, and the joys that we have to go through in order to get binary software for Linux to work at all. It’s not that it’s a difficult process, but it is time-consuming and there is a certain amount of undocumented (black magic) required that you simply have to know. If you are not lucky enough to have this knowledge passed down to you from the previous generation of Loremasters*, resplendent in their robes of penguin hide, then you’re screwed. Good luck trying to figure it out on your own.
* for those confused readers, a “Loremaster” is a very knowing person. In role-play it’s class akin to a bard

What’s even more baffling is that any attempt to get this sort of thing fixed tends to be instantly shot down. I hate to bring up FatELF* again, but the sheer hostility that it faced was kind of mind-boggling. A lot of that hostility probably stems from the fact that FatELF was a project designed to make life better for commercial software, and commercial games in particular. In the Linux community, commercial software is an abomination in the eyes of the faithful – and you know what? That’s fair enough. I love that the open-source ecosystem is as good as it is… but it could be so much better.
* long story short: it’s a programming thing. ELF stands for “Executable and Linkable Format”

It boils down to money. If you want Linux on the desktop to succeed, the Linux community has to accept that commercial, closed-source software, of the sort usually sold for money, is a necessary part of that ecosystem. One of the reasons why Linux in the server market has been so successful is because it’s possible to make a great deal of money off of Linux servers and open source software in the server market. That money then gets injected into the open source development sphere, and everybody wins. If you want a better Linux desktop experience, and a better open-source Linux desktop experience, make it easier for us to make money off of it. If we’re making money, we’re willing to help get things fixed and make life better, and we can use open source software to do so. Everybody wins.