[Spice-devel] Bug#603699: ITP: celt051 -- The CELT codec v0.5.1

[Ron]

Hi Hans,

Hans de Goede writes:
> The if at all part depends on if it will be doable without too much
> pain to support both celt-0.5.1 and celt "1.0" in the same binary.
> This is important to us as we care a lot about protocol
> compatibility.

Are you seriously telling me that you have no plan whatsoever for how
to transition from a random snapshot of an experimental codec, except
to hope that somehow it might in some way become embedded in some other
binary for the rest of eternity?

I really hope you are not, but if you are, and other people accept that
as a valid solution -- then how many apps should the distro extend this
sort of madness to?  I can think of several who would love to if we are
to permit this.  But so far, all have accepted this _is_ an experimental
codec, and they must be prepared to move with it.  The interested app
maintainers and upstream discussed this, and we settled on 0.7.1 as the
next stable epoch for things that want broad interoperability.  We waved
at everyone who would listen about doing the same.

Are you aware there may never be a celt 1.0?  We have roughly 2 years
from today before any version of spice will have a chance to be even
considered for the next Debian stable release - and if things go as
expected, it will not include any version of celt at all.  There will
instead be a new (and standardised) codec that was spawned from it and
merged with other codec work.

At some point, if the experimental spice wants to become a mature and
portable application, it's going to have to acquire the ability to deal
with this sort of thing.  I'd really like to encourage you to use this
release cycle (and Debian) as your lab-rats for getting that right.
Then you don't have to worry about upsetting corporate clients who you
made guarantees to, to get a wide audience for testing -- and by the
time Debian does freeze, I'd hope you have this worked out well enough
for everyone.

If that means an update to the spice bitstream protocol, now might be a
good time to explore one.  I really do want to see people experimenting
with celt.  But I really don't want to see that become a poor facsimile
of The BDB Problem.  If the draft spice protocol isn't really suitable
for the distro as is, then we should fix it until it is.  That would be
a much better outcome than collecting random snapshots of unmaintained
things that only one experimental application actually needs.

Wouldn't it?

(trying not to be grumpy, but kind of frustrated by this sort of thinking),
Ron