Re: bug in date: --rfc-3339=seconds and --rfc-3339=ns options do *not* output timestamps in RFC 3339 format

[Paul Eggert]

Romain Lenglet <address@hidden> writes:

> This interpretation of the NOTE in section 5.6 is not shared by 
> everyone. e.g. cf. 
> http://validator.w3.org/feed/docs/error/InvalidRFC3339Date.html

The explanation for this URL explicitly states that it is applying an
additional check, over and above what RFC 3339 requires.  "The content
of this element MUST conform to the "date-time" production as defined
in RFC 3339. In addition, an uppercase "T" character MUST be used to
separate date and time...".  So this URL underscores the point that
spaces are allowed by RFC 3339.

> http://lists.infodrom.org/infodrom-sysklogd/2003/0023.html

This reference also points out that they are specifying a subset of
RFC 3339, and that one of their extra restrictions (and not the only
such restriction) is that a "T" must be used instead of a space.

> And in this NOTE, replacing "T" by a space is only one example 
> ("(say)"). This NOTE then allows to replace "T" by *(say)* an 
> underscore. How will you parse that?

We don't parse arbitrary dates, but we do try to parse the dates that
we ourselves generate.

> Since this NOTE is ambiguous

Not really.  None of the references you've mentioned have
misinterpreted it.  They all agree that a space is allowed
in RFC 3339.

> making "T" mandatory seems to  be the general consensus:

Not here.

> Could you point me to other discussions or standards which allow 
> to replace "T" with anything else? (except GNU date, of course)

A simple Google search brought up this in the first page.

http://en.wikipedia.org/wiki/ISO_8601
http://www.mdspacegrant.org/iso_dates.html
http://www.cs.tut.fi/~jkorpela/iso8601.html
http://bugs.darcs.net/issue31
http://www.merlyn.demon.co.uk/datefmts.htm

The last reference is the most interesting, since it claims that ISO
8601:2004 allows a space!  I haven't verified this.