Re: DEPENDS semantics (was: removing useless dependencies)

[David Holland <dholland-pkgtech%netbsd.org@localhost>]

On Sun, Jul 12, 2015 at 06:22:50PM +0200, Joerg Sonnenberger wrote:
 > On Sun, Jul 12, 2015 at 04:56:31PM +0100, David Brownlee wrote:
 > > If the possible dependencies are specified as an ordered list,
 > > then a reasonable expectation would be that the first one which is
 > > present is used.
 > > 
 > > So {foo>=5.00, bar>=8.00} if there is a foo>=5 present then bar
 > > would never be even checked
 >
 > [...]
 > assuming

See, there's the problem - you can't build special-case assumptions
into a general-purpose construct like package patterns and expect to
get sensible results.

If foo 5.00 is ten years old and bar is much newer, and you think this
is important, then the pattern should be written in the proper order
so that bar gets chosen by default. I don't see what's difficult about
this.

I might grant that it makes more sense to choose from the right rather
than the left, as we tend to write things in ascending order. But it
sure as hell doesn't make sense to compare numberings of unrelated
things.

-- 
David A. Holland
dholland%netbsd.org@localhost