This message is from the T13 list server.

                                I have posted this to the FTP site,. Dan,
please post it to the web site.

                                Doc. No.:       T13/E02033R0
                                Date:           12/16/2002
                                Project:        1510D
                                Ref. Doc.:      E02032R0
                                                D1510R0h
                                Reply to:       Peter McLean
To:     Membership of T13

Subject: Results of letter ballot E02032R0

For the ballot to forward Draft Standard 1510D, ATA/ATAPI Host Adapter
Standard revision 0h, to INCITS for further processing, the vote was as
follows: 

When the ballot was issued on 15 November 2002, twenty voting member
organizations existed. The results of the letter ballot are 18 yes votes, 1
no vote, 0 abstain votes, 1 eligible organization did not vote. The motion
passes. Details of the vote appear below.

Phil Gardner    Absolute Software       P       Yes     
Ron Roberts     Adaptec P       Yes     
Larry Barras    Apple   P       Yes     
Sumit Puri      Fujitsu P       Yes     
Andrew Vuong    Hitachi P       Yes     
Dan Colegrove   IBM     P       Yes with comments       
Michael Eschmann        Intel   P       No with comments        
Andre Hedrick   LAD Storage Consulting  P       Yes     
Tim Thompson    LSI Logic       P       Yes     
Pete McLean     Maxtor  P       Yes     
Nathan Obr      Microsoft       P       Yes     
        Network Appliance       P       Did not vote    
Tony Goodfellow Pacific Digital P       Yes     
Ron Salmon      Phoenix Technologies    P       Yes     
Arie Krantz     Qlogic  P       Yes     
Marc Noblitt    Seagate P       Yes     
Bryan Cowger    Sierra Logic    P       Yes     
Glenn Lott      ST Microelectronics     P       Yes     
Hiroshi Suzuki  Toshiba P       Yes     
Darrin Bulik    Western Digital P       Yes     

One yes vote and the no vote included comments. These comments are enclosed.

Regards,

P.T. McLean
T13 Chairman
Enclosure 1 - Comments 
Enclosure 1

****************************************************************************
********************************

Comments accompanying the IBM yes vote.

        1.      In clause 7.3.5.1 there are references to a PRD chain but
the following diagram shows APRD chains. References to PRD's should be
changed to APRD's.
        2.      A definition of CPB and APRD in the definitions section
would be helpful. Something like CPB Command Parameter Block. A Command
Parameter Block is an auto DMA mode data structure that describes a command
to be executed by the ADMA engine. (see clause 7.6.1)  APRD ADMA Physical
Region Descriptor. An APRD is an auto DMA mode data structure that describes
areas of host memory that are used during data transfer. (see Clause 7.6.1)
        3.      A definition of PRD in the definitions section would also be
helpful. 

****************************************************************************
***********************************

Intel Corporation votes NO on this letter ballot for the following reasons:

        1.      This specification has not garnered sufficient development
support to validate the technical methods.  
                        a.      Only a single company, the authoring
company, has implemented an ADMA-capable ATA adapter.  
                        b.      With continued requests to Pacific Digital,
no prototypes have been made available for independent validation.
                        c.      With this fact (of no cross-company spec
validation), this specification is only intended to support one company,
which does not make for a good standard. 
                        
        2.      Very little support in the way of involvement or feedback
has been garnered by more than a couple of T13-member organizations.  Only
one host controller vendor has been truly involved in this specification's
development.
        
        3.      This specification has no viable proliferation path to
Serial ATA, so any follow-on for Serial ATA would be a kludge in order to
support both Parallel and Serial ATA environments.
        
        4.      This specification goes beyond defining an interface (which
is T13's charter) and goes deep into defining the implementation.  So Intel
believes this standard stifles innovation of features that could be useful
to a host bus adapter.
        
        5.      Additional comments to the specification include the
following:
                        a.      Table 2, Offset 24h should be titled "Base
Address 5 - Vendor Specific"
                        b.      Table 11, offset 20-27h does not need to be
a 64-bit BAR.  Would be preferable to allow 32-bit BAR's for BAR 4 and 5.
(Additionally update section 7.4.1.10.5 accordingly).
                        c.      Annex A (A.1) makes reference to "Intel x86
PC Legacy ATA Adapters".  I'd suggest a more neutral term that does not make
use of any Intel trademarked names, including Intel's name itself.


Reply via email to