Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in media-sound/umurmur: metadata.xml ChangeLog

[Tom Wijsman <TomWij@gentoo.org>]

[-- Attachment #1: Type: text/plain, Size: 2948 bytes --]

On Sat, 28 Dec 2013 18:26:47 +0100
Patrick Lauer <patrick@gentoo.org> wrote:

> > The discussion is based on some questions that are hard to agree on:
> > 
> > 1. How much of a problem is an unused USE flag in metadata.xml? 
> 
> Cosmetic issue. No functional impact.

This is a description of an unused USE flag instead of an answer;
asked differently, how much does its presence impact quality?

> > 2. Should such repoman warnings be fixed? By whom? When? How?
> 
> Yes. You see it, you fix it.

Once I get around to them, but there are more important things to do.

> Not fixing cosmetic issues (cf. compiler warnings) leads to more and
> more noise until real issues are just drowned in the noise; the only
> way to achieve excellence (in terms of quality) is discipline in
> adhering to rules and standards obsessively.

Does that noise distract us from fixing what really needs fixing?

It seems to me that for example migrating ebuilds to newer eclasses is
much more important than to fix something that's just cosmetic. But it
appears that fixing cosmetic issues is easier and more fun to do...

    "Kill it with fire!"

> If there are (too) many false positives we should add proper
> annotations to silence repoman ...

Are there other false positives than the multi-line readme eclass ones?

> > 3. Can USE flags actually be removed from stable ebuilds?
> 
> Usually removing stable ebuilds makes useflags "disappear", rarely
> does masking stuff (e.g. old cups) lead to the disappearance of
> useflags as they would now be broken

The question is about the removal of an USE flag, not the removal of an
ebuild; it rather addresses changing the ebuild to remove the USE flag,
to address the context of this thread.

> > 4. ...
> 
> Do we need to discuss this?
> 
> No. It's an amazing waste of time, almost as if we had no real
> problems to focus on.

Maybe, maybe not; some parts of it can yield bike-shedding, but other
parts of it can definitely be useful to talk about. Eventually we will
need to discuss as a QA team to focus on getting the harder stuff done;
because well, just shooting at the low hanging fruit and taking the
easy way out and only coming together after "something happened" isn't
the way that works well. This thread is a good demonstration of that...

> > Because this can yield quite some bike-shedding; the alternative
> > "out of the box thinking" package.use.mask solution satisfies both
> > parties, which renders that discussion unnecessary. Nobody has
> > objected this.
> 
> That is a "fix" specific to this package alone, in the general case it
> is not valid.

You either want to keep the USE flag description or you don't.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]