Shawn.Steele at microsoft.com
Thu May 8 22:22:26 CEST 2008
There's no risk to cuneiform. Sorry, I'm not going to be convinced that there is.
Any risk would be along the lines of spoofing, and a simple test for mixed or expected scripts could provide an appropriate warning. Additionally any type of phishing or whatever filters aren't complicated by allowing cuneiform.
Apostrophe is completely different. It is also used as a string delimiter in some programming languages and has various other special meanings depending on how it's used. That reasoning could be used to argue against similar characters in other scripts, but not to discriminate against an entire script.
For spoofing Latin, cuneiform has far fewer problems than Cyrillic. Perhaps you could spoof Chinese, but Chinese is perfectly capable of spoofing itself, so I have a hard time buying into that concern. Heck, even in Latin who's going to notice that a dotless I is missing the dot in Microsoft? Or just use ì or í î, it's easy to miss in IE's address bar. Chinese can be very difficult to distinguish in common font sizes, even for native speakers.
I'm not saying that it’s hopeless anyway, so we should add 10x the number of characters, but most of the cuneiform characters are unambiguous, and the number with a potential for confusion is small.
From: Gervase Markham [mailto:gerv at mozilla.org]
Sent: Thursday, May 08, 2008 1:00 PM
To: Shawn Steele; idna-update at alvestrand.no
Subject: Re: Archaic scripts
Shawn Steele wrote:
> And IMO if those few users would also like a name in cuneiform, then
> that should be permitted. Of course it'd only be usable by a few
> people, but many other small groups have web sites in a
> friendly-to-them form that they only expect a few people to use.
Sites, absolutely. Domain names are an entirely different thing to sites.
Apostrophe is not permitted in domain names, and won't be under any
revisions we might make. Yet the number of people who might want to use
an apostrophe (for example, to have their surname as a domain) vastly
outweighs the number of people who want cuneiform. So, if "some people
will want it" is our only test, then apostrophe should be in.
However, it's not our only test, and rightly so. Every character we add
(assuming some registry permits it; and if none ever would, why add it?)
is a small additional risk - both for issues we know about now, and ones
that might appear in the future. Just because each one individually is
small doesn't mean we should add thousands of them (which is where the
logical conclusion of your argument is).
More information about the Idna-update