On 02. 04. 20 20:07, Stephen Gallagher wrote:
This is a great example of the Nirvana Fallacy. Essentially, you are
arguing that improvements for one group is not acceptable unless it
improves things for every other group too.
(I will answer the entire e-mail hopefully later today, but I could not resist
to make myself more clear on this point now.)
I am not in any way saying that improvements for one group is not acceptable
unless it improves things for every other group too.
I am saying that we should not make changes that have unknown (arguably
negative) impact on any other group.
An example: When we had a proposal for dynamic BuildRequires, we added them to
rawhide only. Obviously, the group of packagers who maintain single-spec and %if
conditionals cannot benefit from that. And that is completely fine. They were
not impacted in any way, they just could not benefit from the new stuff unless
they change their packaging style. If they continue to use their current style,
they are not impacted in any way. Nobody said: "Dynamic BuildRequires are the
only way we handle DynamicBuildrequires and if that doesn't work for you, we'll
send you a Pull Request."
There is a big difference between saying: "all groups should benefit from this"
and saying: "no group should be negatively impacted by this".
--
Miro Hrončok
--
Phone: +420777974800
IRC: mhroncok