de facto obsolete sbc-2 ops - 23h

Pat LaVarre LAVARRE at iomega.com
Wed Feb 20 07:54:08 PST 2002


* From the T10 Reflector (t10 at t10.org), posted by:
* "Pat LaVarre" <LAVARRE at iomega.com>
*
> "Elliott, Robert" <Robert.Elliott at COMPAQ.com> 02/19/02 12:51PM

> ... 23h ... SCSI-2, SBC-1, and SBC-2
> all mark it as vendor-specific for block devices

Scsi 2 I remember, some of the Sbc's I saw recently.

> MMC3 defines 23h as "READ FORMAT CAPACITIES".
> ... Is it being used on block devices too? 

Me, I read paragraph 12.7.10 of the October 1996 r1.1 of Sff 8070i to say devices that plug 'n pray like the Compaq Ls-120 flavour of removable direct access have to support op x23 ReadFormatCapacities as part of Scsi over Ide (aka Atapi).

I know of specific device vendors who, with the advent of Win95B Atapi, abandoned traditions involving op x23 in favour of this de facto Wintel definition.

I had heard in particular that Microsoft uses op x23, particularly when erasing a disk (aka formatting, but not involving op x04 Format).  But if it's not generally accepted that op x23 has a de facto definition from Microsoft, then I need to go collect some bus traces before I can answer more fully.  For example, it might be that only Win95/98/ME use this op.

> The usual rule is once vendor-specific, always vendor-specific.

I hope I recognise the evil of changing to defined from vendor-specific.

For example, my personal history of pain includes the change to defined from vendor-specific that t13 gave to "word" 128 of their Ata op xA1 Identify data (analogous to Scsi op x12 Inquiry data).

Here I mean to suggest a change to obsolete from vendor-specific.

I'm thinking we want to discourage people from creating vendor-specific software whose existence makes significant the somewhat undocumented, but de facto, Wintel definition of op x23.

Thanks again in advance.    Pat LaVarre

>>> "Elliott, Robert" <Robert.Elliott at COMPAQ.com> 02/19/02 12:51PM >>>
* From the T10 Reflector (t10 at t10.org), posted by:
* "Elliott, Robert" <Robert.Elliott at COMPAQ.com>
*
> -----Original Message-----
> From: Pat LaVarre [mailto:LAVARRE at iomega.com] 
> Sent: Tuesday, February 19, 2002 10:52 AM
> To: t10 at t10.org 
> Subject: de facto obsolete sbc-2 ops
> 
> Q1) Do sbc-2 folk hang out here at <t10 at t10.org>?
> 
> > The following operation codes
> > are vendor-specific: ... 23h ...
> 
> Q2) I wish.  Has the time come to acknowledge Microsoft's de 
> facto definition of this op?  Can we make x23 "obsolete" 
> rather than "vendor-specific"?

The usual rule is once vendor-specific, always vendor-specific.  

MMC3 defines 23h as "READ FORMAT CAPACITIES". Is it being used on block
devices too? SCSI-2, SBC-1, and SBC-2 all mark it as vendor-specific for
block devices. 

...


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




More information about the T10 mailing list