<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.59">
<TITLE>RE: Auth 2.0 Extensions: Namespace Prefixes</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->

<P><FONT SIZE=2>On 5-Jun-07, at 8:53 AM, Granqvist, Hans wrote:<BR>
<BR>
&gt;&gt; But it seems superflous: Since you cannot depend on args to<BR>
&gt;&gt; be ordered[1], you'll still need to iterate and match prefix<BR>
&gt;&gt; to values.<BR>
&gt;Martin's proposal seems like a minor improvement to me - iterating&nbsp;<BR>
&gt;thorough openid.ns.* or splitting the value of openid.extensions and&nbsp;<BR>
&gt;then iterating through the result seems roughly equivalent.<BR>
<BR>
Fair enough.<BR>
<BR>
I guess the question is:<BR>
<BR>
Is the cost of semantical additions (as I outlined) worth such a minor<BR>
improvement in processing?<BR>
<BR>
-Hans</FONT>
</P>

</BODY>
</HTML>