Re: [gentoo-dev] bash-4.4 - call for testers

[Kent Fredric]

On Mon, 3 Oct 2016 13:32:34 +0800
konsolebox <konsolebox@×××××.com> wrote:

> "A useflag that entirely goes away depending on the version" (or a
> flag that is implemented in one ebuild but is not on another) is
> pretty common among packages and I see that as totally valid, and is
> way better than a solution that uses dynamic default.

I think the main assumption I'm trying to challenge here is that users
would desire different USE and BUILD configurations based on _rc
status, and I don't think that's much the case.

People IME tend to prefer to set something on a package level, and then
apply consistent enforcement for all versions, or perhaps apply USE
flags to ranges of versions ( ie: everything larger than X, everything
smaller than X ), or on a per-slot basis.

Mostly because its consistent and minimises the amount of effort.

Thus, for basically any argument you can make for having the USE flag
at all, maps to an argument that applies to both _rc and non _rc
versions.

And I get the impression that the desire you have to have different
behaviour for non-rc versions is a bit niche.

( I can appreciate it in a -9999 version, but that produces a situation
which falls under "all versions larger than 9" and thus gives you
version-range based consistency. )

Anything else seems to me to devolve into "being too fiddly", and may
discourage testing.

But I also see it that people who are testing _rc + KEYWORDS=""
versions to be people who's express intent is to optimise for maximum
breakage, in order to find the most possible problems before they
happen, and problems like "readline version is too new or too old" are
exactly the kinds of problems I'd want to make sure we didn't have
before we shipped bash in that configuration to everyone ( even to
~arch )