RE: iSCSI base64 and 12-92

["ERICKSON,SHAWN (HP-Roseville,ex1)" <shawn_erickson@hp.com>]

I agree with Paul & Bill. I don't see the need for base64 encoding of
numbers or for iSCSI caring/knowing about big numbers, treat them as binary
and let the end consumers of the information handle any needed translation.


-Shawn

> -----Original Message-----
> From: Bill Studenmund [mailto:wrstuden@wasabisystems.com]
> Sent: Friday, May 24, 2002 10:19 AM
> To: Julian Satran
> Cc: Paul Koning; ips@ece.cmu.edu
> Subject: Re: iSCSI base64 and 12-92
> 
> 
> On Thu, 23 May 2002, Julian Satran wrote:
> 
> > If base 64 is neede for large integers there is no good 
> reason to test
> > that it is not used for short integers.
> 
> Julo,
> 
> You still haven't explained why we need base64 for large numbers. What
> security negotiation schemes are we using that need to exchange large
> numbers as numbers; where the scheme expects the number in 
> host byte order
> as opposed to a specific on-wire format.
> 
> More specifically you have not explained why the iSCSI parameter
> negotiation system needs to be able to deal with large base64 numbers.
> 
> If we actually have any protocol which strangely wants large 
> numbers in
> local byte order, why not have the keys in question defined 
> as exchanging
> a binary string which is the number in network byte order?
> 
> The advantage of the above is that we then can drop base64 as a number
> encoding scheme. The key negotiation system(s) need only deal 
> with base64
> binary strings, things base64 is good (no, great) for. And we 
> can support
> any cryptographic scheme we choose to in the future; the only 
> thing we're
> loosing is pain.
> 
> So what is wrong with the above?
> 
> I mean do we really want to have to support "MaxConnections=0b0Q=="?
> 
> Take care,
> 
> Bill
>