Re: [spfbis] WGLC: draft-ietf-spfbis-4408bis-14

[Alessandro Vesely <[e-mail]>]

On Thu 09/May/2013 19:55:26 +0200 Murray S. Kucherawy wrote:
> On Thu, May 9, 2013 at 8:31 AM, Alessandro Vesely <[e-mail]> wrote:
> 
>>>    macro-letter     =  %x73 / %x6c / ...
>>
>> No, wait.  Do we really want that?
> 
> We do if current implementations generally do only lowercase macro
> support.  If they're case-insensitive, then what's there is fine.  It's my
> impression that %{S} isn't valid to most implementations.  If I'm wrong
> about that, then I withdraw the suggestion.

I agree the case (in)sensitivity can easily get overlooked.  Yet, the
syntax is the same.  Currently, the paragraph that specifies URL
escaping makes no reference to any grammar rule, but we could rewrite
it as:

   MACRO-EXPAND macros with an uppercase MACRO-LETTER expand
   exactly as their lowercased equivalents, and are then URL
   escaped.  URL escaping MUST be performed for characters not
   in the "unreserved" set, which is defined in [RFC3986].

I don't think that would make it any clearer.  Anyway, if we still
wanted to rewrite it that way, we could do so without having to define
MACRO-LETTER and MACRO-EXPAND, since rule names are case insensitive
as well [1].

>> The syntax of URL escaped expansions is the same as that of lowercase
>> macros.  We don't need two parallel sets of production rules, do we?
>>
>> IMHO, adding a comment, e.g. as follows, such suffice.
>>
>>   macro-letter     = "s" / "l" / "o" / "d" / "i" / "p" / "h" /
>>                      "c" / "r" / "t" / "v"
>>                      ; uppercase or lowercase, see URL escaping below
> 
> "uppercase or lowercase" is implicit in ABNF.  That's the point of my
> suggestion.

Yup, the comment turns implicit into explicit.  IOW, the grammar is
correct but readers overlook it.  Hence it's the accompanying text,
not the grammar, that we need to amend.  I think the reason why that
paragraph is overlooked is because it is too far down in the section,
not because it misses grammar references.  The second part of the
comment is an attempt to draw the reader's attention there.

-- 
[1] For example, the <mechanism> rule has uppercase A / MX / PTR / IP4
/ IP6, while those terms' definitions are lowercase in the subsections
of Section 5 where they are introduced  --a change committed on August
16, 2012 (-04), a period we were lowercasing various stuff.