zstd compression

[Robert J. Hansen]

> I am admittedly unfamiliar with the history here. Can I ask what made
> bzip2 controversial?

Werner knows this history better than I do; forgive me for skipping
this, please.

> From my uniformed vantage point, it seems that a better algorithm is
> available so we should be using it.

"It's a better algorithm" is rarely a good reason to do anything with a
mature standard that has a nontrivial user base.  New features should be
driven by the changing needs of users, not just because something
marginally better is available.

If you can find ten people for whom the lack of this feature
significantly impacts their use of OpenPGP, then sure: user needs have
changed and we should change with them.  But if you can't find ten users
who have a critical need for this feature, then ... why do we need to
add it?