+1<br><br><div class="gmail_quote">2009/9/16 Elisabeth Blanconil <span dir="ltr"><<a href="mailto:eblanconil@gmail.com">eblanconil@gmail.com</a>></span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Jean-Michel,<br>
<br>
I must say I do prefer FE73 rather than PRIVATE USE codes. This is<br>
because PRIVATE USE codes are within the ASCII limits (what may create<br>
confusion with poorly designed implementations) and may poorly print<br>
if it was actually printed. FE73 would react more clearly. IMHO the<br>
best would be to have an ISO 10646 SDO area reserved at the IUCG<br>
reference center taken on the UNASSIGNED area. I am sure that<br>
French.European majuscules (as quoted by Martin Dürst) are not the<br>
only metadata that we would need to carry throught codepoints.<br>
<br>
Elisabeth Blanconil<br>
<br>
2009/9/14 jean-michel bernier de portzamparc <<a href="mailto:jmabdp@gmail.com">jmabdp@gmail.com</a>>:<br>
<div><div></div><div class="h5">> Anyway, my problem is _not_ to know what IDNA will be able to do or not. My<br>
> interest is to make sure that IDNAPLUS is able to match users' expectations,<br>
> and also to be 100% IDNA conformant. Obviously, IDNAPLUS will do things IDNA<br>
> will not do. What I try to make sure is that the difference is only :<br>
><br>
> 1. to support IDNA2003 upper-cases in U-Label<br>
> 2. keeping PRIVATE USE ONE and PRIVATE USE TWO PVALID as PVALID, i.e.<br>
> respecting Unicode's respect for users.<br>
</div></div></blockquote></div><br>