Re: [dhcwg] Re: one more comment about the lifetime option

>>>>> On Fri, 3 Sep 2004 09:56:17 +0200, 
>>>>> Stig Venaas <Stig.Venaas@uninett.no> said:

>> Ah, okay, I was wrong.  The authentication option is a variable-length
>> option but does not take a sub-option according to RFC3315.
>> 
>> > But in any case, its no big deal to specify that the option is restricted to
>> > the message level.
>> 
>> That's correct.  But at the same time, I believe we all agree that
>> there is no reason to allow the lifetime (or refresh timer) option
>> appearing as a sub-option.

> Yes, I don't mind adding some wording to the fact. It's against the
> whole idea to use it as a sub-option.

> I must say though that I'm a little bit surprised this came up. I'm
> not that into the world of DHCP, but I don't recall seeing specs for
> other options saying this, and I had thought that they should only
> be used as sub-options if specified as such. Or is the idea that in
> general, any option could be used as a sub-option?

I think your understanding is basically correct.  RFC3315 says in
Section 22 that:

   Unless otherwise noted, each option may appear only in the options
   area of a DHCP message and may appear only once.

The term "options area" does not seem to be officially defined in the
RFC, but it should be natural to interpret this to mean areas which
are not a sub-option of a separate option.

But at the same time, the RFC explicitly mentions the default
restriction in some places.  For example, in Section 22.5 it says:

   An IA_TA option may only appear in the options area of a DHCP
   message.

And,

> I think though that the IRT (lifetime) option might appear to make
> sense as a sub-option, while many other options pretty obviously
> don't make sense. So I'm happy to add the wording, just to avoid
> possible misuse.

this is exactly my point.

 I believe it depends on whether we want to avoid possible misuse
 explicitly.  And then it depends on how we asses the possibility of
 the misuse.  Of course, the sense on this may vary among people (so I
 won't push my preference).  But if we recall that we are talking
 about this in the context where we are trying to add a note that a
 single IRT (or lifetime) optin should be common for all possible
 information resources, I personally think it's reasonable to add an
 explicit note here.

					JINMEI, Tatuya
					Communication Platform Lab.
					Corporate R&D Center, Toshiba Corp.
					jinmei@isl.rdc.toshiba.co.jp

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg