Josef Reidinger ([email protected]) wrote:
> On Wed, 30 Apr 2014 12:01:41 +0100
> Adam Spiers <[email protected]> wrote:
>
> > Josef Reidinger ([email protected]) wrote:
> > > On Wed, 30 Apr 2014 08:38:26 +0200
> > > Klaus Kaempf <[email protected]> wrote:
> > >
> > > > * Adam Spiers <[email protected]> [Apr 30. 2014 00:41]:
> > > > >
> > > > > This looks like a fundamental limitation of rpm's version
> > > > > requirement system, and I can't think of a good workaround right
> > > > > now :-/
> > > >
> > > > A proper fix has to wait for the next version of rpm which is
> > > > supposed to support boolean expressions in dependencies.
> > > >
> > > > See https://www.youtube.com/watch?v=fz4GJnFokso
> >
> > For those in a hurry, start watching from 6'52".
> >
> > > Well, it is teoretically correct, but practically not.
> >
> > I don't follow; please could you explain this in more detail?
>
> OK, probably I am not fully clear. I wanna to say something like it
> will work in future, but we need to solve it now.

Oh OK :)

> > > I think version
> > > with conflict is fine, just result is that you cannot have
> > > simulatneous installed more gems, which is drawback. So when new
> > > rpm is released and we know in which version, then gem2rpm can
> > > create conditional conflict.
> >
> > Agreed.  Using "Conflicts:" is a kind of evil workaround.
>
> Yes, it also have drawbacks, so we need to decide if current problem
> is bigger then problems from conflicts or not, so we live with current
> problems until new rpm will be released.

I agree.  My hunch is that introducing Conflicts is better than
leaving the status quo, but I'm not 100% sure.
-- 
To unsubscribe, e-mail: [email protected]
To contact the owner, e-mail: [email protected]

Reply via email to