Chronojam, on 04 March 2014 - 06:39 PM, said:
It's definitely the right path forward. But, I still have to question the wisdom of releasing a bad implementation that you know is bad. It won't sell users on the time/effort you've invested in the feature, and if you're only doing it to hit a deadline you'll look really foolish.
I've been there before, with pushing out a suboptimal DirectX upgrade to production before, so I'm feeling empathy here but that's also why I'm screaming "don't do it!" in my mind. Don't do it if you're just trying to prove you're working on it, especially if you have a test server people can go see. Use the test server!
Seriously, use the test server!
Couple of points, for users who recently bought a computer running Windows 8 or that came with DX11 by default they no longer need to downgrade to DX9 to be able to run the game so there is benefit beyond just performance/visuals.
For those that have worked in development you know that engine integrations, especially on a rapidly moving project like this that code branches have a tendency to diverge, its not as simple as keep it on the test servers until we're 100% happy with it. To do that we incur constant integration overhead to keep it up to date with the main trunk and exhaust more QA resources in the process. If you're not happy with where DX11 is at currently you're welcome to continue to use DX9 this is why we didn't switch to DX11 by default at this time, but at least now users have an option.
Personally I'm happy to now be able to push forwards with DX11 to make further improvements to the game knowing its not time invested in a code path that will quickly become legacy code.