AW: Mapping?

[Mark Davis ☕]

Nobody particularly wants to perpetuate mistakes. But for any breaking
change, we have to be aware of both the benefits *and* the costs.

For example, in an area you're familiar with, one can argue that the
placement of the Greek and Hiragana blocks (and a few others) in Unicode is
sub-optimal, because they are not aligned on 128-unit boundaries; that
affects certain kinds of compression and table lookup. Even though arguably
it was a mistake, the value of fixing it would not outweigh the
*incredibly*high cost of fixing it. There are hundreds of like cases
in Unicode alone,
where *if* we could do it all over again -- knowing what we know now -- we
would have done things differently, but we simply cannot change now.

That's why for protocols such as Unicode -- or domain names -- where there
are massive amounts of data and huge numbers of programs using the existing
protocol, and where there is a long transition period before everything
could possibly switch over, we have to assess any breaking change extremely
carefully so that we are absolutely certain that the value outweighs the
cost. If, on the other hand, we can come up with an effective transition
plan for these 4 characters (and I'm hoping we can), then it would be
feasible to change.

Mark


On Thu, Dec 3, 2009 at 08:44, Michael Everson <everson at evertype.com> wrote:

> On 3 Dec 2009, at 10:37, Debbie Garside wrote:
>
> > My philosophy is always to go with backward compatibility if at all
> > possible.
>
> My philosophy is that if a mistake was made in the past it should not
> be perpetuated.
>
> Michael Everson * http://www.evertype.com/
>
> _______________________________________________
> Idna-update mailing list
> Idna-update at alvestrand.no
> http://www.alvestrand.no/mailman/listinfo/idna-update
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.alvestrand.no/pipermail/idna-update/attachments/20091203/d3879195/attachment.htm