> I have been contemplating doing a Video for Linux port for some time but > am dubious about what benefits this confers, since the (wavelet) video > content will be meaningless to standard VfL apps. It seems to me that as > Linux does not a common multimedia architecture then wavelet support > would have to be added on a per-application basis. V4L2 tries to extend things a bit further but the basic framework idea was to make it trivial for a driver to be video4linux compatible. > I'd appreciate any comments. Having lurked on this group for ages, this > is the first time I have seen ADV601/wavelet mentioned. Would there be > any interest in VFL ADV601/wavelet support? If the code is open source (except the firmware obviously) I'd love to see it in the base kernel. You are correct that V4L doesn't offer a great deal but its designed to handle such devices. Other than the basic few ioctls which can return static data (VIDIOCGCAP, VIDIOCGWIN/SWIN) you don't actually need to implement the rest if your card simply does "read frame". I guess you have VIDIOGPICT/SPICT (set/get brightness, hue etc) For ADV601 wavelet format we would also need to allocate a new format type for it but thats also no big deal. In short, I tried to make sure V4L was trivial to plug into existing drivers because that is exactly how it started.