Disallowing code points

[Gervase Markham]

On 16/07/09 02:58, Chris Wright wrote:
> There have been some discussions lately about explicitly disallowing
> specific code points
>
> I believe that the decisions to disallow specific code points should
> be kept to an absolute minimum, with strong technically justifiable
> reasons being required for specific singling out of code points.

In principle, I agree. However, Firefox currently has a "never under any 
circumstances" list of banned codepoints, full of points which look like 
protocol characters. It would be beneficial for interoperability if this 
list (or that part of it not excluded by other IDNA2008 changes) were 
standardized, based on that criteria of resembling protocol characters. 
We can't remove that safeguard, so it's better if everyone does the same 
thing and we don't have domains which work in one context and not another.

Gerv