All things related to multicast

["Lawrence A. Rowe" <>]

Hi -

Wow - I didn't expect to get messages about ancient history (Parallax
Boards) and modern problems (mpeg encoders on Windows).

re: parallax boards -- we had these boards too.  I met one of the
co-designers of the board and talks extensively with the company.  They
were always extremely frustrating because the company never really
solved the problems we had.  I recall Toerless's software fondly.

(fyi - the boards we're using are doing 4:2:2 capture of fields - that
is it actually encoding 60 fields a second.)

re: mpeg2 boards -- we've worked with several of these too.  The most
recent was a board from vision-tech.  We got it to send UDP
encapsulated  data to another host where a decoder board successfully
decoded the result.  The problem in both cases was that boards only ran
on Windows.

The application we are working on is building the infrastructure for
interactive television and webcasting that is remotely managed and
controlled by software automation.  To that end, we put computers in
various locations around campus - classrooms, broadcast centers near a/v
routing switches, and closets.  We want to manage the machines and
launch programs remotely.  While this is certainly possible with Windows
systems, we have found fBSD and Linux boxes much easier to deal with -
and not surprisingly more reliable (e.g., I still have to reboot the WNT
machine on my desktop once a week - something we never had to do with
Unix desktops).  The other problem is that Windows software often does
not have a command line interface so we're left trying to drive programs
remotely using combinations of vnc/timbuktu.  

As Markus says - "...maybe someone will do a Linux driver for board X."
I tried that too with several boards/companies.  This never worked very
well because the company really didn't care whether we finished the work
or not and wouldn't support our effort (e.g., you really need help from
the hardware folks to understand what's happening on the board).

Some reasonable-priced MPEG2 encoder boards for Linux are coming - and I
think they'll be supporting RTP which will be good - but it may still be
a while.

My offer/queston still stands -- anyone want to do some high bit rate
sreaming video experiments using unicast and multicast RTP transport?
        Larry
p.s. If I can get anyone to setup the appropriate system, we'll produce
a live feed of the Berkeley MIG Seminar this fall using this technology.
-- 
Professor Lawrence A. Rowe          Internet:  

Computer Science Division - EECS       Phone: 510-642-5117
University of California, Berkeley       Fax: 510-642-5615
Berkeley, CA 94720-1776            URL: http://bmrc.berkeley.edu/~larry