51 lines
1.7 KiB
Plaintext
51 lines
1.7 KiB
Plaintext
libv4l todo:
|
|
------------
|
|
|
|
-don't automatically enable the software effects (awb, gamma) on cameras
|
|
which need flipping, most of these only produce yuv data and software
|
|
effects are very expensive on yuv data
|
|
|
|
-move pixart rotate 90 hack to v4lconvert_decode_jpeg_tinyjpeg, since it
|
|
is only needed on select pixart cameras, which use this function for
|
|
decoding, this will nicely cleanup the main conversion routine
|
|
|
|
-move usb info gathering code from libv4lcontrol into a new
|
|
libv4l deviceinfo (devicequirks?) sublib, since it really does not have
|
|
all that much to do with the emulated controls
|
|
|
|
-make dmi + usb matching more generic, with wildcards support, etc. To allow
|
|
for a more flexible upside down table where one entry can match multiple
|
|
models, hopefully lowering maintainance costs of the upside down device
|
|
list a bit
|
|
|
|
-some sort of generic rotation framework (for tablets)?
|
|
|
|
-rewrite video effects code to be even more plugin based
|
|
|
|
-add code for software auto focus
|
|
|
|
-libv4lconvert: v4lconvert_do_try_format should always prefer smaller then
|
|
requested resolutions over bigger then requested ones
|
|
|
|
-add support for setting / getting the number of read buffers
|
|
|
|
-add code to v4l2_read to not return frames more then say 5 seconds old
|
|
|
|
-take the possibility of pitch != width into account everywhere
|
|
|
|
-make updating of parameters happen based on time elapsed rather then
|
|
frames
|
|
|
|
-get standardized CID for AUTOGAIN_TARGET upstream and switch to that
|
|
|
|
Nice to have:
|
|
|
|
-support packed yuv as output format so that we aren't forced to do
|
|
conversion on upside down uvc cams
|
|
|
|
-add a software rotation control (0 / 90 / 180 / 270 degrees) for tablets.
|
|
|
|
|
|
utils todo:
|
|
-----------
|