But we have these recipes already.
I'm still low in the learning curve. But last time I tried to use LMS
inside Docker (I believe incl. your image), it didn't work for me. But
that's probably been a year ago.
I'd have quite a few questions wrt. to your image I'd be happy to
discuss with you. And yes, I'd consider building images whenever builds
are run. Just make them part of the build pipeline I already have to
create the other builds. Because why not? :-). And because it would also
allow us to not have to download 200MB of mostly unused binaries. If we
settled for one linux base image, we'd know the binaries required for
eg. x86_64 and aarch64, and could strip out all the rest.
So, if I may:
- locale: can this be defined based on the host system's locale instead
of setting it in the dockerfile?
- do I understand correctly that your image would run on "any" hardware
platform that LMS and Docker support, because you pull in the full fat
_all deb image?
- you install a bunch of tools like espeak, faad, faac, ffmpeg etc.
which either are not used by LMS, or come shipped with LMS. Why are they
required?
- you open up ports 9005, 9010 - what are they needed for?
- would Spotty work inside such an image? I see that you forward 5353
which might be part of what's required.
- what about popular plugins like philippe_44's, which expose virtual
player? Would they be working, because the helper would run inside the
container (while talking to devices outside)? Or would they fail to eg.
find speakers, because they're on a different network?
- I left a comment on github about the MIP patch you apply. I'd be happy
to merge it to LMS if you provided a PR.
--
Michael
_______________________________________________
unix mailing list
unix@lists.slimdevices.com
http://lists.slimdevices.com/mailman/listinfo/unix
