I was intrigued by this ATC post as well. This idea is not new, of course
-- back in the days of storing programs on casette tape, there were some
radio stations which would broadcast programs at specific early-morning
hours -- if you had a programmable radio, or a timer, you could record the
program onto casette and then have a moderate chance of downloading the
program into your computer later. As you might expect, broadcast noise,
read-in problems, and lack of advertising led to the demise of this type of
byte broadcast.
So, with this early byte-broadcast failure sitting in the back of my mind,
and even though I agree with Gurley, I'll take a contrarian view and argue
why air broadcast bits will never take off. Let the ensuing debate resolve
the issues... :-)
To begin with, there are major regulatory and standards issues to contend
with. To reduce the cost of receivers down to negligible levels, there
needs to be standardization of broadcast frequency and broadcast format.
While the data broadcast format could be declared by fiat by the first
entrant into the field, getting the necessary frequencies could be a
difficult process. Yes, you could use spare HDTV frequencies, but this
would require a configuration step by the user (or a frequency-scanner in
hardware, an added expense), a step in the wrong direction (right away
you'll lose part of your audience -- remember, some people can't even use a
radio). Since US HDTV frequencies are unlikely to be the same as European
or Japanese frequencies, the receiver would need to be configured for a
specific market or geographic region. Quite a bummer when the downloadable
bit stream turns off when you go to conferences in Europe...
To work really well, I would vote for a satellite radio approach. This
would allow the receivers in each computer to be standardized, would work
the same across national boundaries, and might allow higher bandwidths than
spare HDTV channels. Plus, control over this satellite network would be
much more centralized than a confederation of broadcast stations (which
need individual tech. support, and individual contracts). Unfortunately,
this would require a much larger capital investment that the HDTV approach,
and getting the frequencies would be a problem. Entrenched broadcasting
interests might oppose this approach as well.
The other concern about broadcast bits is I'm not convinced there are
enough people who want this capability, but are not interested in (and
quite willing to pay for) a richer interactive experience like Teledesic,
which offers most of the benefits (data anywhere) and adds some compelling
extras (on-demand data, interactivity) on top of data broadcast.
Also, it would take a long time to develop a critical mass of people with
computers who are subscribed to data-by-wire, indicating a long period of
investment by early content providers before they see a good ROI. Another
major bummer is ratings -- with this system, how can an advertiser tell
their advertisement has been viewed? This has been a major issue for the
web, and should continue to be a major issue for other electronic media.
So, despite the hype, broadcast data has some drawbacks:
- lack of existing data broadcast standards (I could be wrong on this one)
- lack of standard frequencies across national boundaries
- need for large initial capital investment, with long ROI periods
- inherent difficulty in getting ratings information will make it difficult
to assess effectiveness of the new media during its critical adoption phase
These are not insurmountable issues, but they do make data broadcast look
much less inevitable than Gurley's hype-fest.
- Jim