Re: [Monotone-devel] Proposal for human readable revision IDs

[Bruce Stephens]

Thomas Haas <address@hidden> writes:

[...]

> Using the alphabet (A-Z, a-z) and the ten digits would reduce the
> real estate from 40 to 27 characters.

That would be shorter, but would it be any more readable?

> Additionally, or alternatively, a different, persistent
> representation of the various identifiers could ease the use of
> monotone. E.g.  revisions could be counted as in sub-version or
> monotone limit itself to display identifiers as short as possible,
> while still unique (the result could be fed into monotone complete
> to get the full identifier).

A difficulty with that is that these things would then not be
permanent.  At present the selector "aa0" is unique for the database
of venge.net that I have, but that won't always be so.  So if I cut
and paste "aa0" and store it somewhere, then that's not going to work;
similarly, if I email that revision to someone, then that may well not
be unique for the recipient.

> Personally, I would prefer some short, readable identifier valid
> only for my database.

Why?  What sorts of problems do you have with 40 character hashes?
Are you aware of selectors, which allow you to use (amongst other
things) unique prefixes of hashes everywhere (I think) where you might
use a revision?

Something I've considered is some kind of hook to print out revisions,
so that rather than monotone displaying
aa0a570fc9966431392a9956ea1ce270456c162a, it might produce
"a:njs/b:net.venge.monotone/d:2004-05-12/i:aa0a", or something, which
doesn't save space, and wouldn't be particularly memorable, but it
seems more readable.  It would also be more awkard to select for cut
and paste, though.  (That's a legal selector, by the way.  The a:, b:,
and so on can be omitted, and there's more flexibility that's
possible.)  Overall I'm not sure it would be worthwhile; a converter
might be handy, so you could convert a hash into something meaningful
to email to someone.