Re: [Gnu-arch-users] corrupt archive

[Aaron Bentley]

Jani Monoses wrote:
> Just kidding.
>
> Shouldn't arch_revision_type() print a less heart-sinking message in case the
> revision name given to it is wrong?
...
> Aaron maybe you have this already in your errors branch?

I do have it fixed there, and it also generates a good error if you 
specify a directory that doesn't exist.

Actually, I think the arch_revision_type error is spot on, *if* if arch 
got that revision name from archive data or tree data (which is the 
common case).  So we're validating the user input much higher up, in 
interpret_delta_path ().

Aaron

--
Aaron Bentley
Director of Technology
Panometrics, Inc.