Re: Proposed announcement to rest of IETF

[Costa Sapuntzakis <csapuntz@cisco.com>]

On Tue, 22 Feb 2000, Brian E Carpenter wrote:

> Costa, without requirements analysis you will not get anywhere in the IETF.

Can you point me to some documents that indicate what goes into
a requirements analysis? I'm fuzzy on the concept.

> The way you have phrased it rings lots of alarm bells  -  it reads as if you are
> trying to set up SCSI/TCP as an architectural competitor or replacement for 
> NFS etc, but that is absolutely not the case. That's why the requirements analysis 
> is the first step.
> 
>   Brian
> 

Phrasing aside, here's what I meant for the record:

The intent was to say "SCSI is joining the IP network
bandwagon 15 years late." AFS and NFS have been doing IP networks for
years. Now, SCSI wants to play over IP networks. Our goal is to make
SCSI work over all kinds of IP networks.

The goal is not to replace NFS. In fact, some of the proposals out of
this proposed working group may actually help NFS. Again, though, the
principle goal is to make SCSI work over IP networks.

-Costa

> Costa Sapuntzakis wrote:
> > 
> > > 3. I think it is very odd to refer to SCSI as if it is a protocol
> > > at the same level as NFS, CIFS (which is proprietary), and AFS and DFS
> > > that you don't mention. SCSI is in a completely different class, but you
> > > only half-say this. You will get a lot of flak over this in the IETF,
> > 
> > I agree that if feels odd at first to see SCSI referred to in the
> > same sentence as NFS. However, once both are travelling over the network,
> > they're not all that different, except that SCSI has a simpler model
> > of sharing.
> > 
> > I can access a disk as one large NFS file and it will look a lot like
> > SCSI.
> > 
> > > and you don't at all mention the requirements analysis that will
> > > need to be Phase One.
> > 
> > I don't think everybody agrees that requirements analysis is phase one.
> > 
> > -Costa
> 
> 
>