Re: The IYYY mess again

On Dec 29, 2014, at 7:06 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> In bug #12367
> http://www.postgresql.org/message-id/20141229031218.8013.51171@wrigleys.postgresql.org
> we see yet another iteration of somebody trying to combine to_char's
> IYYY specifier with regular Gregorian MM/DD fields.
>
> It occurs to me that this is largely our own fault, because the fine
> manual just defines IYYY as "ISO year". I'm sure the typical newbie
> thought process is "that sounds like a standard year, I'll use that".
> There is a warning against combining IYYY with MM/DD, but it's buried
> in trivia far down the page.
>
> I did a bit of googling and came across
> http://en.wikipedia.org/wiki/ISO_week_date
> in which this construct is called an "ISO week-numbering year".
>
> Not having a copy of ISO 8601, I'm not sure if that's the standard's
> terminology; but ISTM that if we consistently referred to the Ixxx
> format specifiers as "ISO week-numbering foo" then this type of error
> might become a little less attractive.
>
> Objections, better ideas?

I've seen this problem a few times on IRC too. An explicit warning / cross
reference on IYYY that the user almost certainly wants YYYY instead might
be even better.

Cheers,
Steve