On Mon, Feb 21, 2005 at 09:18:07PM -0800, Erik van der Poel <[EMAIL PROTECTED]> wrote a message of 218 lines which said:
> As George points out, the registries are going to have to start > filtering IDN lookalikes, otherwise they will eventually face > lawsuits from the "big boys" (as George so delightfully puts it).
Quite the opposite: according to our lawyer, if the process is completely automatic (no human eyes involved), you can disclaim any responsability. But if you do screen, you accept a liability if the screening fails (and it will fail, trying to catch homographs is an hopeless task).
I seriously doubt that european registries, which all moved from a "screen every domain to check if it is legal" model to a "accept anything" model in the '90s will go back...
Full agreement. Now, only for the reasons explained below, I am ready to test and propose several ccTLDs a filtering experimentation.
1. I filter the registered names (against foul names, blocked names) at registration level, before accepting payment.
2. the filtering will therefore be on xn--entries strings.
This being accepted:
1. could someone point a "C" source code to carry what has to be carried to filter out the dangerous names? Please help: I have no resource on this.
2. could someone list all the Unicode codes to blacklist that way?
3. could someone point a Perl code to use to enter a IDN and to get it properly punycoded, which could use such a list.
My rationale is that I only want to protect my own operations from confusion. I will only extend the description of the non-authorized characters in the terms and conditions. If this works properly I will describe the solution and experience in a for information Draft and a request to the IANA to list a Unicode black list.
jfc
jfc
