[mythtv] BIG DVB PATCH V3.5 - Logical Channel Numbers

Tim Davies tim at opensystems.net.au
Sun Jan 9 19:00:36 EST 2005


John,

Okay, you can't get to the specs directly but you can see the drafts!

Try:

https://committees.standards.org.au/COMMITTEES/CT-002/N0004/CT-002-N0004.PDF
(Appendix 11)

I also found this:

http://203.147.163.200/documents/OP_41_(DTG_Logical_Channel_Descriptor)_June
2001.pdf

I just looked at scan.c in dvb-apps and the descriptor is identical.  Ahh,
an hour of coding wasted!  

Mapping network_ids in Oz would be a nightmare, the only thing feasible
would be to use the original_network_id.  It still isn't ideal, as networks
change every now and then.

I'd suggest sticking a checkbox on the scanning screen to say "Grab UK/Aus
Logical Channel Numbers" and default it to off.

Anyway, feel free to butcher the patch.  Especially the way it assigns
channel numbers when there is no LCN for the serviceid (but LCN data is
present).

Tim.


-----Original Message-----
From: mythtv-dev-bounces at mythtv.org [mailto:mythtv-dev-bounces at mythtv.org]
On Behalf Of John Pullan
Sent: Monday, 10 January 2005 3:42 AM
To: Development of mythtv
Subject: RE: [mythtv] BIG DVB PATCH V3.5 - Logical Channel Numbers

On Sun, 2005-01-09 at 23:23 +0800, Tim Davies wrote:
> John,
> 
> Not that I know *anything* about the Europe/UK LCNs, but to paraphrase
some
> pertinent parts of the Australian spec (yawn!):

Do you have a link to this spec.

> - The LCN is found in the second descriptor loop of the NIT with tag 0x83.
> I think it is a reserved id, but that's good isn't it?  Reserved for
Aussie
> LCNs ;-)
> 
> - Unlike Europe, Australia uses non contiguous LCNs and three digit LCNs.
> Are we using 10 bits, and you guys use 8 bits?

I think we also use 10 bits, at least this is what I gleaned looking at
scan.c in dvb-apps.

> 
> - There are some rules about what to do with a service that doesn't have
an
> LCN.  This bit of the code might need to be un-Australianised.  I haven't
> yet seen a service here without an LCN anyway...  
> 
> - European standards say that services with the same original_network_id /
> service_id shall have the same LCN.  That way, regional variants of a
> service can have the same LCN.  Not so here!  In Australia the service is
> uniquely identified by original_network_id / transport_stream_id /
> service_id.  There are several networks (approx ten?) in Oz with unique
> original_network_ids.  Each transmitter has its own network_id, it seems.
>
> Not that most of it really matters too much.  I tried to write the patch
so
> that if you get the 0x83 descriptor, it populates the LCN data without
> stomping on another descriptor's code that can also populate it.  Assuming
> of course, any other LCN descriptor just populates a service_id to LCN
> mapping.
> 
> It only falls down if the 0x83 descriptor is used for something else
> somewhere.
> 

That's my point really. I was quite amused/worried when I saw your patch
and realised it was doing the same thing as the UK version of the code.
We intend having a table of network ids to which we can map special
descriptor handling. Hence my request for your network ids.


-- 
John Pullan <jmp at tarantella.com>




More information about the mythtv-dev mailing list