<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Riaz,<div><br><div><div>On Aug 6, 2012, at 1:36 AM, Riaz K Tayob <<a href="mailto:riaz.tayob@gmail.com">riaz.tayob@gmail.com</a>> wrote:</div><blockquote type="cite">
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
<div bgcolor="#FFFFFF" text="#000000">
David Thanks for the clarification. <br></div></blockquote><div><br></div>I hope it was helpful.</div><div><br><blockquote type="cite"><div bgcolor="#FFFFFF" text="#000000">Parminder, ...</div></blockquote><div><br></div><div>Not to intrude on your discussion with Parminder, but could you clarify:</div><div><br></div><blockquote type="cite"><div bgcolor="#FFFFFF" text="#000000"><b>no technical
administration without representation</b></div></blockquote><div><br></div><div>In the context of ICANN, the following technical resources are coordinated:</div><div><br></div><div>- Protocol parameters (protocol identifiers, service identifiers, etc.)</div><div>- IP numbers (IPv4 addresses, IPv6 addresses, Autonomous System numbers)</div><div>- Domain names (top-level domains, root management)</div><div><br></div><div>Out of honest curiosity, in your view, which of these are administered without representation? (I have my own list but am interested in yours)</div><div><br></div>Also, just for clarification:</div><div><br><blockquote type="cite"><div bgcolor="#FFFFFF" text="#000000">You see because for single rooters, whose belief - and it is a
belief - the Chinese proposal should be scary. </div></blockquote><div><br></div>What do you think the right answer should be when an application (like a web browser) looks up "<a href="http://www.gmail.com">www.gmail.com</a>"?</div><div><br></div><div>In my view, the right answer would be the IP addresses Google has assigned to the web servers for "<a href="http://www.gmail.com">www.gmail.com</a>".</div><div><br></div><div>In the "Chinese" proposal (by which I presume you mean <a href="http://tools.ietf.org/html/draft-diao-aip-dns-01">http://tools.ietf.org/html/draft-diao-aip-dns-01</a>), the answer would depend on whether "<a href="http://www.gmail.com">www.gmail.com</a>" resolves in the "local" root. If it doesn't, a box on the edge of the national network would append a suffix (e.g., "<a href="http://www.gmail.com.A">www.gmail.com.A</a>") and submit that query to the "global" root for resolution. The proposal implies that there could be any number of suffixes, presumably corresponding to countries (perhaps they could use ISO-3166 strings?), but does not specify how the box on the edge of the national network is supposed to figure out which suffix to use. One could imagine an implementation that orders suffixes based on politics, e.g., country "A" is more friendly to us than country "B", so we'll query their names first, etc.</div><div><br></div><div>While I can see the attraction of such an approach for folks who want to exert control of what is accessible (by domain name) to the people within the tightly controlled national network or who wish to penalize sites in other countries they don't like, it raises a myriad of technical implementation issues (resiliency, latency, caching, etc). Oh, and I think there might be a few issues related to freedom of expression, but I acknowledge that I might have some non-technical biases in that area.</div></div><div><br></div><div>As such, the "Chinese" proposal isn't really scary, rather it yet another reiteration of the "same old stuff" (I think I first heard a proposal similar to this back in the late 80s) and one that isn't particularly well specified or actually even implementable.</div><div><br></div><div>Regards,</div><div>-drc</div><div><br></div></body></html>