%config loses

[Elan Ruusamäe]

On 30.08.2016 22:34, Jeffrey Johnson wrote:
> Fix the following flaws in your bug report (sic).
this is not bugreport. complete reproducer and expectations were sent in 
previous thread.

> 2) Don’t use -F —freshen; instead use -U —update.


-U will install any packages matched, -F installs only packages that are 
actually installed. come on, jbj doesn't know that difference?
> 3) Don’t use *.rpm because I have no idea what operation is being performed.

ah come on. because i can type rpm -Fhv foo-1.0-1.rpm foo-1.1-1.rpm via 
shell glob doesn't give excuse rpm to destroy data. i've already 
explained what is my expectations which rpm5 does not obey.

> 5) Stop throwing multip[le instances of identical packages at rpm and expecting the right thing to happen.
you even understand yourself that is the key to the reproducer. if rpm 
can't handle this it should abort(3), assert(3), not destroy systems.

> I again suggest<rpm-devel at rpm5.org>  as discussion forum, because I will
> not change %config without discussion.
i don't see it going any differently than it has in pld-devel-en. there 
was complete reproducer last time i reported this. you refused to even 
try to understand and make your own reproducer in your favourite distro.

i don't have time or wish to "discuss" this. i consider it flaw in rpm, 
you don't. obviously you *win*.

> And I again suggesthttp://launchpad.net/rpm  as a bug reporting system, so that
> we do not have to repeat these exchanges every 3 months.

done that in the past. no difference. not interested wasting my time.

all other 1-7) are answered in previous thread.

-- 
glen