Hi,

On 2017-01-27 11:08:10, Nicholas Boel wrote to Wilfred van Velzen:
  about: "FSP-1040.001 Draft #3":

 WV>> I don't think the 'minute' can ever be "60" !?

 NB> What is "the ABNF notation of the format that messages SHOULD use?" I see
 NB> that ABNF notation is mentioned here:

 NB> http://www.odata.org/documentation/odata-version-3-0/abnf/

 NB> ..and is defined like so:

 NB> zeroToSixty = [ "0" ] oneToNine /
 NB> ( "1" / "2" / "3" / "4" / "5" ) DIGIT /
 NB> "60"

 NB> This is also mentioned in the bibliography later on in the document,
albeit
 NB> maybe not the same link I posted above. That seems to include 60 and has a
 NB> couple RFC's to back it up.

 WV>> The 'second' can be "60" when there is a leap second inserted into
 WV>> UTC. It would be very cool if anyone ever accomplished to write a
 WV>> message with that date/time stamp! (Without cheating of course! ;))

 NB>>> I'm no programmer, but maybe that's exactly what it's defined
 NB>>> for, in the very rare case that it would need to be used?

 WV>> For the 'second' it is. But I've never heard of a leap 'minute'. And
 WV>> google doesn't know about it either...

 NB> Apparantly the link I provided as well as the bibliography in the
 NB> proposal know about it, though. *shrug*

I think the ABNF format is overly complex to specify such simple things. I
don't see any RFC's mentioned in the bibliography only other ftsc documents.
And FTS-0001 does it in a simple clear way:

      HH         = "00" | .. | "23"
      MM         = "00" | .. | "59"
      SS         = "00" | .. | "59"

Why not stick to that?

Or if you want to be more specific, take as example rfc3339:

   time-hour       = 2DIGIT  ; 00-23
   time-minute     = 2DIGIT  ; 00-59
   time-second     = 2DIGIT  ; 00-58, 00-59, 00-60 based on leap second
                             ; rules

And none specify that a minute can be 60 !


Bye, Wilfred.


--- FMail-W32 1.73.14.71-B20170123