The future of Pylink

Pylink was experimental, to see what could be done in a popular dynamic language. It was also a test bed for trying things out. One of the things I have been trying out is Ice (http://www.zeroc.com/). This has reached the point where it looks a very good approach to building a true plug and play bus. The pylink component implementations are now working completely with Ice, no Pyro and no net-jack, shipping both audio and control signals over Ice using point-to-point connections and the IceStorm service for multi-point event services.

This project will remain for a while as it may be of interest. It will not be further enhanced in this form. The project continues in its new form at http://code.google.com/p/acorn-sdr/.

(current version 2.1)

Pylink-SR Version 1.0

My thought was this - how simple could a functioning SDR be?

There is no doubt that the state of the art moves ever forward and as a result complexity increases.

I have tried to apply the 'what can be achieved with simple equipment' philosophy to software. So, simple and functional means something that works, can be tinkered with by a lot more people than could ever hope to get inside the very good, but very complex systems that have evolved over the last 5 years or so, and can be put on-air (all in good time) and make contacts. I would love to be able to hook this up to the simplest possible hardware, but maybe SoftRock already got there (or could it be even simpler and easily home brewed?).

The V0.1 Version will do RX SSB and AM with the usual selection of filters and has AGC. It will drive the SDR1000 hardware and I'm sure a SoftRock version won't be far away.

This project is 100% Python, yes, even the DSP. The DSP is not fully sorted yet. AM is a little strange and the AGC could do with more work. There is more info on getting up and running, and about the design in the Wiki pages. Yes, there is much to do but Release Early, Release Often is not a bad way to proceed.

I have nicknamed this system JunkBox SDR because I pulled virtually all the pieces out of other efforts in terms of ideas and actual software I had lying around. So it was equivalent to building your QRP rig from stuff in your junk box. It's official name is however PyLink-SR as it shared my head with ErLink-SR which is a more full featured SDR based on Erlang. That one has not reached sufficient maturity to be put up for public scrutiny yet.

Pylink-SR Version 2.0

I was pretty happy with the first attempt so I've pushed on further and its still looking good. This is experimental. If you want a complete system then don't come here - but if you want to tinker then I hope you get some pleasure out of these efforts.

This has been on the stocks for a while. It's given me a pretty hard time but its 95% meeting my objectives now so it's probably reasonable to publish. Don't expect perfection, there are still things to address.

This release concentrates on expanding the capability but keeping it in Python and keeping it simple. There is no enhanced functionality, that will be in the next release, what it does is add transparent distribution and database functionality. You may rightly say, doesn't that add a lot of complexity? In a static language, especially writing mostly from scratch it would add significantly to the code base. However, this is Python, and although I've not counted the code lines I doubt if there is much bloat as I've taken out almost as much as I've put in. There is some restructuring as well to better separate out the functions and particularly there is more in the common area as this is used by both local and remote systems. The secret is using some very powerful libraries.

I have added a Version 2.0 section to both Wiki pages to explain the new version.

Pylink-SR Version 2.1

A minor update to the main-line code but includes an extra plug-in to play with. Please see the Wiki for details. Read the install and particularly the bit about Jack.