Re: [Bug-tar] Proposed enhancements to "Sparse format 1.0"

[Sergey Poznyakoff]

Tim Kientzle <address@hidden> ha escrit:

> I've just finished implementing read support
> for the new GNU tar sparse file formats in
> libarchive/bsdtar, and have a couple of suggestions
> you might consider if you ever create a format 1.1:

Thank you.

>    * Don't pad.  The 1.0 format pads the prepended
> sparse block description out to a multiple of 512
> bytes.  There's no compelling need for this;

Actually, there is. It was designed this way to facilitate splitting
a condensed file into sparse headers and data blocks. Here, a
"condensed" file means the file created by extracting the archive with
a tar implementation that is not aware about 1.0 format. See:

  http://www.gnu.org/software/tar/manual/html_node/Sparse-Recovery.html

>    * Include a format specifier ("magic number")
> in the prepended block description.  Specifically,
> I'd like to see something like the following added
> to the beginning of the block description:
> 
>     #!gnu_sparse_file
>     #format=1.1

It is possible. But this information is already stored in PAX variables
`GNU.sparse.major' and `GNU.sparse.minor'. Why it is necessary to
duplicate it? 

Regards,
Sergey