On 05/19/2009 02:36 PM, Jesse Keating wrote:
On Tue, 2009-05-19 at 14:23 -0700, Toshio Kuratomi wrote:
I agree with this sentiment 100%. We're here to make better packages,
not whine about how making better packages is hard. However, I'd point
out that in this specifc case Jesse raised the "easier" argument and
Till was just replying to that.
I think that my point was that it is an easier to understand policy.
It came across as implementable, not as understandability:
""""
There is an easier option 3, which is no flags in Fedora period,
regardless of what spin. Far easier to implement.
"""
It
may be more difficult to enact, but it is a far sight easier than trying
to figure out what other packages might require your flags, and if any
of those packages are on something called a "default spin", which can
change it's contents at any time.
And here you're again arguing about implementation, not about
understandability.
Where we deal with implementation, we deal with Bill's note about
"easy". It's easier to trust that your packagers haven't packaged
anything that has flags than to write a script that looks for the
Provides (or filename) in the package set but it's not necessarily right
for Fedora.
-Toshio
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
