[mdx] Fwd: comments on metadata-retrieval-protocol-03.html

[Chad La Joie]

Sending Ian's comments to the list so that everyone can see them.

-------- Original Message --------
Subject: comments on metadata-retrieval-protocol-03.html
Date: Wed, 5 May 2010 12:51:07 +0100
From: Ian Young <ian at iay.org.uk>
To: Chad La Joie <lajoie at itumi.biz>

Mostly typos.

2.4  Shouldn't the response contain something to say whether compression
has been used?

In 3.1 you use "proceeds" when you mean "precedes".

End of 3.1: "appended to path" --> "appended to the path", I think.

3.2.1 "adventagous" --> "advantageous"

3.2.1 "For example some network protocol will transmit" -->
"For example, some network protocols transmit"

3.2.1 "printable ASCII character" --> "printable ASCII characters"

3.2.1 "tranformation identifier" --> "transformation identifier"

4. "metadata agent-drive content" --> "metadata, agent-driven content"

5. "Responsed" --> "Responses"

6.1.1: you use {id} here where I'd have expected <id>.

6.2.1 the text uses ID where the template is id (lower case).  The {}
vs. <> thing applies here too.

7. "Effecient" --> "Efficient"

"Requester's SHOULD support" --> "Requesters SHOULD support"

"requester's SHOULD maintain" --> "requesters SHOULD maintain"

You don't indicate any rules by which a requester could know when it
should expire negative cache information.  I seem to recall there are
such rules for the DNS (minimum TTL from the zone SOA, from my reading
of section 4 of http://www.faqs.org/rfcs/rfc2308.html) so perhaps we
should say something explicit here rather than leaving it completely
open.  Not sure what, though.

	-- Ian