Willard McCarty <MCCARTY@vm.epas.utoronto.ca> (02/06/90)
Humanist Discussion Group, Vol. 3, No. 1004. Monday, 5 Feb 1990. (1) Date: Fri, 2 Feb 90 11:44:00 EST (15 lines) From: N_EITELJORG@cc.brynmawr.edu Subject: Re: 3.981 the ideal workstation (2) Date: Sun, 04 Feb 90 19:26:16 CST (7 lines) From: Amanda C. Lee <ALEE@MSSTATE> Subject: re:perfect workstation: myth,dream,mystery? (3) Date: Mon, 5 FEB 90 13:42:14 GMT (46 lines) From: S200@CPC865.EAST-ANGLIA.AC.UK (4) DATE: 05 FEB 90 15:55 CET (40 lines) FROM: A400101@DM0LRZ01 SUBJECT: Perfect workstation or perfect mainframe? (1) -------------------------------------------------------------------- Date: Fri, 2 Feb 90 11:44:00 EST From: N_EITELJORG@cc.brynmawr.edu Subject: Re: 3.981 the ideal workstation I would like to take issue with a couple of the comments made recently about the ideal workstation. One, 4 MB of RAM (particularly if one is using CADD or desktop publishing software) is insufficient unless virtual memory is available. Even then it is marginal. Without virtual memory 8 MB is a minimum. Two, the NeXT has recently been reviewed in BYTE. The review showed the machine to have been more than a little over-hyped. It's slow - and that's without color. Three, I question the need for large disks at the workstation. Storage can be more economically done centrally. Nick Eiteljorg (2) --------------------------------------------------------------12---- Date: Sun, 04 Feb 90 19:26:16 CST From: Amanda C. Lee <ALEE@MSSTATE> Subject: re:perfect workstation: myth,dream,mystery? The Amiga is one of the only computers that can be mac compatable, ibm compatable, or both. In fact, you can run ibm programs and amiga programs at the same time, with both running at full speed. (3) --------------------------------------------------------------58---- Date: Mon, 5 FEB 90 13:42:14 GMT From: S200@CPC865.EAST-ANGLIA.AC.UK "Workstations for Humanists" This is to be the title of a panel discussion to be held at the joint ALLC/ACH conference at Siegen, West Germany in June 1990. Hence, as chairman, I find the current discussion of great interest. However unlike the science community, I am not sure whether humanists agree on what is a workstation. A scientist's workstation tends to be a UNIX system such as a SUN Sparkstation, a DEC 3100 or a Silicon Graphics IRIS. They are usually purchased to perform a single interactive CPU/Graphics intensive application. Electronic circuit design, geographic information systems, molecular modelling are typical applications. Properly configured systems rarely cost less than 15000 pounds sterling. However even at small UK universities such as UEA, money is raised to buy and maintain such systems. BUT as yet I do not see Humanists, especially literary Humanists, rushing to buy such systems. I am not aware of many Humanist applications available for such systems. Word processing, desk top publishing, trivial databases can all be catered for on relatively cheap IBM micros or MACs. Multi-megabyte databases are possibly best held on a mainframe system and accessed using micros. Biblical scholars probably do have a modest workstation available to them. However if you wish to capture, store and selectively display high resolution coloured images of illuminated manuscripts the scientific workstation will have its uses. Perhaps a humanist's workstation is much more modest. Perhaps it has to be cheaper. Perhaps the computer industry has already provided Humanists with most of the software it really needs. Humanists! Come out of the woodwork and describe your workstation. Describe what applications you run on your workstation beyond the most obvious, but be practical and only consider what you can afford. If we all disagree, the panel session at Siegen should be invigorating. John Roper, University of East Anglia, Norwich, UK S200@CPC865.UEA.AC.UK (4) --------------------------------------------------------------45---- DATE: 05 FEB 90 15:55 CET FROM: A400101@DM0LRZ01 SUBJECT: Perfect workstation or perfect mainframe? Andrew Gilmartin criticises dreams of ideal workstations (here I tend to agree - sounds like hot-rodding to me), and then suggests as an ideal not autarchy but division of labour between micros and "large centralised computing". It's attractive, but I can think of several large buts. a. a timeshared large computer isn't necessarily quicker than a modern micro (especially if you have to shunt lots of data over the link). b. mainframes (I know this begs a question, but let's use the term as shorthand for large central facilities) are expensive discrete items. You can't buy half one year and half the next (whereas with micros you can buy say 100 one year and another 100 the next). That has several consequences: 1. people other than humanists (notably engineers, chemists, physicists, economists and computer scientists) with more financial clout will dictate *what* mainframe gets bought, and they have in practice other priorities (for example, number crunching rather than terabytes). 2. the possibilities for staggering the replacement of one mainframe by another are limited. That means, with a life cycle of five to ten years for mainframes, that we all have to spend a good deal of our computer- using time coming to terms with new operating systems (or with new versions of existing ones, which can be almost as bad). Alternatively, the mainframe providers may choose to stick for generations with a particular type of mainframe + OS combination even though it's unsatisfactory for some purposes (probably ours), because there is too much know-how etc. invested to make the change easy. Probably in principle one could now get round these problems by having a "virtual operating system" simulating some kind of standard OS, with changing real computers and real operating systems underlying it - but it wouldn't be efficient or easy, or it would already have been done. With a micro I can largely decide for myself whether I am happy with what I've got or whether I want to make changes to take on new possibilities. I can't protect myself from the unwanted consequences of technological advance altogether, but I can keep a lot of them at bay. Given enough storage space I can reduce my need for centralised computing to the minimum I can cope with (and hence also reduce my need for junk knowledge about access, command syntax, etc). The more computing services are provided remotely, the more this decision is taken out of my hands, and the more I'm forced to make frequent changes in my working habits, whether I want to or not. Timothy Reuter