Re: iso-timestring fix / proposal for saner timezones

On 2/2/2010 10:43 PM, Harald Hanche-Olsen wrote:
>> It now outputs
>>       YYYY-MM-DD HH:MM:SS.UUUUUU
>> which should be a valid iso8601 timestamp and is parseable by
>> parse-timestring et al. Additionally, it does not print the ".UUUUUU"
>> part if there are no microseconds; useful for the databases that don't
>> support fractional seconds.
>>     
> Actually, as far as I understand iso8601 that is not a valid
> timestamp. I have here a pdf of ISO 8601:2004(E), and it quite
> explicitly specifies the letter "T" as the time designator, i.e., the
> character between the date and the time. For example, one should write
> 2010-02-03T03:34:16Z and not 2010-02-03 03:34:16Z for a certain moment
> that passed while I was writing these words (the letter Z stands for
> UTC). A space is often used in place of the T for human readable
> output, but then it is no longer strictly ISO 8601.
>
> I am not necessarily saying that the letter T should be used in this
> context; all I am asking that one departs from ISO 8601, one should at
> least be aware of the fact (and perhaps even document the departure).
>
> Amusing detail: The standard says
>
>   Unless explicitly allowed by this International Standard the
>   character “space” shall not be used in the representations.
>
> I cannot find any place in the standard which explicitly allows a
> space, rendering the first half of the above quote pointless.
>
> - Harald
> _______________________________________________
>   

Excellent to know, we're working off of secondhand documents (mostly
wikipedia) for that standard and I was thinking it was optional. I've
changed to include the 'T' and tests are still passing!

-- 
Nathan Bird
[email protected]
http://www.acceleration.net/ 
Custom Programming, Design, Hosting, and Broadband.

_______________________________________________
CLSQL mailing list
[email protected]
http://lists.b9.com/cgi-bin/mailman/listinfo/clsql