Re: I-D ACTION:draft-zigmond-tv-url-03.txt

   Some fundamental limitations identified as early as last summer have still
   not been fixed, this draft should not have been published until these issues
   have been resolved.

   1) This proposal only supports the broadcast model, it provides no support
   for the increasingly important on-demand model; a workable TV URI scheme
   requires support for BOTH content and broadcast addressing.

This functionality is addressed by the associated lid: URI scheme and
(should be (:-)) described in the assciated "guide to using this stuff" RFC.

   The removal of channel numbers is unjustified, it is not obsolete.  Most DTV
   consumers select by channel number either directly or channel  +/-, many do
   not even use menus, very few will ever use a URI.  Devices using this scheme
   are aimed at end consumers where usability is perhaps the more important
   issue.

Channel numbers are not unique identifiers: what is on channel 4 here
is not the same as what is on channel 4 in New York.

   >The channel numbers generally correspond to tuning frequencies in the
   various national broadcast frequency standards; for example, "tv:4" in the
   United states would be found at 66 MHz.*

   This is supposed to be an international standard, not a US standard.  If
   channel numbering in the US is fixed to a specific radio frequency, (which I
   find difficult to believe) channel numbering should be included on a
   "should" or "may" basis.  Current limitations should not be reason to
   cripple this standard.  We should be aim for an ideal.

For NTSC (analog) TV in the US, channel number to frequency mappings
are fixed.

However, the reason for omitting the usage was not the mapping, but
the fact that channel number are not unique identifiers.

Craig

Received on Tuesday, 11 January 2000 10:38:34 UTC