I am listening to you- I'm just not agreeing with you in the slightest.
There was a period of time where the person who was qualifying the drive
f/w for Sun (Jim Kahn) had a bunch of the drives accepted but then some
of them had a lot of workarounds done in sd itself (like our quirks).
This was in the period of 1990 or so thru about 1994 IIRC.

The information about which of these drives are around is harder to find
than it used to be. However, it's likely still to be the case certainly
as people *consolidate* older sun hardware. You'll get lots of cases of
people pulling drives from SS1s and SS2s (which did *not* have good
T-qing support) and drop them into an SS10 (which shipped *way* before
Zeus was finished).  After all, 3.5" narrow drives are just about only
useful these days as internal drives in older h/w like this.

So, the *sensible* engineering manager, planning for a release for a
class of machines which is *likely* to have broken drives makes the
feature an *option* with a RTF note that says "here's how to get the new
feature if you don't have older drives". This keeps a project within
budget and is more *likely* than not to make the installation experience
for *everyone* work well. The *other* approach is fraught with risk and
saddled with somewhat pointless assertions about "well, all bad drives
should be in the quirk table"- well, yes, they should be. But we *know*
such a list would not be exhaustive and we *know*, again, that it is
more likely than not somebody will hit the problem.

Note that this is *completely* the opposite of what we would do with the
alpha or i386 (e.g.) platforms as *those* platforms are *much* more
likely to have drives that do the basics quite well. The drives that do
*not* do well (e.g., some broken Quantums) *are* in the quirks table.

An alternative approach would be to follow your model and have the
people who fail to install just go install OpenBSD instead.

-matt

On Fri, 6 Sep 2002, Greg A. Woods wrote:

> [ On Friday, September 6, 2002 at 00:33:20 (-0700), Matthew Jacob wrote: ]
> > Subject: Re: 1.6 on IPC (tagged queuing bug still there)
> >
> > 
> > thirded, despite GAW's objections
> 
> You are not listening to the content of my objections, nor apparently
> the suggestions I would consider acceptable
> 
> This issue affects all drives of the claimed vintage on _all_ machines
> with _all_ kinds of scsipi-enabled drivers.
> 
> Sun systems are in fact the _least_likely_ systems to have drives which
> are susceptible to tagged queuing bugs.  Sun explicitly QA'ed drives
> because Solaris-2 was shipped with tagged queuing enabled and they
> explicitly document what users who've chosen buggy third party drives
> must do.  Sun users are very aware of these issues.
> 
> If this change is to be made I instist that it be made across the board
> for all scsipi-enabled drivers.  See if that flies before you go hacking
> at a very important performance feature in on a platform where it's
> least likely to have adverse affects.
> 
> IMNSHO it would be far more intelligent to simply gather up published
> lists of drive models which are known to have buggy tagged queuing
> support (eg.  from the other *BSD quirk lists, etc.) and to make a
> combined master list that can be used in the default scsiconf.c.
> That'll cover the vast majority of problem drives that might still be in
> service, and a simple warning note in the INSTALL documents will cover
> the rest.
> 
> -- 
> 								Greg A. Woods
> 
> +1 416 218-0098;            <g.a.woods@ieee.org>;           <woods@robohack.ca>
> Planix, Inc. <woods@planix.com>; VE3TCP; Secrets of the Weird <woods@weird.com>
>