Re: [Sip] Last Call: draft-ietf-sip-ipv6-abnf-fix (Essential correction for IPv6 ABNF and URI comparison in RFC3261) to Proposed Standard

[Paul Kyzivat <[e-mail]>]

inline

Vijay K. Gurbani wrote:
> Attending to pending email ... sorry for the delayed response.
> 
> On 03/05/2010 06:17 PM, Paul Kyzivat wrote:
>> I see this is very close to done.
>> I'm sorry to only be looking at it now.
>> While I don't see a problem with what is in this,
>> I do see a logical omission:
>>
>> This calls out that comparison of the binary forms of the ip address,
>> and this fixes a problem with ipv4 as well as ipv6.
>>
>> A similar problem exists for the port number:
>> are <sip:foo@bar:1234> and <sip:foo@bar:01234> the same???
>>
>> ISTM that binary comparison should also be used for port numbers.
>>
>> But is it worth pulling this back to fix that???
> 
> Paul: <sip:foo@bar:01234> could also be interpreted as the digits
> comprising the port "1234" to be in base 8 (leading 0 signifies
> an octal base.)

Well, it *could* mean that. (I was never a fan of that notation even 
from day one in C.) There certainly is nothing in 3261 that would 
suggest that the octal convention applies to hostport in sip URIs. The 
ABNF currently is:

port =  1*DIGIT

If the octal notation was intended to apply, then I would expect that 
the ABNF would be:

OCT-DIG =  %x30-37
NON-OCT-DIG =  %x31-39

port =  (NON-OCT-DIG *DIGIT) | "0" *OCT-DIG

If there is any question that some might believe the octal notation 
applies, then that is more reason to say something.

> Writing leading zeroes for IPv4 address is not prevalent; by the
> same token, representing ports in octal base is not prevalent either.
> So I am inclined to let sleeping dogs lie.
> 
> However, if the sponsoring AD or anyone monitoring this list feels
> strongly, I don't mind adding a sentence or two to this effect in
> the draft.

I don't feel strongly about it. But now that you mention it, I think I 
remember some question about octal ports coming up on one of the mailing 
lists some time in the past.

I also don't feel *strongly* about it. Lets see if anybody else has 
something to say about it.

	Thanks,
	Paul