On Sat, 2002-03-16 at 05:45, Mark McClelland wrote: > >Comment #2 (the negative one) > > > >What an utter waste of time. > > > >I'm sorry to say this, and I've thought for a while if I even should > >mention this, but this is how I feel about it. <someone who just started v4l2'ification of his driver) that's how I thought about it first. But for more advanced driver features, v4l2 introduces many many new features and better options than v4l1. I know finally understand why we needed so many API extensions for our driver, and better, that we don't need these for v4l2 anymore. (and what is v4l1.1?) > >Yes, but I'd have to make my driver ready for V4L2 (which is only backward > >compatible). And since that isn't in the main kernel yet, we have a > >chicken-and-egg problem... > > Agreed. This is why my ov511 is still a V4L1 driver. With all the talk > of a V4L3 a few months ago, I didn't want to risk writing for any API > that wan't available from kernel.org. There was some talk about v4l3, but I *know* that v4l2 is extendible enough for almost any purposes. If it isn't, you could always add options in the reserved fields. Most of this v4l3 talk actually seems based on Gerd Knorr's comment on his website (http://bytesex.org/) saying "and it looks like even the second try isn't perfect". People who don't understand v4l(2) will directly think that v4l2 "thus" sucks. But v4l2 isn't that bad, it just takes some getting-used-to :-). This sounds a bit zealous, it's not supposed to ;-). Just want to say "it's not gonna be that bad" :-). Ronald -- - .-. - /V\ | Ronald Bultje <rbultje@xxxxxxxxxxxxxxxxxxx> - // \\ | Running: Linux 2.4.18-XFS and OpenBSD 3.0 - /( )\ | http://ronald.bitfreak.net/ - ^^-^^
