Re: [Nbd] [PATCH] docs/proto.md: Clarify SHOULD / MUST / MAY etc

[Wouter Verhelst <w@...112...>]

On Wed, Apr 06, 2016 at 01:41:31PM -0600, Eric Blake wrote:
> On 04/06/2016 01:21 PM, Wouter Verhelst wrote:
> > 
> >> * Put the above terms in bold.
> > 
> > I'm a bit in dubio about this bit:
> > 
> > - It's something that will tend to be forgotten, which would then result
> >   in a text with some things in bold and some things not in bold, which
> >   would be confusing.
> 
> Most contributors are either familiar with the document conventions, or
> are copy-pasting, so it will probably be correctly done either way (as
> long as the document starts from a consistent all-bold or no-bold state).

Well, we've had a few normative should/may thingies in lower case too,
so that's not a given.

[...]
> > - RFC2119 doesn't use bold (mostly because RFCs are plain text, anyway,
> >   but hey).
> 
> The RFC-to-html converters that add bold as a post-processing pass are
> optional.
> 
> > - It's more type work.
> 
> That one is true, particularly if a post-process pass can do the work.
> 
> > - Bold *and* uppercase isn't that much stronger than "just" uppercase, I
> >   think.
> 
> I tend to agree with this one; ALL_CAPS is enough to call attention to
> the word used in normative sense vs. in casual language.  I'm 50:50 on
> whether bold makes it better (I won't object to bold, but I wouldn't
> have proposed it either).

I'm going to reject it, then.

(have a look at the first paragraph of the "Newstyle negotiation"
chapter; it gets fairly close to a shouting match there)

-- 
< ron> I mean, the main *practical* problem with C++, is there's like a dozen
       people in the world who think they really understand all of its rules,
       and pretty much all of them are just lying to themselves too.
 -- #debian-devel, OFTC, 2016-02-12