Re: [apps-discuss] Unified User-Agent String

[Andreas Petersson <[e-mail]>]

On Tue, 10 Jul 2012 16:05:12 +0200
Mateusz Karcz <[e-mail]> wrote:

> Good afternoon! I'm Mateusz Karcz. I'm author of "Unified User-Agent String (UUAS)" I-D (draft-karcz-uuas-00). I had sent it to RFC Editor as Independent Submission, but after I got suggestion from ISB to publicate it by appsawg, because it extends application layer protocol (HTTP) header. Can I develop that draft with You?
>      Sincerelly
>      Mateusz Karcz

Some quick comments on the draft:

1. Is this supposed to be put as value in the current
   "User-Agent"-header field, or is this a new header field?
   If this is to be put in the "User-Agent"-field, how does
   this relate to Section 10.10 of draft-ietf-httpbis-p2-semantics-19
   or 14.43 of RFC2616?


2. I think the use of sub-sub-sub-sections make the structure a
   bit... strange. :-)


3. What is security level? How is it used? References? Should it be
   possible to extend the values?


4. I think it would be better if you use productions from RFC5234,
   instead of specifying ranges of hexadecimal-values.

   I also believe that you can write:
     uuas = agent [ SP engine ] 
   instead of:
     uuas = agent *1(SP engine)


5. RFC2068 is listed as a normative reference, but it is not mentioned
   in the text. Further, it is obsoleted by RFC2616 since like over ten
   years ago. :-) RFC2616 is also to be obsoleted by httpbis pretty
   soon, I assume.



Best regards,
 Andreas