On May 23, 2011, at 4:03 PM, Klaus Kaempf wrote: > > Thanks, thats good information for getting satsolver to work with rpm5. >
Which -- as you know -- I started porting libsatsolver in April 2008.
The issue I have with SATsolvers is mostly this:
When a depsolver based on SAT fails, you know
No solution exists.
When a depsolver that is NOT based on SAT fails, you know
No solution was found.
While I fully understand the differences between "doesn't exist" and "not
found",
does any end-user really care why an upgrade did not work?
BTW, I had some brief meet-up with YAST/Zypper while checking
RPM portability on s390x at IBM. YAST/Zypper beat the crap out of anaconda
yum for "usability" on s390x.
OTOH, I have seen -- because I typed a ^C -- zypper set about
removing python in order to SATisfy the hungry downgrading
depsolver. At least zypper handled ^C, yum took 4+ years to figger
how to handle ^C mostly correctly.
73 de Jeff
smime.p7s
Description: S/MIME cryptographic signature
