On 10/2/24 08:56, stefan11111 wrote:
> Both removals definitely not still being contested and debated.

You've conveniently ignored the context immediately below the line that
you chose to quote and somehow decided to try and shoehorn in discussion
of a completely different (and settled) issue. Congratulations, it's not
often that I get to open my emails in the morning and untangle several
different flavours of wrong in only a single line!

> Maybe support as much choice as possible, and not act like you know what
> users want better that the users themselves.

FYI I don't appreciate your tone. Purge it down a little next time,
please?

Knowing what's best for users, especially when the users in question
appear to be willfully misinformed, is actually part of a developer's
job description. We even have a fancy section of the dev manual that's
all about not adding superfluous USE flags because we've learnt through
experience that providing a toggle for every single little thing in a
package is not, in fact, beneficial to users overall.

https://devmanual.gentoo.org/general-concepts/use-flags/index.html#when-not-to-use-use-flags

Users _rely_ on developers to make sane decisions on their behalf,
particularly around which USE options should exist and be togglable and
which options it makes sense to just hard enable or disable in the
configure step. In general we don't expect users to have an in-depth
knowledge of an upstream build system; they don't need to be across the
implementation details of every single configure option, they just need
to trust the options enabled by ebuild developers (and by extension the
decisions made by the wider project, including [and especially] areas
like QA).

In addition to this there are mechanisms through which a user can choose
to override the decisions made within an ebuild, including but not
limited to the already-mentioned EXTRA_ECONF, MYMESONARGS, MYCMAKEARGS
in package.env.

TL;DR: Methinks thou dost make too great a mountain of that which is but
a molehill

On to the main subject, we already default USE=unicode to on in the
ebuild and a browse over the sbcl docs suggests that it's also the
upstream default and is unlikely to break anything given that it's been
that way since version 0.8.17 and we're well into major version 2 by
now. Just drop the USE and default it to on - QA is right in this case:
in the absence of a compelling reason to keep the USE it should be
dropped.

Cheers,

Matt

Reply via email to