On Tue, 4 Feb 2003, John Alvord wrote: > On Tue, 4 Feb 2003 12:32:29 +0000, Dougie G Lawson > <[EMAIL PROTECTED]> wrote: > > >> I think that the prerequisite is a user community for the OS that > >demands > >> that security holes be fixed, and a developer who is committed to fixing > >> the holes. IBM isn't perfect, but they have been taking security > >> seriously for quite some time now. It remains to be seen whether > >> Microsoft is truly committed to security, or whether it is just doing a > >> PR exercise. The Microsoft user community does not seem to be demanding > >> that holes be fixed. > > > >You can say that again. The commitment for zSeries, S/390�, z/OS, and > >OS/390� dates back to 1973. > >http://www.ibm.com/servers/eserver/zseries/zos/security/systemintegrity.html > > That is true in concept, although fuzzy in the details. In 1973, S/370 > virtual had just been delivered, there was DOS/VS (born of DOS), VS1 > (first born of MFT), VS2 R1 (or SVS - first born of MVT). MVS came a > few years later. [No one talks about FS any more...]
I've long-sinced discarded my sysgen listings, but I think VS2 Rel 2.0 surfaced in Canberra briefly in 74 which is about when we took delivery of our first 168. > In the late 1970s, IBM made a public commitment to resolve any The commitment was made with the introduction of MVS. I recall some training we received from IBM folk at the time, and the statement was made, "We will accept APARs for any security problems." As a junior sysprog about then, one of my tasks had been to examine VS1 dumps. I quickly decided an 0C1 with PSWKEY=0 was IBM's problem, and a few more inches (this is pre decimal) of paper went on the IBM stack.. > problems related to security/denial of service/etc ... in MVS. Now MVS > begat OS/390, which begat z/OS etc etc. It was the commitment to cure > security holes which made the difference. In '94 I discovered a neat trick for killing off initiators. I don't recall whether I was running COBOL or HLASM2 (the translators, not application programs), but if the translator was told to use all the region, the initiator died and dumped. The sysprogs didn't seem interested in persuing it; I thought it constituted a security-type problem. True, JES2 simply started another. -- Cheers John. Join the "Linux Support by Small Businesses" list at http://mail.computerdatasafe.com.au/mailman/listinfo/lssb
