From: Vladimir Dergachev <volodya@mi...> - 2004-01-03 19:08:41
First of all I would like to wish a Happy New Year to GATOS users and
Secondly, there are several new releases (http://gatos.sf.net/) :
* XFree86 4.3.0 drivers , experimental 14 - this version fixes an
old bug in rage128 drivers that resulted in bitmap corruption upon
exit from AVview.
* km 0.5.0 - this is the latest code from CVS, which has been stable
enough on my main computer for a while now. I do see occasional lockups
when using CC (VBI), however the workaround (for now) is to make a
snapshot before enabling CC. For some reason km is stable after this.
For developers, this version separates km_v4l from hardware-specific
code in km and all interaction is done via km_api.
* AVview 0.80.2 - AVview has switched to using ffmpeg-0.4.8 for encoding.
Also, the capture interface was reworked to use "encoding themes" -
combination of encoding and capture parameters - that should make it
easier to switch between different formats. There are a number of
predefined themes available.
Let me know if you find predefined themes inadequate for some reason.
In particular, I am not an expert with ffmpeg settings, maybe there are
better combinations of parameters..
Thirdly I would like to make some comments on the state of GATOS
development in general:
* Federico Ulivi has done some wonderful work on new tv_output branch
of GATOS CVS. As I understand it should provide support for TV-out
on Radeon cards - without using BIOS calls as much older devel branch
* The big jump in version numbers of AVview came from my realization
that most hardware features I was aiming for are implemented
(capture, identification between XFree86 screens and v4l devices,
CC/VBI support) and it is now time to focus on making it more
user-friendly. In other words, AVview is not just a test app
* The bug in rage128 was due to chunk of code that went missing during
transition from XFree86 4.2.0 to XFre86 4.3.0. This should really have
been fixed much earlier.
Therefore, I would like to admit that mach64 and rage128 cards are
This does not mean that identified bugs will not be fixed - as once
the problem is known it usually takes just several minutes to fix.
Rather, this is an acknowledgement of status quo - that despite my
wishes or attempts I plain simply do not have time to perform
systematic testing of older chipsets.
In hindsight, I should have seen this earlier, as simply listing
different areas of GATOS makes obvious that I can't possibly test
and hunt down bugs for them all:
AVview TV playback, AVview capture, AVview CC, ati.2 2d for three
chipsets (mach64, rage128 and radeon), ati.2 TV playback for three
chipsets, ati.2 3d for plain radeon and r200 (8500+ cards), km_api,
km_v4l, km hardware drivers three chipsets.