[mythtv] Channel changing Bugs...

[Tony Clark]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sunday 13 April 2003 08.38, Andrew M. Bishop wrote:
> Andy Davidoff <dert at pobox.com> writes:
> > On Apr 12, at 17:50, Leandro Dardini encoded a 1.0K recording:
> > > I am using mythtv from about three months and still I am discovering
> > > new features. I live in Europe too, and I use the chanid
> > > ordering. Is there a way to select a channel by its chanid with the
> > > remote? I'd like to press 1 and watch chanid 1.
> >
> > This is the original problem, AFAICT from the month or two I've been
> > following/working-on tuning.
> >
> > Why don't we rename or at least divorce chanid from its current
> > role and use it as the second layer of indirection that the folks
> > in Europe have been asking for.  I realize that chanid is currently
> > being used in this capacity (sort of) by some people.  I'm suggesting
> > that we make this the only behavior world-wide, for consistency and
> > simplicity.

This idea sort of makes sence as what chanid really is, is tuner setting 
information.  But I guess renaming it in the database may break many things.  
It's internal name doesn't really matter from the users perspective.  He just 
needs to be clearly told what information needs to be put in it during setup 
time.

> >
> > We'll sort on chanid and take inputs from the remote against the
> > values of that column.  For N. America, this means chanid == channum
> > by default.  For Europe, chanid == sort-order-of-channels.  The chanid
> > values would be arbitrary, so everyone could renumber them to values
> > that are more familiar or easier to input via remote.

Whats the point of indexing on chanid anyway.  At least for me, there is no 
relationshit between chanid and how I want to scroll through the channels.  
Scrolling by channel slot number makes sence. ie 1 2 3 4 5.  Duringing setup 
I simply assign my chanid to my channel slot number.  ie 1 = E11, 2 = E3, 3 = 
SE11, 4 = E12, 5 = E10.  Now the 3 stations I want to watch most are 1 2 3.  
This works for everybody and the default for North America is 1 =1, 2 = 2 
etc.  They can change the order during setup if they wish, by running 
mythfilldatabase --manual.  I actually don't program in all the channels 
available to me as some of them I will never watch.  ie. wrong language, 
religion etc.

> >
> > It seems to me that this simplifies the code and gives everyone new
> > functionality without breaking anything.
>
> Can you explain clearly what the problem is that you are solving by
> doing this.  The description above seems only to describe two
> different sorting methods and ways for entering channel "number" for
> live TV; either by channum or by chanid.  This can surely be handled
> by a configurable setting to use one or the other.  In fact there is
> already a setting so that pressing the up and down keys go in chanid
> or channum order.  There seems only to be missing the ability to type
> in a chanid instead of channum.

Sorting by channel slot number (channum), station ID and maybe program types, 
ie sport, sex, comody, sex etc make more sence than what information is sent 
to the tuner.

> In most cases chanid == channum anyway because that is what
> mythfilldatabase does when adding new channels.  I thought that the
> only exception to this was by running mythfilldatabase --manual.  In
> fact running mythfilldatabase --manual allows everybody to change
> chanid values now (if they clear existing channels from the database
> first).
I suspect "chanid == channum" is probably only true for North America and a 
few places that tend to follow NA.  It is certainly not the case in Western 
Europe and I think Australia also.

>
> The solution is therefore to use chanid for all sorting everywhere.
> Everybody that sorts by channum and doesn't have channum == chanid
> must clear their databases and re-run mythfilldatabase --manual. :-)
The solution is to never use chanid to sort on, use channum, station id maybe 
or program catogories + channum maybe or user perferred + channum.  I don't 
this there is program catogories or user preffered information in the 
database yet. 

Tony

- -- 
Contract ASIC and FPGA design.
Telephone +46 702 894 667
http://pgp.mit.edu:11371/pks/lookup?op=vindex&search=0x633E2623

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iQCVAwUBPppTv2kjRGZD9pCjAQImAAP9HEiXjsNs+DOCuCXBVLIuglzs3bhbi8CK
PVPjsA85nwwrkHonJS9bUBRlIDQdV/7i274BfVeyEgLHQmxkhC6LkaW30nWBGBzH
b7UPpO3bgSP7BP/EvDEDDuS/W66QaXF5RTDSfTXckwRYSn1zEuXfC0FWrJ/N2r3F
m2xX/CmW2pw=
=8WZm
-----END PGP SIGNATURE-----