francis@ircam.fr (Joseph Francis) (05/21/91)
In article (Masataka Ohta) 'writes': Interestingly, what the conversation seems to revolve around is someone who doesn't understand unix 'philosphy' speaking a language that those who do seem to understand, and vice versa. The logical extension of Ohta-talk is that one has a large, very fast machine running a single mega-program; 'logically', this is the most 'rationally' fast, since there is no user-oriented work done, no housekeeping per se.. It is a single machine, for a single purpose, for running a limited number of megadeth modules. (Mondo-emacs, perhaps) However, he concedes to using unix, which therefore implies a division of labor into inter- and intra-communicating segments. However, this is limited because he doesn't use machines which have a variety of tasks running, he seems to be interested in sole-proprietorship (including perhaps just a single TTY console). An idea more complex than a workstation is disallowed implicitly, and because he has a limited 'rationally small' environment any work that is relatively complex is ignored. He cannot understand (somewhat ironically, for the origin of the messages) that for large complex tasks, (no, not programs) subdivision, tool, pipelining and resource-sharing are both efficient and necessary; 'unix philosophy'. What precipitated this 'agenda' of proprietorship, oddly, was X11, probably the largest 'system' within unix that medium-level university people will come up against: not a mondo user-interface program, but related tools, on a large scale. What I propose, is for him to discuss 1) Very large, complex transaction-processing environments (commodity exchange, office automation) 2) Why multi-user, multi-tasking? 3) What does he think unix and unix variants should be used for? Typing? Can't wait: but, why are people reacting to him? -- | Le Jojo: Fresh 'n' Clean, speaking out to the way you want to live | today; American - All American; doing, a bit so, and even more so.