[mythtv] [mythtv-commits] Ticket #993: FireWire Scientific Atlanta 3250HD channel change problem

[Carl Moore]

On Tue, January 10, 2006 8:56 pm, MythTV wrote:



> #993: FireWire Scientific Atlanta 3250HD channel change problem



>
-----------------------------------------------------+--------------------



> --



> Reporter:  Wylie Swanson   |       Owner:  ijr



> Type:  defect                                   |      Status:  new



> Priority:  major                                    |   Milestone:



> Component:  mythtv                                   |     Version:



> Severity:  medium                                   |



>
-----------------------------------------------------+--------------------



> --



> I have brought up a slave backend with only one encoder connected,
the



> Scientific Atlanta 3250HD over Firewire.



> 



> 



> I am able to successfully stream HD content using test_mpeg2,
however,



> after configuring the slave backend to use the tuner and starting the




> backend, I notice some flawed tuner behaviour.



> 



> I have pre-programmed the tuner to switch to channel 708 (PBS HD,
Cox,



> Phoenix).  When the slave backend starts, it tunes on to channel 708,
as



> expected, but then pauses and changes the channel to 7. Channel 7
cannot be



> streamed, so the tuner is skipped (and rendered not usable).



> 



> --



> Ticket URL: 



> MythTV 



> MythTV_______________________________________________



> mythtv-commits mailing list mythtv-commits at mythtv.org 



> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-commits



> 



> 









I would have addressed this earlier, but
this is the first time anybody has said anything about the SA3250HD since
support was first introduced.



I modified the sa3250ch.c file some 7 months or so ago when I first put my
hd system into production.  It's been running like a champ ever
since.



I assumed I was the only person on the planet (Phoenix maybe), that was
running HD capture with firewire on the SA3250HD, because it was
impossible to tune to channels above 99 with it and nobody seemed to
notice (I guess everybody must be using lirc to change the channels..which
works of course).



Anyway, after some experimenting (I don't know beans about 1394 protocols
or the sa3250 internals for that matter), I discovered that if I sent a
key-up event with the keycodes reversed immediately after the key-press
event, then the channel changes worked perfectly.  Don't ask how I
figured it out, blind luck most likely.  I don't know if it's the
"correct" way to do this, but I don't care...it works.



I've attached a .diff of the sa3250.c file from current svn against my
local copy that has the changes I made to make this work.  If someone
would like to review this and see if it can be incorporated back into
current svn, it would be greatly appreciated.



It appears that my editor of choice (visual slickedit), made some
whitespace changes in the file so it appears as if the whole thing has
changed, when in reality it's mostly the part at the end.



Also, on that note, the backend has similar code built-in (or it used to),
to change the channels, and that is broke as well.  I never modified
the built-in code, I just tell the backend I have an unknown firewire box
and tell it to use the sa3250ch program to change the channels.  But,
fixing the backend code to match should be no problem anyway...







-- 







==============



=Namuch-Noway=



==============
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mythtv.org/pipermail/mythtv-dev/attachments/20060111/79e3a606/attachment.htm
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sa3250ch.c.diff
Type: text/x-patch
Size: 9377 bytes
Desc: not available
Url : http://mythtv.org/pipermail/mythtv-dev/attachments/20060111/79e3a606/sa3250ch.c.bin