5 Jan
2014
5 Jan
'14
10:55 p.m.
Hi,
On 05.01.2014 18:04, Joris van Rantwijk wrote:
If you know that the hardware is going to freak out, I
think returning
-EINVAL is a more useful thing to do.
Thanks for the suggestion, after some further testing I've now pushed
a change that excactly does this.
Normally I expect a function to either do what it
promised or return an
error. Because RTL-SDR is reverse-engineered, it is not fair to judge
it to such high standards. Still, it would be good to document known
limitations of the RTL chip in the library via comments or error codes.
I've also documented the function in rtl-sdr.h now.
That is good to know. So in general I should try to
keep the hardware
sample rate high, and resample in software if needed.
Yes, that indeed is the best approach with those devices.
Regards,
Steve