Re: [rtcweb] I-D Action: draft-ietf-rtcweb-ip-handling-01.txt

[Martin J. Dürst <[e-mail]>]

> On 3/27/16, 01:42, "rtcweb on behalf of Harald Alvestrand" <[e-mail] on behalf of [e-mail]> wrote:
>
>> Den 23. mars 2016 10:09, skrev Drage, Keith (Nokia - GB):
>>> Assuming it does not contain normative language in its own right, then
>>> it might well be a BCP, but then it needs to tighten up on some of its
>>> internal language – if we don’t mean RFC 2119 language then we should
>>> try and avoid the lower case forms as well, in order to avoid confusion.
>>
>> Matter of principle:
>>
>> I've argued against this idea for the last 10 years .... avoiding common
>> verbs makes it harder to write English. Enough people have a hard time
>> with that already that we shouldn't make it harder.
>>
>> If we mean something by the UPPERCASE form, we should stop trying to
>> police the lowercase ones. Indeed, the policing of the lowercase forms
>> serves to increase confusion between the forms, not decrease it: "if
>> this lowercase should is still there, does it mean that it wasn't
>> policed out, or that it was intended to be uppercase?"

<comment severity='minor' hint='see later for major comment'>
If I'd have to decide, I'd lean to the other side, because while for 
some people, the difference between upper and lower case is very strong, 
for others, it's not perceived that strongly. Examples may include 
people who's native script doesn't make such a distinction, and those 
people for whom software makers create OSes which treat upper and lower 
case as (mostly) equivalent, or default to case-insensitive for search 
and similar stuff.

In addition to that, there's the issue of getting it right, which is 
more difficult if the difference is only one of case, and the issue that 
styling may change case. Also, for those who have difficulties finding 
additional words, there's 
https://tools.ietf.org/html/draft-hansen-nonkeywords-non2119.

On the other hand, for those really addicted to case differences, there 
was also an RFC that defined keywords such as MUst and mUsT. You can 
guess the date of that one.
</comment>

<comment severity='major'>
As Harald says, this issue has come up many times over the past 10 (or 
more) years, in many different places. It would be good if the relevant 
entity (I have copied the IESG and the RFC Editor, but somebody else may 
be in charge here) would just decide. That way, WGs wouldn't have to 
spend time on this. And readers (who may not check the start of a 
document every time they check some detail), when they see a "must", 
won't have to worry whether it's a "MUST" erroneously spelled lower 
case, or just the equivalent of "need" or so.
</comment>

Regards,   Martin.