Re: [Gnu-arch-users] Re: Making --setup default in tag and import

[Miles Bader]

On Wed, 9 Feb 2005 10:25:13 +1100, Ben Finney <address@hidden> wrote:
> This then breaks the "one and only one [...] forever and ever" that
> Aaron said, above.
> 
> Once something has happened in the archive it *has happened*; any "undo"
> or "redo" is a further change, recorded in addition to the original.
> That's the model on which GNU Arch is founded; a deliberate design
> decision.
> 
> You seem to be suggesting technical "solutions" to subvert something
> that is a fundamental part of the design of GNU Arch.

Christ, can we be a bit more pedantic please?  The "11th deadly sin",
as used recently against arch :-):  "11.  to rate the beauty of
mathematics above the usability of your compiler."

This attribute of arch is great in general, but it only really
_matters_ if someone has a dependency on the information in question. 
In practice, a command to delete stuff from a private local archive in
certain cases would be a nice tool in a limited set of circumstances,
and people would appreciate it.

If such a command were actually offered by tla, it should try to
automatically enforce safety (e.g., once if a branch/revision has been
mirrored/checked-out, it shouldn't be deleted) and maybe implementing
such enforcement is too annoying.  But that's a practical decision.

BTW, on the original subject, the idea that requiring -S is actually
going to make it safer seems ... ludicrous.  As has been pointed out,
it will simply become an automatic part of such commands, a
"--yes-really" option, and people will create bogus branches because
of typos with exactly the same frequency as if there were no -S
option.  So while the question of whether certain archive deletions
should be allowed is interesting, I'm not sure it really affects the
issue of -S very much.

-Miles
-- 
Do not taunt Happy Fun Ball.