[email protected] (Paul Gilmartin) writes: > There's always a reason. Rarely is it an analogue of Gresham's Law, > to which one partisan attributed the triumph of UNIX over VMS ("Bad > software drives out good!") Betamax succumbed to the greater capacity > of VHS cartridges; a decisive advantage in the eyes of consumers at a > tipping point in time despite the higher quality of Beta in professionals' > view. For many years thereafter I saw Beta only in the kits of TV news > reporters on location. I think VHS had caught up in quality and Beta > in capacity, but both camps has too much capital investment to switch.
re: http://www.garlic.com/~lynn/2013n.html#16 remember, sna didn't have internetworking and networking ... there was central vtam that had mapping to device. for a time i reporting to the same executive as the person responsible for APPN (internal architecture document awp164, i mentioned previously my wife much earlier was co-author of peer-to-peer networking, awp39) ... which provides networking layer ... i would periodically chide the person to not waste their time trying to help sna (because they wouldn't appreciate it) and come work on *real* network. it turns out that the communication group did non-concur on the draft announcement letter for APPN ... and it took six weeks of escalation to resolve the issue ... where the APPN announcement letter was carefully rewritten to not imply any relationship existed between APPN and SNA. trivia ... person responsible for DNS had worked at the cambridge science center when he was student at MIT. also in the mid-80s, i had gotten sucked into and effort to taking some support done by one the babybells and turning it out as product ... we tried very hard to isolate the effort from internal political influence of the communication group to block it ... which they managed to do anyway (which can only be described as truth is stranger than fiction). I probably didn't help things by doing a presentation on the effort at one of the regular SNA architecture review board meetings ... which had top technical and executives in the audience. part of that presentation http://www.garlic.com/~lynn/99.html#67 basically what the babybell had done was implement an NCP/SWSCP emulation on Series/1 ... which was significantly more powerful computer that what was used for the NCP/37x5 controllers ... and then actually run everything in real networking infrastructure ... except at the boundary spoofing to host vtams. all resources were simulated as "cross-domain" ... but was really fully distributed resource management with no-single-point-of-failure. the use of real networking within the operation of the infrastructure made possible a lot of things that weren't possible in a pure sna/vtam environment ... as well as having a much more powerful processor than what were used in 37x5. part of the effort also included moving the implementation from series/1 to rios (rs/6000) after the initial release. post with part of presentation made at series/1 common user group meeting http://www.garlic.com/~lynn/99.html#70 -- virtualization experience starting Jan1968, online at home since Mar1970 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
