Obsolete mixed command and data, mixed data and response

Talluto, Dennis dennis.talluto at eds.com
Thu Nov 19 08:25:08 PST 1998


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* "Talluto, Dennis" <dennis.talluto at eds.com>
*
Bob,

I completely agree with your comments regarding the need to standards that
are completely and unabiguously documented.

So let's start a New Year's Resolution for 1999 -

No more T10 and T11 proposals/ballots/standards without STATE DIAGRAMS.

This surely will stamp out YEARS of FUD !!

Dennis Talluto
EDS/Infrastructure Systems Templates
office: 248.853.3521
email:  dennis.talluto at eds.com


> ----------
> From: 	Bob Snively[SMTP:Bob.Snively at ebay.sun.com]
> Sent: 	Wednesday, November 18, 1998 3:56 PM
> To: 	Gene_Milligan at notes.seagate.com; Bob.Snively at ebay.sun.com;
> Douglas.Hagerman at digital.com
> Cc: 	fc at nsco.network.com; t10 at symbios.com
> Subject: 	RE: Obsolete mixed command and data, mixed data and response
> 
> 
> Sorry Doug, I really have no nefarious agenda here.  The function is
> poorly, incompletely, and ambiguously documented, which is true of most
> of the "creeping elegance" features that we have allowed into our 
> documents and which I have, with far too little success, attempted to
> stamp out every time I have found a good target.  Their documentation is
> poor because nobody has tried to implement them and been forced to ask
> the hard question, "what does that really mean?".  
> 
> To make a product interoperable, robust, low in cost, easy to test, and
> high in performance requires that the standards that define the interface
> be simple and clean.
> 
> If this particular element of creeping elegance has struck your company's
> fancy as being particularly useful, then I would expect to receive
> a "please do not remove this function" message together with a dozen
> pages of comments and criticisms so that we can document it 
> unambiguously in the standard.  This in fact is exactly what has happened
> with most "creeping goodness" features that achieved the coveted goodness
> label,
> including such examples as tagged queueing, arbitrated loop, concise 
> exchange status ELS's, persistent reservations, high speed parallel 
> transfers, Command Reference Numbers, and FCP_CONF.
> 
> Bob
> 
> > 
> > I agree with Gene that the recent wholesale effort to eliminate
> > "obsolete" or "unused" functions from established standards is an
> > objectionable practise. For one thing, it puts yet another burden on all
> > users of the standards because now we need to keep track on a monthly
> > basis to make sure that something we're using doesn't disappear. For
> > another thing, the method used seems to have some of the flavor of a
> > fishing expedition. What if I'm using this and don't want to tell you?
> 
> 
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com





More information about the T10 mailing list