** This is the quasi-official and semi-temporary T13 email list server. **
Let me correct some misstatements about the SFF Committee.
For starters, the only reason your disk drives are consistent in size with
their connectors in the same place is because the form factors were defined
that way by the SFF Committee. All the form factors except 1.8" are now EIA
standards, and the 1.8" drives are active projects at SFF with editors from
IBM and Toshiba.
> >When reading SFF8070 they refer to ATA/ATAPI-4, and they say they
> >are compliant with this standard. What about ATA-5 ?
The date on INF-8070 is November 1998. At the time the specifications was
developed it was current with the ATA standard of that time which was
ATA/ATAPI-4. No SFF or INF specification has been prescient and claimed to
be compliant with a standard that had not been written at the time it was
completed.
> The problem with all the SFF documents is that they mostly have been
> develop completely outside of the T10 or T13 community.
Rubbish. Until T13 decided to absorb all the SFF specifications which
involved ATA, the attendance at SFF Committee meetings listed the same
engineers as regularly attended T13 and T10.
> In general
> the SFF documents are out-of-phase with the T10 and T13 activities.
SFF specifications were never 'out of phase' at the time they were developed
because SFF projects were developed with the tacit support of T10 and T13.
Both of these committees used SFF to develop specifications in areas where
the committee did not want to do any work.
There is more SFF in ATA/ATAPI than you may want to admit, including the 80-
conductor Ultra-ATA cable. SFF specifications remained active until they
were absorbed by T13 into ATA/ATAPI-x, at which time they were Expired by
SFF.
The same thing happened in T10, with several projects completed by SFF being
absorbed by the standard after technical completion and market acceptance.
SFF specifications and INF specifications are different, as the INF prefix
identifies a document that has been released for general distribution
because the members felt it was of broad industry interest.
A long list of companies including IBM and Intel used SFF to distribute
material that would not otherwise have been made available to industry at
large (ATA SMART used to be INF-8055i).
> The "correct" way? Well, my opinion is you ignore ALL SFF documents
> (with two exceptions) and use only the T10 SCSI command set
> documents. The only SFF-80xx document that has not found a home at
> T10 is the SFF-8070 (super floppy) and QIC157 (tape).
Don't be ridiculous. Hundreds of companies rely on SFF specifications as the
only source of information on high speed signaling, connectors, enclosure
interfaces, disk form factors, etc.
There are only two SFF specifications associated with ATA/ATAPI available
from SFF, INF-8070 and INF-8090. Every other specification to do with ATA
and ATAPI has been absorbed by several standards, and have Expired.
> Yea, I know,
> there are some cases when you just can not ignore those SFF-80xx
> documents even when most are obsolete or very much out-of-date.
And how does an engineer with no history know what parts to trust and which
pieces not to? Nobody with any grasp of reality would use an Expired
specification to develop a product. These Expired and obsolete specs are
only being used because they contain content that is not available in a
standard.
The last revision of SFF-8020 was January 1996, so there has been plenty of
time to put that content into a standard. Don't blame SFF as the source of
ATAPI's woes, look to the failure of the standards process to provide
everything that developers need to know.
> But really a
> super floppy is a "block device" and should be desribed by T10's SBC
> document.
When Matshushita approached T10 to make changes/additions to SBC for super-
floppy their requests for changes/additions were rejected. The company had
no alternative but to document the device via an external document that was
submitted to SFF for distribution as an INF specification.
> The point is that anyone can write a SFF
> document and make it look like it is or should be a widely
> implemented "standard". But only ANSI or IEEE or ISO, etc, can
> produce true STANDARDS.
Absolute crap.
There is less chance of putting out an incorrect SFF Specification than
there is a bad standard because SFF has a 'majority of one' rule. It only
takes one valid negative technical comment to prevent publication.
- In a standards committee, a vote of 20:1 is a ringing endorsement.
- In SFF, a vote of 20:1 forces the spec to be revised and corrected.
Check out the list of Expired SFF Specifications and you will find that most
of them became ANSI, EIA, or IEC standards.
Dal Allan
--
If you have any questions or wish to unsubscribe send a
message to Hale Landis, [EMAIL PROTECTED] To post to
this list server send your message to [EMAIL PROTECTED]
For questions concerning Thistle Grove Industries or TGI's
list services please send email to [EMAIL PROTECTED]
