Closed russel closed 6 years ago
This seems to be mostly "not our problem" if it has to be fixed in the repositories. IMHO you can close this issue once you fixed the README.
Or switch to soapysdr which would enable dabtools to use many more devices than just the rtlsdr.
Hummm… I hadn't even thought of that route, I had thought it was /dev/swradioX or rtlsdr, but then for my hardware it is (I guess I should get an AirSpy Mini). I wonder if soapysdr stand is an analogous position to libdvbv5 for Linux DVB and is therefore very much the right way for application level code to go. I am guessing we are rapidly discussing our way to a dab2eti rewrite from scratch.
Switching to an abstraction like soapysdr would be absolutely preferable; I would really like to use an HackRF or AirSpy Mini instead of the deaf RTL stick.
I suspect then that means C++11, since I can't find any Rust bindings that are not vapourware, and I haven't been able to try DStep on the API to create a D binding. I am assuming we do not want to work in Python.
Just to note though that whilst Debian packages soapysdr, Fedora does not.
I'm also in favour of using SoapySDR. I'm fine with both keeping librtlsdr or replacing it.
Just to add, this seems like a decision made, which is fine, and that there is a Rust binding for SoapySDR.
It has just been pointed out to me that librtlsdr is actively maintained. Although everyone looks to the OsMoCom website and repository, or even the https://github.com/steve-m/librtlsdr mirror of it, these have been unchanged for a long while. Is seems though there is https://github.com/librtlsdr/librtlsdr which is actively maintained and releasing regularly.
I intend to post bug reports to Debian and Fedora asking them to switch from the OsMoCom repository to this one, and to update their packages.
We should make a decision what to say in our README.
PS the link in our README is wrong anyway so a change of some sort is necessary.