Phil asked: Love to know what a 9662 is, though.
This is (was) a S/390 MicroProcessor Complex (PCI Card) supported in an
RS/6000.
US Announcement Letter is 197-164 (from July 22, 1997)
Birger Heede
IBM
Phil Payne wrote:
My point was it is possible to cheat IBM (illegally of course)
without Hercules or PSI machine. So, it is not good excuse
for denying software for PSI machine owners.
a) I don't see the connection.
b) That isn't the "excuse" in any case.
In fact, IBM has not said in public what its response to a request to license
its software on
a PSI platform would be, just as it never said the same about Hercules. There
are very strong
parallels between the two. It has only been pointed out that no PSI machines
are listed in
IBM's tables at
http://www-03.ibm.com/servers/eserver/zseries/library/swpriceinfo/
Well, there are a lot of other systems not listed there that have received
licenses, so I
wouldn't regard that as definitive just yet. Even IBM's own emulation (xSeries
430) isn't
listed. Love to know what a 9662 is, though.
I do feel that both PSI and T3 are being less than forthcoming about the state
of their
licensing agreements with IBM. It's obviously up to them how much they tell
the market, but I
hope they're telling their VCs the truth. I'd like to hear something about
z/VM support,
too - I can't see any box being a success today without it.
I/O seems to be a major issue, just as it was with Fundamental's FLEX-ES. I
held a prototype
PCI ESCON card in my hand around five years before it shipped. Few FLEX-ES
boxes use external
I/O because Flex's DASD emulation and FAKETAPE are so good, but if you're
aiming further up
the MIPS chart you _need_ external I/O support - and these days that means
FICON.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
