On Jan 2, 2009, at 3:47 PM, Per Øyvind Karlsen wrote:
Hehe, FWIW DistTag doesn't have any relevance for EVRD (or whatever combination one wants to have) comparision in my world, it's just meant to be ditched from Release piggybacking to avoid it being taking into consideration for comparision, while adding it to file name as an identifier. :)
Glad to see that you've outlived your New Year's hangover ;-) But Distepoch: is string valued, unlike Epoch: and the patternsthat are not enforcing [0-9]+: matching, and so the "mdv" or "fc" or "epel" or ...
string can be easily buried within Distepoch: rather than separately
carried in RPMTAG_DISTTAG. Note that the data normalization
isn't quite as good, but you'll likely never notice.
Note also that all the oh so many many Fedorable postings from dweeb's
worried that The World As We Know It was going to end because
of the LC_COLLATE used by
strcmp("fc9", "fc10")
was going to suddenly do the Wrong Thing have all been rendered moot.
(*rpmvercmp) already does the right thing with digit strings.
All that remains is for Michael Jennings to realize that cAos can
quite easily get rid of E everywhere, no fuss, no muss. But
perhaps KainX is still busy editing *.spec files with mezzanine RE's
using perl.
And quite predicatbly someone is going to point out that RPM sux because
not only '-' but also ':' is denied everywhere.
Hint: Send Joey Hess another bullet point for his timeless APT <-> RPM
comparison page. RPM forbids the ':' character now, read all about it!
The problems of denying ':' and '-' in EVRD strings are solvable if
anyone other
than the Debian Borg really cared. 73 de Jeff
smime.p7s
Description: S/MIME cryptographic signature
