Hey folks, So we distribute our application as a binary to Linux users with our build of Qt 5.2 included. We rely on the libGL.so.1 from the users system. Everyone has this installed so that's not a problem - but it seems like many users have broken video stacks and now have trouble running our application now that we are using QtQuick 2. Well users with broken Linux video stack - no big surprise. While these users could probably fix things by getting their drivers installed properly, in the end we need something that Just Works™.
On Windows we just started shipping the Mesa software rendering version of opengl32.dll, mostly to support remote desktop and virtual machines. The performance impact has been negligible for us, the initial rendering lags and then its fine. I suppose we could do something similar on Linux as well, but it would be nice to use hardware acceleration if it works and only fallback on software rendering. (And eventually we'd like to figure that out on Windows as well). Anyone solve this situation? Should we just ship our own libGL.so.1 working with Mesa's swrast? Deploying software-rendering on Linux looks more complicated than Windows. Thanks, Ian _______________________________________________ Interest mailing list [email protected] http://lists.qt-project.org/mailman/listinfo/interest
