Technical Director of Banner Saga comments on state of Linux-Version and on Linux as platform for games

Banner Saga Header

John Watson, the technical director of the praised Indie-game "Banner Saga", comments in a post in the company-forum to the current state of the Linux-version and the status quo of Linux as a gaming-platform.

Currently the game is nearly unplayable on Linux in the main-branch on Steam because of the very low FPS in the battles. The developer know this already from the prolonged feedback from Linux-users, and already released a beta-version called "linuxfix". In this beta, the FPS in the battles are good, but the beta is lacking many other features like music or localisation. However the company hopes to get it out with all bells and whistles in November. See this forum post at Steam for details.

Please read the full post of John Watson below. Sentences marked with ">>" at the beginning are from other users of the forum.

>> Being left on your own fate to roam into uncharted territories is kinda bad, yes... Even more so if you relied on that high-level/complex technology to make your life easier, and it just sweeps your feet from under you =/ If you were thrown back to 2012, would you have gone with AIR again?

Yes, I probably would. It certainly made the initial development of the game go faster, as well as the porting to mobile. However, I might consider writing the entire engine in a language like C++ instead of AS3. Still using AIR for the platform-specific framework and the its low level renderer, but keeping all that behind a nice abstract interface that can be swapped out later. I'd have to think about that some more though, to determine if it's really feasible.

I did make a prototype of the game in Unity back in 2012 as well, but based on some technical limitations at the time, I felt it was not quite ready to trust for the project. Of course, these days it seems like Unity is the best choice for many things.

>> Thanks for the heads up on the PS & XB ports as well as for the Linux problems. That Linux community has developed a patience throughout the long years of waiting (usually last in line, after Windows/Mac etc), but it's better not to overtax it!

Agreed with that. Before I got into professional game development I was 100% on Linux. My first amateur game was actually developed on Linux and cross-compiled for Windows. I wrote a Linux/Unix level editor from scratch for Quake because I hated having to reboot to use Worldcraft or whatever I was using back then to make Quake maps. I can still vividly remember the joy I had, and the hours I played, when Loki games ported Heroes of Might and Magic III to Linux.

The downside for small developers like us is that Linux is such a tiny fraction of the gaming market that developing for it is pretty much a labor of love and there's no expectation for it to even pay for itself in sales. I can tell you that Linux amounts to less than 1% of TBS sales, and that 6 months from the TBS Linux release, we are about 25% of the way to the port paying for itself, and if the current 3 month moving average holds steady it will pay for itself in only 70 more months of sales. However we both know that game sales always decline rapidly over time. Now, if we could manage a simultaneous release, things might be a bit rosier financially. I'm also very curious what will happen to the Linux market when the Steam Machines hit the market in November. If anybody can pull off a game-change like that, Valve can.