Re: [PATCH v2] Provide some linguistic guidance for the documentation.

From: Marc Branchaud <marcnarc@xiplink.com>
Date: Aug 2, 2013, 14:16 UTC
Message-ID: <51FBBF42.5040102@xiplink.com>
In-Reply-To: <20130802062511.GA3013@elie.Belkin>

On 13-08-02 02:25 AM, Jonathan Nieder wrote:

Show 20 quoted lines

> Junio C Hamano wrote:
>
>> Is that accurate?  My impression has been:
>>
>>      The documentation liberally mixes US and UK English (en_US/UK)
>>      norms for spelling and grammar, which is somewhat unfortunate.
>>      In an ideal world, it would have been better if it consistently
>>      used only one and not the other, and we would have picked en_US.
>
> I'm not convinced that would be better, even in an ideal world.
>
> It's certainly useful to have a consistent spelling of each term to
> make searching with "grep" easier.  But searches with "grep" do not
> work well with line breaks anyway, and search engines for larger
> collections of documents seem to know about the usual spelling
> variants (along with knowing about stemming, etc).  Unless we are
> planning to provide a separate en_GB translation, it seems unfortunate
> to consistently have everything spelled in the natural way for one
> group of people and in an unnatural way for another, just in the name
> of having a convention.

Personally I find it distracting when the norms are mixed. I don't think the current mishmash pleases anyone (as evidenced by the steady stream of patches that change spellings).

Show 6 quoted lines

> I am not sure it makes sense for the documentation to say "A huge
> disruptive patch of such-and-such specific kind of no immediate
> benefit is unwelcome".  Isn't there some more general principle that
> implies that?  Or the CodingGuidelines could simply say
>
> 	The documentation uses a mixture of U.S. and British English.

I'm hoping this patch will help the list avoid seeing patches that merely flip between alternate spellings. (Perhaps the commit message should state this?) I think it's important to be clear about the kind of work the git community wants to see. So I don't think that single sentence by itself would be helpful.

In the case of alternate spellings in the documentation, I think there's a temptation to create a blanket patch that "fixes" lots of perceived misspellings since such changes are mere typographic tweaks. It's a bit easy to overlook the disruptive nature of such a patch, so IMO it bears repeating here.

Are you suggesting maybe that there should be no "official" dialect? That the guidance should simply say that we don't want to see patches that merely flip between alternate spellings (because we don't really care)?

M.