[openal] 1.16.0 info bug?
chris.kcat at gmail.com
Wed Aug 27 19:10:46 EDT 2014
On 08/27/2014 11:49 AM, LAM wrote:
> Also note in the output that it turns
> out the identifier code is exactly the same for the dupe, if that helps.
Indeed. Odd that it works with 1.15.1 then, since the problem is that
Windows is returning a different IMMDevice pointer for the default
device query. This should actually be causing a buffer overrun in 1.15.1
since it only allocated enough space for the number of reported devices,
while the default would get added twice.
So then I wonder what's the correct way to test if two IMMDevices refer
to the same endpoint. The device ID is a stringified UID of some kind,
which theoretically can be equal even if a case-insensitive compare
fails. MSDN also says apps should treat the ID string as 'opaque', and
not try to parse it since it can change between systems.
More information about the openal