Re: [Discuss-gnuradio] s/Eric/Tom/g

[John Gilmore]

> As some of you know, I've been involved with GNU Radio for a long
> time.  The idea that became GNU Radio started as a conversation over
> dinner in San Francisco with John Gilmore, something like 10 years
> ago.

As one of the guys present at that dinner in early 2001, let me
suggest that Eric has done an incredible job picking up that idea and
running with it for a decade.

We saw that commercial companies were using digital signal processing
to radically simplify and improve their products, but that the free
software world hadn't learned those lessons.  That meant there was
a real opportunity hanging wide-open.  Eric jumped on it.

Part of the deal was that I'd pay his salary for the first year or
two, because I knew you can't really get much public support or
financial support for a free software project until it can actually do
some useful job.  Eric spent the first year learning modern signal
processing, surveying existing hardware and existing free software,
then settled on MIT's "spectra/pspectra" code base as a good place to
start hacking.  After the first few years, he found enough academic
and commercial support for GNU Radio that I didn't need to pay him
full time -- and he weaned himself fully off my support shortly
afterward.

Matt Ettus was an early volunteer who also saw the real-world promise
in free signal processing software.  We had reasonable software, but
the available high speed A/D and D/A hardware cost thousands of
dollars and was pretty lame.  Matt finished his job designing
Bluetooth circuitry, and then risked everything to design and build
what became the USRP.  With Eric's help, he built up from nothing to a one-man
design/procurement/manufacturing/stocking/shipping/sales/customer-support/programming
shop, which over the years has matured into the thriving and valuable
business it is today.

Jay Lepreau was another early contributor who saw how GNU Radio could
enable active academic research into cognitive radios.  Jay brought us
into his lab at the University of Utah, encouraging researchers at
dozens of other institutions to design their experiments on GNU Radio
and the USRP.  He brought us into the academic funding that
significantly matured GNU Radio's ability to do packet-based
communication.  Jay died in 2008 but his contributions live on in this
community.

Along the way we took a few detours into application areas that tested
and honed GNU Radio's strengths.  While Hollywood was trying to force
the FCC to outlaw TV receivers that could receive free over-the-air
digital TV signals (because they'd forgotten to put DRM into them),
Eric and a small team successfully implemented an HDTV receiver using
old PCI-bus digitizer boards and GNU Radio.  Hollywood's engineers
said it couldn't be done, and we knew they were liars, so we did it.
Indeed, it ran 30x slower than realtime on a dual Athlon motherboard.
But it ran, decoded actual TV signals, and proved to the regulators
and to the standards committee that you'd have to not only outlaw
hardware demodulators, but also software -- which EFF had recently
proven to a Federal appeals court was a violation of the First
Amendment.  The fucking bastards at the FCC passed the regulation anyway
(https://secure.wikimedia.org/wikipedia/en/wiki/Broadcast_Flag), but
the American Library Association and Public Knowledge litigated in
federal court, proved that the FCC had no authority to regulate what
receivers do with their signals after reception, and the rule was
struck down.  This HDTV demodulator code is *still* not running in the
latest version of GNU Radio, but I hope someone will work out the
kinks.  Modern hardware should be able to do it in realtime.

A second big attempted application area was passive radar.  We read
that the US Army's favorite tactic when invading somewhere was to blow
up all the TV and radio stations because it's easy to track airplanes
by watching their signals bounce off the planes.  It works with
cellphone tower signals, too.  Eric spent several years researching
the topic, writing GNU Radio code, and designing antennas and
hardware.  Ultimately none of it worked reliably; it took more dynamic
range (or custom differential hardware) than we had, but we learned a
lot and have made it easier for future generations to do this as the
hardware improves.

Eric, it's been a great decade, and I'm looking forward to the next
big trouble you get into!

        John