Print

Print


On Thu, 20 Jan 2011 07:33:03 -0500, Gerard Ashton wrote
> Considering all the subtle problems we have been discussing, I would 
> think that some people will want to take advantage of the format as 
> a way to originally record information on paper with a pen, or in a 

My point.

> simple computer program such as Wordpad or EMACS. This might be 
> particularly helpful for people in a multi-cultural environment 
> where subtle distinctions of spelled-out words like "weekdays" might 
> be misinterpreted. So I think the format should, as much as possible,

My favorite example is 9-11 which has a historical
significance as 11 Sept but also as 9 Nov-- among other
events the Kristalnacht, Beer Hall Putsch, the opening
of the Berlin Wall, ....
The ISO format is compact and well defined.. it lets us
talk about what we want to talk about.. 

>  be suitable for direct human use with no computer intermediary. 
> Certainly some aspects of ISO 8601 are used in this way (or maybe 
> ISO 8601 adopted what some people were already handwriting, I don't 
> know).
> 
> It follows that some expressions will be regarded as wrong when read 
> by humans, and the people will contact the author to determine the 
> correct value. The date-time 2011-01-19T25:29Z is likely to generate 

This is why I made the distinction between normative form
and generic expression. Non-normative expressions such as
2011-01-19T25:29Z I see as for computer-computer communication
and explicitly for "dumb" clients. It lets clients specify
for any given date-time a date-time some seconds, minutes,
hours, days etc. later. 

> a phone call if it was written by a person, no matter what the spec 
> says, just as Friday, January 20, 2011 would. So one wonders whether 

This is something else.. The specification of Friday is a
kind of check-sum.... even in programs parsing RFC 822 style
dates such dates need to be triggered as wrong. At the very
last either day name (Friday), the month, day or the year is wrong..

> values should be allowed that would be rejected if read directly 
> rather than through a software intermediary.
> 
> Gerry Ashton


--

Edward C. Zimmermann, NONMONOTONIC LAB
http://www.nonmonotonic.net
Umsatz-St-ID: DE130492967