Re: [Discuss-gnuradio] Why Isn't GNU Radio Used More?

[Michael Dickens]

On May 9, 2011, at 1:22 PM, Marcus D. Leech wrote:
> this is, in fact *software* defined radio.  So why is it always a big 
> surprise when hardware types encounter an SDR platform and become 
> more-than-vaguely-queasy at the though of having to, perhaps, learn a little 
> bit about software.

If I got paid a dime for each time someone in a meeting made the comment "Why 
is there so much hardware in -software- defined radio?" I'd be wealthy!

I think your points about SDR being the "perfect storm" by requiring both 
software and hardware knowledge is quite true -- and, because of this joint 
complexity, one reason why folks aren't adopting it in droves.  As you say: 
there aren't too many practitioners out there who straddle the fence acceptably 
well at this point.  I'm a software guy; I can integrate software with 
hardware, but I don't do hardware beyond knowing specs and writing the software 
for integrating it.  I have no desire to be a hardware guy, though I know that 
understanding hardware limitations is important for DSP and SDR in particular 
where "real time" means something.

One cannot have every possible block available in GR/C; but, if enough blocks 
are available to do 90+% of what common users need, then that's good enough for 
the GR/C release.  And, then much like MATLAB / Octave, there needs to be a way 
to install new blocks easily from an archive or local compile -- I really don't 
know if that's the case right now.

In my thinking, what blocks to include comes down to what end-users actually 
-need- to do, versus what we engineer / developers think they need to do; there 
is often a substantial gap between these areas.  In order to attract "the 
masses", this gap needs to be reduced / closed. - MLD