Re: Fwd: Last Call Draft 01

On Monday 07 April 2003 17:12, Joseph Reagle wrote:
>   http://www.w3.org/2001/XKMS/Drafts/XKMS-20030331/xkms-part-1.html
>   http://www.w3.org/2001/XKMS/Drafts/XKMS-20030331/xkms-part-2.html
>   http://www.w3.org/2001/XKMS/Drafts/XKMS-20030331/issues.htm
>
> Here is the tar file for the last call draft with the documents in html,
> also for convenience here are two pdfs generated on my machine.

I can verify that the html files print out fine in my environment; the bugs 
in the last version aren't present. Otherwise, reviewing the text I'm 
prepared to say I agree this text meets the requirements we set out to 
satisfy. 

When I looked it at with respect to publishing as a Last Call I noted the 
following nits. We just talked these over on the phone but I'll capture 
here for reference. (Generally, the documents look great and some of these 
crept in when you fixed earlier more troublesome bugs.)

1. Do we want two specs, or two chapters of one? For example SVG just 
published a title page with many chapters:
  http://www.w3.org/TR/SVG11/
But SOAP and Schema are all seperate but related specs. We agreed these 
should be two seperate specs each with their own URI.
2. Do we name them "part-1/part-2" or core/bindings? We agreed to call them 
Core (Part 1) and Bindings (Part 2) and to update the title/h1 elements of 
each document accordingly. However, on further reflection, I think we 
should just call part 1, "XKMS", and part 2 the "XKMS Bindings"
3. Do we bother to give the bindings a version? For consistency's sake I'm 
including it version 2.0 too...(?)
4. Do we need to in-line the whole schema at the end? We agreed it's not 
critical, but we'll want to provide a link to the schema resource anyway 
and removing it could cut down on the size.
5. In Binding (Part 2), why is the SOAP message in yellow. We agreed that's 
an editorial error and its class class needs to be changed.
6. In a few places, "Key" is capitalized as is "Registration Service", but 
in other places they aren't. Additionally, I noted that 
some elements appears as between angles some don't. And at least one case 
missed it's open angle:
  [338] The <RSAKeyPair> element specifies the public and private 
  parameters of an RSA Key Pair. The contents of the RSAKeyPair 
  element are specified in [PKCS1]. The RSAKeyPair> element contains 
  the following parameters:
7. There are some borken links that point to:
  "http://www.w3c.org/2001/XKMS/Drafts/XKMS/#ref-KEYWORDS"
instead of
  "#ref-KEYWORDS"
Including:
  Lines: 114, 423, 430, 445
  Broken fragments and their line numbers: 
    ref-KEYWORDS: 430
    ref-XML-NS: 445
    ref-XML-schema: 423
8. As noted above, a couple of the links to bibliographic entries include 
"ref-" but others don't.


Then there were to minor substantive issues that can probably be easily be 
addressed and wrapped into the last call:

9. Should we say something about  our use of the terms validate and 
verify/unverifiable? I generally try to go by RFC2828, and in xmldsig we 
had a couple specific "validate" definitions. In XKMS one might send a 
"validate" request but get a UnverifiedKeyBinding. So perhaps a sentence on 
whether we use these interchangable or not would be worthwhile.
10. My suggestion with respect to the UseKeyWith didn't make it into this 
version:
  http://lists.w3.org/Archives/Public/www-xkms/2003Mar/0016.html

I've attached a zip file with the target directories and the opening (head, 
status, abstract) of the target "Overview.html" files.