Hi,
Just fyi I'll look at those (for osmo-gmr and libosmo-dsp) in detail over the weekend.
I'll have to see exactly what they do because the state of the repo might not reflect what I have in mind for the future. (for eg I see the gmr1 code become a library with the installed headers, or even if l1 is the only one used in both binary in the public repo, I still have some test / future binary that depend on both sdr/ and l1/ ... things like that).
In general I'm not a big fan of the flattening ... I kinda like that each repo is responsible for defining it's own source files. Beside build speed (which for me is not really an argument given it's so small), what are the advantages ?
Cheers,
Sylvain