[lsb-discuss] RPM decision process

Theodore Tso tytso at mit.edu
Wed Apr 16 18:36:06 PDT 2008

On Wed, Apr 16, 2008 at 05:35:13PM -0700, Wichmann, Mats D wrote:
> again, this is an overloading of terms.
> everybody supports the rpm4 package format, and has since sles8 was
> the straggler (I think it was sles8, could have been older - this
> was the *only* reason the spec called out the v3 package format).
> the rpm5.org technology is not incompatible with the v4 format, 
> and is itself not under discussion anyway because the LSB doesn't 
> talk about the tool you use for package management, only about the 
> file format (actually we do need a tool you can call, but that's 
> a different story - the creature that's been known as lsbinstall).

OK, it wasn't clear to me whether or not the rpm5.org technology was
incompatible with the v4 format.  It was my impression that rpm5.org
was going to be adding new tags that wouldn't be understood by rpm
4.4.2, but if that's not true, I'm happy to be corrected.

As I said earlier, my main concern is that the rpm v4 format isn't
documented in a form where we could easily use it as the basis for
uplifting the LSB specification for the RPM file format.  I thought I
heard Russ claim that it was in the Maximum RPM book, but that doesn't
seem to be the case.  I looked through both the rpm.org and rpm5.org
sites, and couldn't find any such RPM file format specification.  And
again, if such a specification does exist already, again, I'm happy to
be corrected and I invite someone to send us the URL.


						- Ted

More information about the lsb-discuss mailing list