On Jan 15, 2009, at 5:42 AM, devzero2000 wrote:
So which GUPI approach, through "arbitrary" tags, or through a
common mechanism?
Through a common mechanism, for sure. But this should not be taken
within a committee of standardisation? LSB perhaps. Heresy? But a
"rpm" Foundation does not exist yet and, unfortunately, not ever
exist.
If LSB was paying attention (or there were a "RPM foundation"), then I
would definitely
pursue a GUPPI "standard" through those means.
FWIW, all of --xml, and RPMTAG_PKGID/RPMTAG_HDRID/RPMTAG_FILEID were
added to rpm to __ASSIST__ LSB package "standards". None of that
functionality
has ever been looked at for various hysterical reasons.
But the ideal world where "standards" are perceived as being in
everyone's interest
is not the world I live in. So a unilateral implementation @rpm5.org
is all that I can pursue.
The examples that you place confirm that no vendor has decided to
comply with a common choice and, perhaps, is not even been discussed.
To be fair to vendors, all of "decided" and "comply" and "common
choice" are not exactly obvious.
And inter-vendor discussions are always difficult, particularly with
competing
products and approaches to identical problems.
73 de Jeff
______________________________________________________________________
RPM Package Manager http://rpm5.org
Developer Communication List rpm-devel@rpm5.org
