Re: [PATCH 1/2] docs: Fix botched rST conversion of 'submitting-a-patch.rst'

[Kashyap Chamarthy <kchamart@redhat.com>]

On Mon, Nov 22, 2021 at 03:01:58PM +0100, Thomas Huth wrote:
> On 22/11/2021 14.53, Peter Maydell wrote:

[...]

> > I don't think we should be recommending to new contributors that
> > they do things that established contributors generally do not do.
> > The document has enough "things you should do or think about" already.
> > My preference would be for simply not mentioning spelling-checking.

Fair points; and yes, the doc can be intimidating as is.

> > (If we do want to come up with some process for dealing with
> > spelling issues in the codebase, then we either need to put it
> > into CI so it's run automatically, or we need to have something
> > that works on the individual patch level.)

For individual patches, some projects use commit hooks for `aspell` /
`codespell`.  The contributor still needs to wade through false
positives, though.  For Sphinx-based documentation, there's
"sphinxcontrib-spelling"[1].  I don't know how effective it is, but it
lets one configure project-specific private dictionaries[2], which could
eliminate many false positives.

> Ok ... In any case - seems like this needs more discussion, so I'll
> drop it from the patch for now. We can still add some wording or CI
> magic later, but that's certainly something that we rather want to do
> after version 6.2 has been released...

Yeah, dropping it sounds fine.

[1] https://github.com/sphinx-contrib/spelling
[2] https://sphinxcontrib-spelling.readthedocs.io/en/latest/customize.html#private-dictionaries

-- 
/kashyap