[Tools-discuss] [Cellar] non-ascii characters (fwd) Dave Rice: [Cellar] non-ascii characters

[Michael Richardson <[e-mail]>]

reposting to tools-discuss for comments.
I suspect that the idnits tool may be in error in this case.

We are also using xml-v3 for ffv1, since it is full of math which gets
rendered to SVG.  

From: Dave Rice <[e-mail]>
Date: Tue, 27 Aug 2019 10:24:41 -0400
To: Codec Encoding for LossLess Archiving and Realtime transmission
 <[e-mail]>

Hi cellar,

I’m doing an idnits review on the latest ffv1 draft and am confused about the
IETF rules for non-ascii characters in txt rfc drafts. The update at
https://tools.ietf.org/html/rfc7997 <https://tools.ietf.org/html/rfc7997>
seems to allow the RFC Editor to use some discretion in using non-ascii
characters, but idnits <https://tools.ietf.org/tools/idnits/> still complains
about that. One instance is using a Unicode minus sign to be more
semantically clear than an ascii hyphen. Also since the Unicode minus sign
uses more bytes than the ascii hyphen, idnits will say the line is too long,
even though in appearance the line displays within the correct width. 

In the RFC for ffv1, we’ve used unicode symbols for floor and ceiling
functions, such as ⌊a⌋ and ⌈a⌉, but rfc2xml transforms these into their
hexadecimal expressions such as &#8970;a&#8971; and &#8968;a&#8969;. Any
advice, should we listen to idnits and avoid non-ascii characters or listen
to rfc7997 and use them when it makes the content more clear? 

Kind Regards,
Dave Rice

-- 
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-