Browse Prior Art Database

Nonhistory of IBM Time-Sharing

IP.com Disclosure Number: IPCOM000129958D
Original Publication Date: 1996-Sep-30
Included in the Prior Art Database: 2005-Oct-07
Document File: 2 page(s) / 16K

Publishing Venue

Software Patent Institute

Related People

Robert Frankston: AUTHOR [+2]

Abstract

278 Lake Ave. Newton Highlands, MA 02161 e-mail: bobf@frankston.com I was disappointed in the article ";'Prestige Luster' and 'SnowBalling Effects': IBM's Development of Computer Time- Sharing"; by Judy E. O'Neill (Annals, vol. 17, no. 2, pp. 5~54). There are some correct statements to be sure, but overall it was an inaccurate portrayal of history. I was involved in both Multics and CP/CMS (VM) development from 1969 through the 1970s. I'm writing this note more to set the record straight than out of a blind assumption that Multics was perfect. It was far from perfect, as ongoing design discussions and reexaminations demonstrate. With the principals still active in the field and available, it should be possible to do a more complete history. For starters, one can examine the references in ALT.OS.MULTICS on Usenet. Or, better, see file: //ftp.stratus.com/pub/vos/multics/multics.html. -- Yes, IBM offered a sales team and GE offered an engineering team to MIT, but MIT was forced to choose the latter.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 2

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

Copyright ©; 1996 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Used with permission.

Nonhistory of IBM Time-Sharing

Robert Frankston

278 Lake Ave. Newton Highlands, MA 02161 e-mail: bobf@frankston.com

I was disappointed in the article "'Prestige Luster' and 'SnowBalling Effects': IBM's Development of Computer Time- Sharing" by Judy E. O'Neill (Annals, vol. 17, no. 2, pp. 5~54). There are some correct statements to be sure, but overall it was an inaccurate portrayal of history. I was involved in both Multics and CP/CMS (VM) development from 1969 through the 1970s. I'm writing this note more to set the record straight than out of a blind assumption that Multics was perfect. It was far from perfect, as ongoing design discussions and reexaminations demonstrate. With the principals still active in the field and available, it should be possible to do a more complete history. For starters, one can examine the references in ALT.OS.MULTICS on Usenet. Or, better, see file: //ftp.stratus.com/pub/vos/multics/multics.html.

-- Yes, IBM offered a sales team and GE offered an engineering team to MIT, but MIT was forced to choose the latter.

-- Yes, Multics was a market failure but not because the market had changed. It was because Honeywell (which bought out the GE computer division) worked hard not to sell it. Customers asking to buy the system were turned away, and marketing concentrated on legacy GECOS systems. One example was the pricing of memory. Memory was priced for GECOS systems, which used a relatively modest amount. This price was very high (about 10 times what others were charging), since it covered the base costs of the system rather than reflecting the marginal costs of memory. But Multics made very effective use of RAM as a resource. Given the declining costs of memory and its value in improving system efficiency, this is an insightful decision but one that was frustrated by Honeywell's policy. Part of making a product offering viable is a buy-in from marketing. This doesn't mean that the development effort itself was mismanaged. Fred Brooks (whom the author quotes) wrote the Mythical Man Month in response to IBM's own delays. I've viewed Multics as one of the better managed projects, due to the fact that the project was actually managed (not an assumption in those days) and that the participants themselves bought into the need to manage the process and to cooperate.

-- Yes, Multics was way behind schedule. But TSS was even worse and no...