Some of the best resources I've seen in that regard are nand2tetris and make-a-lisp.
The SDR community seems to mainly consist of people interested in using existing software and focusing on the received signals. There's still a lot to be discovered if more people start writing software to decode the signals.
You might be interested in the GNU Radio project, a modular suite of tools for building software receivers.
You may also be interested in WSJT (ultra-efficient data modes) and FreeDV (ultra-low-bandwidth voice). A number of people in the amateur radio community are working on modern modulation and codec systems.
As usual in this business, setting up the development environment will be 90% of the hassle. But it looks like it'd be well worthwhile for those just getting started in the field, especially if you're already comfortable with Python and Linux.
GENI also sounds extremely interesting. Being a Windows-dweller for the most part, I'd never heard of that before.
I was thinking like 3.5 mm from ipod to RCA & then RCA to Raspberry pi Mic. Then processing the mic input for streaming.
I think I'll loose the dual channel & also not sure whether I would be able to control the channels without creating a breakout switch in the line to ipod's 3.5 mm.
Can anyone suggest a better way to do this? There's no Internet Radio in India & smartphones doesn't have FM radios now a days.
Then again, an SDR is overkill if all you'll ever want is FM radio. A dedicated FM receiver will use less power, take less space and hopefully have better usability.
Edited to add: Since this is a digitally controlled tuner, you can also change the station remotely if you'/re using the right setup. The websdr project, http://www.websdr.org/ , is a really nice way to do this and there are plenty of other setups.
HN reader messaged me that there's even Gnome extension for some.
What I bought was https://www.tindie.com/products/microwavemont/dsp-radio-vers... - I'll be hooking this up to a Pi with some kind of audio capture and hopefully modifying the firmware so the frequency can be set via UART.
What I'd really like to know is how hard it would be to hook up the same kind of chip/circuit to another chip that does USB Class Audio so there's no extra capture step to get audio to a Pi.
For more, there's another really nice SDR basics course online you can find at Great Scott Gadgets (https://greatscottgadgets.com/sdr/) associated with the HackRF One open-source SDR hardware. It goes into it a bit with GNU Radio, which is a really incredible project.