C4898@UMSLVMA.BITNET (Larry Pickett) (02/06/90)
On Mon, 5 Feb 90 11:48:32 PST Richard Hintz said: >Our equipment classification people are trying to come up with a >way to distinguish among micro, mini, and mainframe computers >which is likely to be recognized as technology marches on. > >One metric suggested was number of users supported? Comments? Any absolute metric you care to come up with will lose its value quickly. Todays micro is yesterdays mainframe equivalent. The three levels are marks on a continuum and only useful in relation to each other. And you left out Super which seems to have developed a meaning of more than mainframe. For example is a MAC II a micro or a mini. Micro you say? What if it's running A/UX and more than one person comes in with a terminal? One of our past directors defined a micro as smaller than the mainframe and was teased about that simplistic definition for years - now I think he was as right as one could be.
TOM@PENNDRLS.BITNET (Thomas D. Denier) (02/07/90)
I have seen one proposal for this type of classification. I don't know whether it was intended as an entirely serious proposal. A system was considered a microcomputer if the system (including external peripherals, if any) could be moved in one trip using a wheelbarrow. A system was considered a minicomputer if it could be moved under the same ground rules with a pickup truck but not with a wheelbarrow. A system was considered a mainframe if a vehicle larger than a pickup truck was needed.
magill@OPERATIONS.DCCS.UPENN.EDU (PENNnet Oper/Planning) (02/07/90)
> One of our past directors defined a micro as smaller than the mainframe and > was teased about that simplistic definition for years - now I think he was as > right as one could be. Never trust a computer you can cary... Mainframe - you need a truck and more than one person to move it. Mini - it has wheels and one person can push it around micro - one person can pick it up and carry it. Another definition I've seen - which might actually turn out to be the most consistant over time - 1-5 simultaneous users = micro 5-25 simultaneous users = mini 25+ simultaneous users = mainframe Or by another measure... It runs out of the box - a micro It runs out of the box with "configuration" - a mini It doesn't run without a "support staff" - mainframe. I've seen too many "mini/micro" installations who are really disguised "mainframe" shops - The hardware's cheap (another definition means), but the software and the critical nature of the application(s) being run are no different than any other "MIS/DP" environment. William H. Magill Manager, PENNnet Operations Planning Data Communications and Computing Services (DCCS) University of Pennsylvania Internet: magill@dccs.upenn.edu magill@eniac.seas.upenn.edu magill@upenn.edu
EDB85007@NOBIVM.BITNET (Espen Andersen) (02/07/90)
As to the definition for micro/mini/mainframe, I tell my students:
- if you can throw it - it's a micro
- if you can overturn it - it's a mini
- if you hide in it - it's a mainframe
This definition has served me well for a long time - but I am beginning to
feel a little insecure ("portable mainframes" etc.).
On another discussion list someone proposed a definition of a personal
computer as "a computer you have the rigth to turn off" - well, what do you
think...?
__ Espen Andersen, Data Consultant
/_ _ _ _ __ Norwegian School of Management
(___/_)_/_)_(__/ /_ BITNET : EDB85007@NOBIVM.EARN
/ BIX : espen
Compuserve : 71361,3425
Phone : +47 2 47 07 05PMILLER@BOWDOIN.BITNET (Pete Miller, Academic Computing) (02/07/90)
From: BITNET%"TOM@PENNDRLS.BITNET" "Thomas D. Denier" 6-FEB-1990 12:43:3 8.62 To: PETE MILLER - BOWDOIN COLLEGE <PMILLER@BOWDOIN.BITNET> ... A system was considered a microcomputer if the system (including external peripherals, if any) could be moved in one trip using a wheelbarrow. A system was considered a minicomputer if it could be moved under the same ground rules with a pickup truck but not with a wheelbarrow. A system was considered a mainframe if a vehicle larger than a pickup truck was needed. Another version uses this theme but suggests that if the user can be moved in one trip with a wheelbarrow, the system is a micro, ... and so on. Anyhow, more seriously, we don't have an official set of definitions here, but semi-formally we distinguish the following cases because their support requirements seem usually to be distinct: * Single-user systems -- always micros or "workstations" here; little expectation that the h/w and s/w on one will be the same as on any other; high expectation that user wants no one else to have any control over the system. Examples: individual micro in office, science department workstation. * Small, usually "locally"-dedicated, multi-user systems -- sometimes comprising a LAN of single-user systems whose users are always doing related work; higher expectation that the same h/w and s/w are used throughout the system; general acceptance that someone controls the system. Examples: cluster of micros in admissions and physical plant, even if not literally networked; cluster of PS/2s in public Academic Computer lab; separate cluster of Macs in public lab. * Large, multi-user systems -- Varied and concurrent work by unrelated people; certainty that someone controls the system.