This message is from the T13 list server.


Nathan,

The purpose of the standards is not to make the negotiation between vendor
and customer easier.  If that was the purpose we could all just get into the
room and hash out a common purchase specification - and be thrown in jail
for anti trust violations.

What people seem to forget is that the market is the default mechanism for
all of this.  You actually have to have a good, positive reason to bypass
the market and put something in a standard.  That's why requiring something
in a standard takes so long, it subject to so many reviews, and in practice
requires the near unanimous consent of the affected industry.

Underlying all of this is the belief that market mechanisms in the long run
are better at promoting the common good.  But just like in the rest of the
economy, there are always issues that can cause temporary disruptions.  It's
the price we pay.

Jim


-----Original Message-----
From: Marushak, Nathan [mailto:[EMAIL PROTECTED]]
Sent: Thursday, March 21, 2002 9:12 AM
To: 'Hale Landis'; [EMAIL PROTECTED]
Subject: RE: [t13] vendor-specific as good as optional, outside the
o.s.?


This message is from the T13 list server.


I agree that those commands do not offer much value these days.  Although,
I'm curious to know what ATA is gaining by obsoleting commands?  You keep
the current specification a little more clean, but you confuse the history
of the specification.

As mentioned in the previous email, obsoleting commands makes the dialog
between the host and drive more difficult.  Is it possible that for these
obsoleted commands a drive vendor may return an error?  If so, you have now
required all host drivers to change their code to stop using these commands.
At a minimum, they may need to be aware that the command may not succeed and
appropriately ignore the possible error.  It seems to me this leaves a lot
of room for ambiguity.

If everybody only ever had to worry about the most current incantation of
the spec, then obsoleting pieces of the spec isn't such a concern.  The
problem is that people always have legacy requirements.

Regards,
Nate

-----Original Message-----
From: Hale Landis [mailto:[EMAIL PROTECTED]]
Sent: Thursday, March 21, 2002 9:27 AM
To: [EMAIL PROTECTED]
Subject: Re: [t13] vendor-specific as good as optional, outside the
o.s.?


This message is from the T13 list server.


On Thu, 21 Mar 2002 08:32:22 -0700, Pat LaVarre wrote:
>This message is from the T13 list server.
>Hang on a minute.  There is a small gap between 
>Optional and Obsolete, but a wide wide wide gap 
>between Optional and VendorSpecific
>The point of keeping Seek, Read/WriteLong, Recalibrate, 
>etc. etc. Optional is to give device folk a chance of 
>implementing something host folk want in advance.

My point is this... Making RECALIBRATE, SEEK and FORMAT TRACK
obsolete (and mostly implemented as a NO-OP these days) does not
break anything. These commands don't really have any value on a
modern drive. R/W LONG is a slightly different story... If these
commands are REQUIRED and if they are really going to be used for
anything PRODUCTIVE then they probably need to have a vendor specific
implementation because the old MFM implementation is probably
inadequate.

>By quietly neglecting to carry forward these commands 
>past the 28 and 32 bit Lba limits, T10 & T13 have in 
>effect made it more difficult than it was for host & 
>device folk to cooperate.

Nothing quiet about how T10 and T13 operate. Again... If you are
using ATA or SCSI devices then you need to be watching what T10 and
T13 are doing. And you need to complain or make alternative proposal
when you see things happening that you do not like.



*** Hale Landis *** www.ata-atapi.com ***

Reply via email to