It would be nice to be able to probe the opengl state just once and cache the results - this would speedup the client startup.
On Linux, this could be cached by graphics card so that a change in configuration would cause xpra to re-run the opengl probing:
$ lspci | grep VGA 00:02.0 VGA compatible controller: Intel Corporation Skylake GT2 [HD Graphics 520] (rev 07)
Looks like this would involve parsing the output of
lspci which is going to be slow.. Python, In linux obtain VGA specifications via lspci or HAL?
Preferably, this cache should be per-host: Get a unique computer ID in Python on windows and linux. And ideally, we should somehow record the driver version in case a driver update / downgrade breaks things.
Similar to #2349.
Detecting the driver makes this too hard: by the time we get the driver information, we've already done most of the job required for a full opengl device test.
this ticket has been moved to: https://github.com/Xpra-org/xpra/issues/2345