[linux-dvb] bt8xx problems - a call to all PCTVSAT users and bt8xx maintainers

Uwe Bugla uwe.bugla at gmx.de
Wed Mar 22 14:59:43 CET 2006


Hi everybody,
Yesterday I tried the new official kernel 2.6.16 in connection with the
latest mercurial tree.

As a first impression everything looked very fine so far. So I'd like to
send a big "Thank you" to Manu Abraham, Edgar Toernig, Johannes Stezenbach,
Sigmund Augdal Helberg and many other contributors.

But unfortunately there are still critical issues about this solution:

1. I am not sure whether people like Klaus Knopper do merge their actual
kernel tree with the mercurial tree before publishing their latest Live-CD
or Live-DVD material. In a couple of days Knoppix 5.0 will be published
containing kernel 2.6.15. So at least watching DVB TV / Radio with a live CD
/ DVD will be impossible as far as bt8xx cards are in use. Isn't that a
pity!

Question 1: What is the exact time interval in which the official kernel is
merged with the mercurial one?

2. The driver solution at least for the PCTVSAT card is still not
satisfying:

After watching TV for more than one hour without changing the channel at
least I suffer an application breakdown in three steps:
a. for about half a minute audio / video synchronization suffers
b. audio playback breaks down for a couple of seconds
c. the playback window dies

Now if I talk about an application breakdown I'd like to confirm that the
nature of the error cannot be application specific at all:
In how far?
No matter if I use the latest xawtv-4.0 snapshot or the latest MPlayer with
its own builtin dvbstream engine I get the same application breakdown after
a certain amount of time. I do not use Mplayer with the cache parameter as
that hasn't been necessary any more for a very long time.
AND:
I've tried Edgar's patch collection that I published here a couple of weeks
ago.
I've also tried the latest mercurial tree as I stated before.
I swear I never experienced this kind of problems with kernels <= 2.6.13.

Question 2: Are there any PCTVSAT users around having experienced the same
problems?
Be aware: It takes time to find that out - the error does not appear after a
couple of minutes. It needs testing for some hours for this error to appear
at all.

If somebody needs debug printouts etc. to resolve that issue please tell me.

Perhaps Manu Abraham was right in calling Edgar's code cleanups a
regression. I still do not know, but I guess to find that out we all are
here for.
Hope this message will not be ignored after all those quarrels and
fightings.
Fact is that above all quarrels and fightings that happened around this damn
bt8xx issue there meanwhile are two kernel versions out that do not supply
support for bt8xx cards without additional material.
This is anything else but a real advantage neither for Linux users nor for
the Live-CD or Live-DVD concept.
And even if your material is installed on a harddisk drive you never can be
sure whether your long term recording material will be consistent at
all...............
It is a real pity and I still do hope that this exhausting issue can be
resolved soon in a quality being satisfactory for everybody. At least I have
not been passive in the meantime. I cleaned all documentaries in the dvb
tree that I thought to be insufficient and / or full of errors and it makes
me very happy that all those cleanups are at least part of the mercurial
tree now. So where is the effort of people being more capable in C
programming that I am?
I am still waiting for a stable solution because I am tired of those goddamn
breakdowns of whatever kind. And I'm sure I am not the only one........

Cheers

-- 
"Feel free" mit GMX FreeMail!
Monat für Monat 10 FreeSMS inklusive! http://www.gmx.net



More information about the linux-dvb mailing list