New drafts

[David Lapp <[e-mail]>]

Hi,

I just took a look at the new IDs for DHCP (Authentication,
FQDN, and option 127). My compliments on their brevity :-)

I have no complaints about the Auth and FQDN docs. They
seem straight forward.

I have a question and perhaps a suggestion for the option 127
doc. The definition says:

"  Option code 127 indicates that the DHCP option has a two-octet option
   code.  The format of these options is:


    Code         Len   Data...
   +-----+-----+-----+-----+-----+----
   | 127 | XXX |  n  |  d1 |  d2 | ...
   +-----+-----+-----+-----+-----+----

   Other than the two-octet option code, these options are encoded and
   carried in DHCP messages identically to the options defined in RFC
   1533 [2]. "

Perhaps I'm confused (it happens often :-) but this format doesn't
look "identical... to ... RFC 1533". That is unless XXX is the length
of the entire option ? I think it needs to be for compatability
with deployed clients. 

I'd suggest a format similar to the FQDN format:

                  subcode   sub
     Code  len   (2 bytes)  len   data ... 
    +-----+----+-----+-----+----+----+---+---+----------------------
    | 127 | n  | SC1 | SC2 | sn | d1 | d2| d3| ....
    +-----+----+-----+-----+----+----+---+---+----------------------

     ----+----+-----+-----+----+----+--------------
     ... | dn | SC1'| SC2'| sn'| d1'| ...
     ----+----+-----+-----+----+----+---------------

This preserves the options-encoded-in-options format we have used
before.

I've shown the "subcodes" as 16 bits. I'm not sure how others feel
about it but I'd prefer not to run out of option codes again.

Dave Lapp