Re: [edk2-devel] [PATCH v1 0/8] Fix new typos reported

["Ard Biesheuvel" <ardb@kernel.org>]

On Tue, 17 May 2022 at 21:32, Michael Kubacki
<mikuback@linux.microsoft.com> wrote:
>
> As noted in the patch, this BZ was filed to follow up and review those:
> https://bugzilla.tianocore.org/show_bug.cgi?id=3929
>
> I don't like doing this either but the spelling errors do exist. I am
> trying not to make CI policy changes as those can be controversial even
> among maintainers in the same package and is an orthogonal conversation
> to addressing pre-existing issues within the presently defined CI policy.
>
> In this specific case, the ignore list in the package CI YAML file can
> be used to explicitly identify known typos and the BZ explicitly tracks
> reviewing those so there's a well defined path to resolve and fix the issue.
>
> I personally feel that's better than ignoring the problem entirely but
> it also depends on where your package code ends up getting consumed and
> the requirements and burden it might place on those consumers. For
> example, if it ends up in auto generated documentation and that
> documentation has spell check enabled, it becomes a downstream override.
>
> There's currently several PRs active that fix typos so others see some
> value in this work (as opposed to disabling spell checking):
>    - https://github.com/tianocore/edk2/pull/2900
>    - https://github.com/tianocore/edk2/pull/2789
>    - https://github.com/tianocore/edk2/pull/1955
>
> For future changes, I suggested lock the cspell version and I think
> that's an option to prevent these from appearing at unknown points in
> time. I'm not appointed to make authoritative decisions about that (to
> my understanding) so I am making that suggestion for the community to
> consider.
>
> Again, I don't have a strong opinion on this topic, I've been waiting 4
> weeks to get the v5 patch series merged (other busy work in between),
> and you're the maintainer. It sounds like if I take ownership of BZ
> 3929, you might be okay with leaving it enabled? I can do that but
> there's so many words in this instance, I wanted someone closer to the
> package contents to look at it.
>
> If you still strongly feel you would prefer to have it disabled, I will
> pull that change in and see if any opposing opinions surface. However, I
> wanted to double check this is what you want to do right now.
>

If you feel it is worth your time to fix typos in existing comments, I
won't stand in your way. But I don't feel it is worth my time, given
that it doesn't actually improve the code, except for by some
artifical measure of spelling-correctness, which has no bearing at all
on what runs on people's machines, and as far as I can tell, these
typoes do not create any confusion regarding what the comments intend
to convey.

Adding typoed words to the ignorelist is the worst possible solution,
because you will be wasting your time only to placate the machine,
accumulating technical debt in the code base without actually fixing
the problems. So that is out of the question for me.

If you want to fix these issues, that is also fine. I will review/ack
with priority provided that I actually have any bandwidth available.

But if we are working for the CI instead of the other way around,
something is seriously wrong. If we can't roll a stable tag because
the CI wants us to fix our typoes first, we have to be able to
override it. And corrupting the codebase by adding typoes to the
ignorelist just to placate the CI is preposterous..