James Carlson <[EMAIL PROTECTED]> wrote:
> Shawn Walker writes:
> > On 27/06/07, Joerg Schilling <[EMAIL PROTECTED]> wrote:
> > > Well, as I mentioned, the trusted extentions for Sun tar are undocumented.
> > > It seems that there is no need to have them (at least in the wway they
> > > are currently implemented). Customers may only ask to keep documented
> > > features to stay stable and compatible.
> > >
> > > Going this way would save at least half of the effort.
> >
> > I don't know that what you say is necessarily true. Undocumented
> > features are sometimes the most important ones to keep for backwards
> > compatibility. Especially when it comes to programs like tar which
> > people often use to do "backups."
> >
> > Any engineers that can speak for this point?
>
> I don't know what Joerg means by "undocumented." The -T option itself
> is clearly documented in tar(1):
>
> T
>
> This modifier is only available if the system is config-
> ured with Trusted Extensions.
>
> When this modifier is used with the function letter c,
> r, or u for creating, replacing or updating a tarfile,
> the sensitivity label associated with each archived file
> and directory is stored in the tarfile.
>
> Specifying T implies the function modifier p.
> [...]
With this documentation, we may safely replace Suntar by a star emulation
that implements -T but uses a completely different archive format in order
to implement the feature.
For compatibility with existing archives that are not from Solaris, it would
be sufficient to retain a copy of the old suntar implementation under the
name /usr/bin/otar.
Jörg
--
EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin
[EMAIL PROTECTED] (uni)
[EMAIL PROTECTED] (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
_______________________________________________
opensolaris-code mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
