Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> espoused:
> __/ [ Mark Kent ] on Friday 04 May 2007 20:04 \__
>
>> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> espoused:
>>> Online TV viewing 'on the rise'
>>>
>>> ,----[ Quote ]
>>>| Almost half of European broadband users are using their computers to
>>>| watch television online, a survey claims.
>>> `----
>>>
>>> http://news.bbc.co.uk/1/hi/business/6617013.stm
>>>
>>> Internet TV may pose threat to Cable Companies by Decade's End.
>>>
>>> ,----[ Quote ]
>>>| The Cable Version of "Wireless Conversion" is going to happen. We just
>>>| need a name for it.
>>> `----
>>>
>>> http://pulverblog.pulver.com/archives/006905.html
>>
>> Until 3-mode networking is grasped, this is not going to work. Consider
>> how much bandwidth you need to broadcast a football match by satellite,
>> and then how much to do it by wired network, even *with* 3-mode
>> networking.
>>
>> I can see some types of programming being web-based, as they're a little
>> off-beat, but mainstream material will have to go radio, as this is the
>> singular area where radio excels over wired connections - it uses
>> /exactly/ the same amount of spectrum to broadcast to a million as it
>> does to one person. It scales. The cost doesn't rise for the extra
>> people.
>>
>> I find it hard to understand why so many people seem to find this so
>> hard to grasp...
>
> Can network capacity and implementation be extended (in the long term) to
> accommodate the need? Our network ground to a halt before we started
> disconnecting people who watched TV over the Internet. With capacity comes
> greed to utilise it. Limits are indeed needed.
>
It's not just an issue of capacity, rather, an understanding that
connectionless packet-switching does not carry streams reliably;
basically, you need to separate the streams, and manage their bandwidth
along with everything else. Voice streams use relatively low bandwidth,
and even compressed video isn't so bad, the problem, however, for
current packet-switching network designs is two-fold - that every viewer
is another stream to the server (no economic scaling), and that the
brain is part of the connection (eyes/ears & brain) and cannot tolerate
jitter in the stream (beyond certain limits, anyway). This means
that you need to be able to isolate & protect these time-critical
streams from normal data which generally doesn't care about maintaining
its temporal relationship that much.
Telco networks in their TDM form carry voice and video something like 10
times /more efficiently/ than IP networks, and with guaranteed temporal
relationship between information instances; unfortunately, these
networks are expensive to build, and nobody wants to pay for phone
calls any more, so they'll have to pay for the massive increase in
bandwidth to get their "free" voice and video... funny old world! And
even then, we will still need connection-oriented packet switching in
order to reliably move streams anyway. See the wiki page on PBB-TE for
more info.
--
| Mark Kent -- mark at ellandroad dot demon dot co dot uk |
| Cola faq: http://www.faqs.org/faqs/linux/advocacy/faq-and-primer/ |
| Cola trolls: http://colatrolls.blogspot.com/ |
|
|
