Print

Print



On Oct 11, 2015, at 10:18 AM, Saašha Metsärantala <[log in to unmask]> wrote:

Hello!

So, I would suggest that we follow the two-digit ISO 8601 for centuries
I do not agree with that, because of:

- the problem with years 1 - 99 CE, which could be missinterpreted as centuries (or the other way round)

A fair point. Using the +/- prefix allows distinguishing them, but there is a reasonable chance of confusion.

Another point is that in ISO 8601 form, it is currently impossible to specify a BCE century — this is clearly an oversight that should be resolved. 8601 also lacks a form for decades or millennia, which could be potentially useful.


- the lack of unambiguous definition of the concept of century (as lengthly explained on this list a few yeras ago)

Also a good point, though in both EDTF and ISO 8601, I don’t believe anyone would assume that “19xx” or “19” would span 1901-2000, regardless of how they define “century.”


Let's use ranges for centuries (and decades) to improve readability and avoid ambiguously defined centuries (and decades), I suggest.

This sounds like an excellent idea. Using an EDTF interval (e.g., “1900/1999”) removes any potential ambiguity, provides flexibility for decades and other ranges, is reasonably concise, and seems to tick the boxes for ISO (no letters, ASCII only, etc.).

Users of ISO 8601 must have run into these issues, perhaps they should ditch their century form and follow suit?

Regardless what form is settled on, I would love to see the final EDTF and new ISO 8601 match up as closely as possible, rather than having separate ways to convey the same concept. I’m sure they won’t want to adopt everything in EDTF, but this seems like an opportunity to fix a few bugs in their own spec.

—Richard