>>> On 11/5/2013 at 08:59 PM, Jon Perryman <jperr...@pacbell.net> wrote: > These aren't imagined ills.
Of course they are, as I discussed. > They are "ill's" that have been healed in recent > year's (in the scheme of things). Which means to repeat them _now_ is to talk about ancient history. So, don't do it. > Maybe Unix isn't as rusty as I thought. In > the last few years, it seems to have matured some but it still doesn't make > it better than z/OS. I wasn't arguing that, although I could. z/OS has its place, as does Linux, as does midrange hardware. Each has its advantages over the other. "Right tool for the job" is still the right mantra. In some cases that will be z/OS, in others it won't. The problem a number of people besides myself see is that in this mailing list, that's frequently not recognized. > Gilmartin is the one who is stating imagined "ill's" about z/OS which > started this. TSO alloc & exec's have existed for decades that could easily > merge datasets. True, but he's right. The TSO COPY command used to be part of an _extra_cost_ package. Pretty insane at the time, and even more so as I look back at it. Whereas with every UNIX, Linux and z/VM shop in the world, it was included with the base. Not something to brag about, 30 years after the fact. > Use of UNIX rename over the use of GDG's which has also > existed for decades. Ah, now this is another thing I would love to see done right in Linux. At various times, I have greatly missed GDGs. > 10 to 15 years ago is about the same timeframe as the cloud so there is some > good that has come about partially due to it's requirements. Living 30 years > without LVM and CPU hotplug (both came in about the same timeframe) or some > sort of shared resources seems like a long time. Actually it's been about 50 years, but back then we called it "timesharing." -snip- > In the cloud and Unix, is there a standardized tool or a standardized method > for vendors to participate and influence workload? There must be > methodologies but what is their commonality. I won't speak for UNIX, since I don't work with it. In the Linux world (which is a superset of "cloud" as you seen to use it), control groups (cgroups) would be the equivalent. That is fairly new and doesn't have a lot of automation built for it yet. But, enterprises have only begun to adopt Linux in significant proportion lately, so that's not surprising. -snip- > My point is that z/OS is no worse than UNIX. It's not the beast that UNIX > people believe it to be. Granted there could be improvements but that's true > for anything. Having worked with mainframes, mostly MVS, for most of my career I agree. But then UNIX and Linux aren't the child's toys that a lot of z/OS people believe they are. Knee-jerk derision isn't conducive to sharing knowledge, that's for certain, which is why I try to combat it when possible. Mark Post ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
