On Thursday, 09/23/2010 at 04:01 EDT, Ivan Warren <[email protected]> wrote:
> <rant> > Sheesh... Would it have been hard to add a TM (On the STFLE results) > around the PFD/PFDRL ? Don't tell me the labs decided to cast off an > entire line of (<4 yo) machines (which cost >$100K) for just *ONE* > instruction (and more than that - it's only for performance.. PFD/PFDRL > have no operational impact other than performance). > > The "underlying" problem I see is that it is putting a severe dent in > people's mind as far as IBM's view of "Invest in long lasting > technology" is concerned.. z9 is what ? 4 years old ? and IBM is already > issuing SCPs that are no longer compatible.. > > I remember a time (yeah.. I'm an old geezer) where the time when IBM > would release SCPs that were no longer guaranteed to work on system that > were more than 20 year old - or rather had been withdrawn from marketing > for more than 20 years.. (I'm quite convinced that with some effort, you > could have run VM/SP5 on a S/370 138.. And possible VM/ESA S/370 option!) > > PS : We're not talking about the z900/z800.. or the z990/z890.. But the z9 ! > </rant> We do not dual path the code as that creates more complicated (read: expensive) test scenarios and lengthens the time to market. Any use of any instruction from the Extensions would have to be dual-pathed. No thanks. z/VM did not cast off the z9. z/VM 5.4 has continued to receive significant updates and will stay in service for the next several years. If Single System Image and Live Guest Relocation are of interest to your business, then it is time to plan a processor upgrade. That's why we're talking about SSI/LGR so soon - to give folks time to plan hardware upgrades: z10 or z196, more LPARs, more CPUs (maybe), extra memory, and FICON CTCs and switches, Alan Altmark z/VM and Linux on System z Consultant IBM System Lab Services and Training ibm.com/systems/services/labservices office: 607.429.3323 [email protected] IBM Endicott
