FreeDV 2_0_0_ RADE Doc

[drowe67]

Collect, draft, review and publish documentation on RADE. Like the software, this should not be ad-hoc, but carefully thought out and reviewed, a team effort.

• Collect info from disparate sources, such as (a) FreeDV 2.0.0-20241018 release email (b) RADE primer (c) any new instructions/notes that have been generated since 2.0.0-20241018
• Add content on our roadmap. We need to avoid making any promises on feature enhancements the PLT has not committed to.
• Document PLT policy re Windows/Linux/OSX distributions. For example "currently only Windows packaging is being maintained by FreeDV team - Linux users are welcome to compile from source, but must perform their own support".
• Decide where to publish, It would be good to make this a single location, e.g. README.md or freedv.org web page.
• Draft
• Review
• Publish

[drowe67]

https://freedv.org/radio-autoencoder/ has taken care of most of this.

Looks like all we have left on this one is the (RADE?) Roadmap. We need to be careful about making any big promises or timelines on public mailing lists - the initial RADE demo release is already massive delta. Much of the work from here on in will be "under the hood" (e.g. C port, bug fixes, minor features).

[Tyrbiter]

Looking at where we are now, I think we have to make it clear that whatever is released as RADEV1 with EOO text is going to be it until the C port is essentially done and all the ctests show it is working correctly.

Once people can see who they are receiving in good signal conditions then we can point to this as progress for everyday users.

The roadmap should avoid being definitive until the features under development can be shown to work as intended. And perhaps those feature improvements should be kept a bit quiet until it's clear that they do work in the real world with real channel impairments.

[tmiw]

Looking at where we are now, I think we have to make it clear that whatever is released as RADEV1 with EOO text is going to be it until the C port is essentially done and all the ctests show it is working correctly.

IIRC the C port is effectively done. The only part that hasn't been ported yet is the DSP stuff, and I don't know if that's going to happen before V2.

[Tyrbiter]

Looking at where we are now, I think we have to make it clear that whatever is released as RADEV1 with EOO text is going to be it until the C port is essentially done and all the ctests show it is working correctly.

IIRC the C port is effectively done. The only part that hasn't been ported yet is the DSP stuff, and I don't know if that's going to happen before V2.

I'm clearly not fully up to speed with what porting is needed. But when I refer to RADEV2 I mean a version with some of the improvements discussed with @drowe67 recently which is a step on from EOO text and RADEV1.

If I'm missing the details then I ought to know where to look to grasp the details.

[drowe67]

Take a look at "C Port of Core ML" in my Nov progress report

[Tyrbiter]

Oh yes, I had read it before but after reading it again I finally realised what it meant.