SHARE reflections

i still have 3-4 (360 reference) green cards (GX20-1703) as well as two 360-67 "blue" cards (229-3174)

One "blue" card, I borrowed it from "M" at the science center (it still has his name stampped on it) ... 4th floor, 545 tech sq:

besides doing virtual machines ... recent reference to Melinda's history & mentioning Bob Creasy's work:

the internal network:

technology also used for bitnet and earn networks

Compare-And-Swap instruction ("CAS" is the initials of the person at the science center primarily responsible)

GML was also invented at the science center by "G", "M" (one of my 360-67 blue cards has his name stamped on it), and "L, which then morphed into sgml, html, xml, etc (all the "markup" languages):

there also are a bunch of other "cards" ... several 370 reference "yellow cards", a number of "internal use only", 1980, REX yellow cards (before it was released as a product and renamed REXX), an orange 1980s VMSHARE card. vmshare archives can be found at:

which was originally provided by tymshare starting in the 70s ... some collected posts about cp67-vm370 commercial time-sharing services

There is also an "internal use only" AIDS Reference Summary yellow card, ZZ20-2341-4, Jan, 1973). There were growing number of APL-based applications supporting field, sales, and marketing. They were available on internal timesharing service HONE ... starting with CMS-APL on CP67 (evolving into vm370 along with APL-CMS, APL-SV, VS-APL, etc). from the card

IBM Aid Programs are proprietary and are clbuttifed as either Marketing (M), Services (S), Marketing and Services (MS), Testing (TI) or Internal (I) useage as defined below. IBM Aid Programs and their documentation may not be made available to customers. Instructions regarding the usage permitted for a specific IBM Aid Program and its output are supplied with the program.

... snip

in 77, the US HONE datacenters were consolidated in Cal. (as it happens not very far from the tymshare dataceter) ... and clones of the US HONE operation were springing up all over the world.

in the late 70s, the consolidated US HONE datacenter had possibly the largest single system image cluster in the world. In part because of earthquake concerns, the consolidated US HONE datacenter in 1980 than had replicated subset first in Dallas, and then a 2nd replication in Boulder (with fall-over and load balanching working between the three datacenters).

Part of our ha-cmp product

grew out of my wife's stint in POK in charge of loosely-coupled architecture ... where she came up with peer-coupled shared data architecture semi-related recent post

but also some of it grew out of the experience with clustering HONE and then having geographically replicated fall-over & load-balanacing sites. misc posts on continuous availability, disaster surviviablity and geographic survivability

--