Thanks for testing.
However seeing all the problems this program gives, I am thinking about switching to the method that just picks the desktop resolution.
Cases where the current method selects an unlikely resolution are probably related to my way of comparing aspect ratios (floating point errors) but even then it shouldn't result in crashes and lock-ups; this seems to be caused by the selected resolution not being supported after all.
Adding a list of well-known resolutions could partly solve this (and the floating point problem) but I would never be 100% sure of it functioning correctly on all machines, unlike picking the desktop resolution which is known to be supported by the monitor. (I am just thinking aloud a bit. ;))
I like the idea of letting the user choose the resolution and waiting about 10 seconds before switching back to their desktop resolution if "is this a valid resolution?" question is not answered.
Yes that would be a good idea.However seeing all the problems this program gives, I am thinking about switching to the method that just picks the desktop resolution.
Cases where the current method selects an unlikely resolution are probably related to my way of comparing aspect ratios (floating point errors) but even then it shouldn't result in crashes and lock-ups; this seems to be caused by the selected resolution not being supported after all.
Adding a list of well-known resolutions could partly solve this (and the floating point problem) but I would never be 100% sure of it functioning correctly on all machines, unlike picking the desktop resolution which is known to be supported by the monitor. (I am just thinking aloud a bit. ;))