Dear Alireza,<div><br></div><div>Thanks for your reply.</div><div><br></div><div>As far as I understand, ZWJ/ZWNJ and ß are similar in that they are treated differently in IDNA 2003 vs. 2008, and present similar problems (new distinctions of domain names or failure to connect).</div>
<div><br></div><div>From the user's perspective, these cases are a bit different though: As you point out, the joiners are necessary for proper spelling of some words and names in a number of languages, and it is possible to write regular expressions to reasonably approximate the contexts where that's the case. In contrast, the "sharp s" is really just a ligature of two differently-shaped 's' letters. In all but government usage (registration of names of persons), the difference between sharp s and ss can be ignored, and often is. There is no way to create a contextual rule for when to allow a sharp s. At best, one could use a dictionary, but that does not scale and would only upset users who don't know or disagree with the standard spelling.</div>
<div><br></div><div>Best regards,</div><div>markus<br><br><div class="gmail_quote">On Wed, Dec 10, 2008 at 10:29 PM, Alireza Saleh <span dir="ltr"><<a href="mailto:saleh@nic.ir">saleh@nic.ir</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Dear Markus,<br>
<br>
We have the same problem for Arabic-Script character ZWNJ ( Zero With Non-Joiner ) . IDNA2003 says that this character should be removed before generating the A-label. But in IDNA2008, this character categorized as CONTEXT because its unicode's property is 'joining'. As it is a necessary character for some languages using Arabic-script, it is possible for us to define a safe contextual rule for it and use it in IDN lables. Without having a contextual rule, IDNA2008 treats the same as IDNA2003 in case of ZWNJ. I don't know if it is possible to propose the same solution for 'eszett' and the other mapping characters.</blockquote>
<div><br></div></div>
</div>