[rfc-i] dates, was: Comments on draft-hoffman-xml2rfc-06

Paul Kyzivat pkyzivat at alum.mit.edu
Fri May 2 09:40:38 PDT 2014

On 5/2/14 10:55 AM, Elwyn Davies wrote:
> On Fri, 2014-05-02 at 09:30 -0400, Heather Flanagan (RFC Series Editor)
> wrote:
>> I'm not sure I see the value in adding support for numbers to represent
>> months, requiring the processor to convert the numbers to names.
> Well, it's not exactly the most costly piece of code one might ever
> write!  Since the processor will have to have a list of month names to
> check against...
> No big deal either way.

The cost of implementing alternatives is certainly insignificant.

IMO of more concern is that every extra thing added increases the 
overall cognitive load on users - especially those learning for the 
first time. And any individual one of these is insignificant in 
isolation. Rather it is death by a thousand cuts.

If we were starting from scratch, and choosing *between* name and 
number, then one might make a case for choosing the number. (Though I 
always have to mentally translate to numbers and there is always the 
possibility of getting it wrong.)

But we have a backward compatibility requirement for names. Numbers will 
be an alternative, not a replacement. IMO this doesn't meet the cut of 
providing more value than added complexity.


More information about the rfc-interest mailing list