[mythtv-users] Scan for channels not working with 2.6.20 kernel

Michael T. Dean mtdean at thirdcontact.com
Fri Apr 13 12:58:13 UTC 2007


On 04/13/2007 03:17 AM, Jarod Wilson wrote:
> On Friday 13 April 2007 01:40:17 Michael T. Dean wrote:
>   
>> On 04/12/2007 12:08 PM, Jarod Wilson wrote:
>>     
>>> On Monday 26 March 2007 10:42:09 Jarod Wilson wrote:
>>>       
>>>> DVB channel scanning doesn't work for me using Axel's latest 0.20 mythtv
>>>> builds,
>>>>         
>>> This is now fixed in the release-0-20-fixes branch. Turned out to be
>>> related to the us-cable-irc (in my case) frequency tables being out of
>>> whack just enough to throw off the scanner. Of course, I don't know
>>> exactly why it used to work and stopped working, but...
>>>       
>> Yeah.  There must have been another change (perhaps in drivers or
>> something), too, that had some effect.  But, as long as it's working,
>> now...
>>     
>
> Grr. Argh. I suck. Its NOT working. :(
>
> I stupidly had both an rpm install and source install on the box. The source 
> install was a sufficiently old svn trunk build that looked much more like 
> 0.20 than current svn trunk, so I stupidly didn't notice that's what I was 
> scanning with, not the 0.20-fixes build.

Well, that explains why something that didn't seem like it should need 
changing "fixed" things.  :)

>> BTW, Jarod, nice work on the backport...
>>     
> Heh, you think that was something... My working local build basically 
> backports everything from ~11682 to 11702... :)
>   

I'm sure that took some serious work...

>> I verified it a bit ago (not a 
>> lack of trust, but I knew the whole history of the changes and which
>> were most critical, so I wanted to verify nothing got lost).  The only
>> thing that surprised me was that you didn't put in US_MAX_CHAN and
>> EIA_542_FREQUENCY (which were for long-term maintenance, so aren't
>> important) since it meant you had to calculate values yourself.  I would
>> have put them in just so I didn't have to think as hard.  :)
>>     
> Hrm... basically just took the relevant trunk changesets I happened upon and 
> threw them in. Can you point me to what I missed? Wouldn't mind adding that 
> to the mix.

It's really not important--not worth putting in.  You've got the freq's 
correct, so I'd leave it as is.  It was just a couple of macro defines 
for the digital frequency tables (frequencytables.cpp) that were used to 
ensure consistent frequency table sizes in the US and to allow for easy 
changes to the number of channels we support without having to change 
multiple parts of code (in a couple of years we may increase the table 
size as cable companies use higher and higher freq's for their digital 
packages).

Mike


More information about the mythtv-users mailing list