"Alfred M. Szmidt" <[EMAIL PROTECTED]> writes: > > > So you admit that there has been no _fixed_ definition. > > > Maybe you are too old to remember those things, but in those > > > times, "operating system" and "kernel" was used pretty much > > > synonymously in computer science circles. > > > > This is true, but in computer marketing circles the definition > > extended far beyond the kernel. See, for example, what AT&T > > included in System III or for that matter what all came on the > > BSD tapes. > > > > It had nothing to do with "marketting circles", check what came > > on the LMI CADR, or LAMBDA tapes. Or what came on RSX-10, or > > even ITS. > > "came on" most certainly had to do with marketing circles. > > You realise that no marketdroids touched ITS, right?
Well, if we take a look at an early ITS manual at <URL:ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-161A.pdf>, we notice that _not_ _once_ the system is referred to as "operating system". The terms used are "control program", timesharing system" and just "system". The described entity is the combination of kernel and some basic programs. A later report <URL:ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-238.pdf> starts out calling the system an operating system in the abstract. However, the introductory text then states "The ITS system resides in the DEC (Digital Equipment Corporation) PDP-10, where normal user programs are executed. It controls almost all input-output and it allocated hardware resources among users..." The whole report does not mention any user program as part of the "operating system". Mentioned are "core job", "system job", "dragon" and "pseudodevices", stuff that nowadays would be called "kernel thread". > In academic circles, "operating system" and "operating system > theory" was very much centered about the concept "arbitration of > system resources like CPU, memory, file storage, devices". > > You agreed that there was no fixed definition, but now you claim > that there was. What about "very much centered about" did you not understand? > Some people refered to everything that made up the system, some like > you, refer to the kernel. > > You'll actually find that while UNIX including its central > utilities is often informally referred to as "the system", the > more formal term "operating system", stemming from computer > science, is commonly reserved to mean the kernel itself. > > If it is formal, then there is an semi-agreed definition of the > term. Since you just agreed that there was not, insisting that > "operating system" was reserved for the kernel itself in computing > science is quite stupid. In your circles it might have been "kernel > == operating" system", but in others, that was not the case. Well, the ITS manuals not once refer to something as "operating system" which would include any applications. So it is not really making a good case for you. Care to quote something, _anything_ from that time that would actually support your case about "operating system" being applied to whole systems including applications? -- David Kastrup, Kriemhildstr. 15, 44793 Bochum _______________________________________________ gnu-misc-discuss mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnu-misc-discuss
