Re: Amanda, bsdtar, and libarchive

Don Lewis wrote:
> On 24 Apr, Tim Kientzle wrote:
> 
>>If you want reasonably accurate estimates without reading the
>>file data, you could ... build some sort of "size estimate" capability into
>>libarchive ...
> 
> There are three variations on the archive size estimate:
> 	Fast and exact
> 	Fast and approximate
> 	Slow and exact

You can already do the last one with libarchive;
just register an output function that counts the
bytes but doesn't do anything.  Done.

"Fast and exact" is simply not possible in general,
so I don't think I would bother with that option
at all.

Incorporating some sort of estimator into the API does
sound like a good idea.  I'll have to think about it
some to see if I can figure out a good way to do that.

Thanks for the feedback,

Tim