On 20 Apr 2010, at 22:27, Derrick Brashear wrote:

> As before, I've written up a draft based on the 2004 Stockholm AFSig hackathon
> discussion of the PTS alternate authentication names proposal, as
> modified based on further feedback and the 2009 Edinburgh Hackathon.
> Comments welcome and encouraged.

I've finally had a chance to review this. I've split my comments into ones of 
substance, and ones of style.

Substance:

> 10.4.  Authentication Name Type Rewriting

I'm still uneasy about requiring the rewriting of GSSAPI obtained Kerberos 
names to use the Kerberos name type. If we believe that GSSAPI is the future, 
then I would prefer that we use the GSSAPI exported name for
all GSSAPI mechanisms, rather than special casing Kerberos.

Style:

>    Some deployments provide several mechanisms to obtain AFS
>    authentication; While mappings between Kerberos 4 and Kerberos 5
>    [RFC1510] authentication names allow use of most Kerberos 5
>    deployments with AFS, supporting more than a single realm requires
>    matching usernames in all realms; Additionally, support for other
>    systems is not provided at all.

I'm not sure about the readability of this paragraph - in particular the use of 
the semicolon.

> 3.  Background information on operation of AFS

Whilst this background information is of use to a reader inexperienced with 
AFS, I'm not sure that every draft we produce needs to explain what AFS is, and 
how it works. Given that AFS novices are probably not the intended audience, 
I'm not convinced that this section is required.

>  permitted for administrations

for adminstrators?

> section 7above

missing space

Apart from those, looks good to me.

S.


_______________________________________________
AFS3-standardization mailing list
[email protected]
http://michigan-openafs-lists.central.org/mailman/listinfo/afs3-standardization

Reply via email to