From: MCCARTY@EARN.UTOREPAS 9-JUL-1988 03:56 To: LOU Subj: Received: from UKACRL by UK.AC.RL.IB (Mailer X1.25) with BSMTP id 0280; Sat, 09 Jul 88 03:54:30 BS Received: from UTORONTO.BITNET by UKACRL.BITNET (Mailer X1.25) with BSMTP id 0274; Sat, 09 Jul 88 03:54:18 B Received: from UTOREPAS (MCCARTY) by UTORONTO.BITNET (Mailer X1.25) with BSMTP id 4048; Sun, 03 Jul 88 13:55:35 E Date: 3 July 1988, 13:43:55 EDT From: MCCARTY@EARN.UTOREPAS To: LOU@UK.AC.OXFORD.VAX ========================================================================= Date: Fri, 01 Apr 88 17:21:33 EST Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Wordprocessors & minds (159) (1) Date: Fri, 01 Apr 88 11:47:34 EST (11 lines) From: Diane Balestri Subject: the mind in the program (2) Date: 1 April 1988, 10:25:48 EST (45 lines) From: Norman Zacour Subject: The folly of comparing (3) Date: Fri, 1 Apr 88 10:59 PST (80 lines) From: Sterling Bjorndahl - Claremont Grad. School Subject: word processing and what makes good software good (1) -------------------------------------------------------------------- Date: Fri, 01 Apr 88 11:47:34 EST From: Diane Balestri Subject: the mind in the program I appreciated Willard's comments about the difference between features and the (in some sense) human level of interaction between a user and the structure of a program such as a wordprocessor. They seemed connected to the thesis of a book that I have just started to read, called Understanding Computers and Cognition (Addison Wesley, I think?) by Terry Winograd and Fernando Flores. Have any other humanists read it, and would they be willing to offer an opinion about its value? (2) -------------------------------------------------------------------- Date: 1 April 1988, 10:25:48 EST From: Norman Zacour ZACOUR at UTOREPAS Subject: The folly of comparing I doubt that a comparison of the "features" of different word processors like Note Bene (=NB) and WordPerfect (=WP) is going to get us any forwarder. As time goes on the leading word processors, like the commercial applications of spreadsheets and data bases - indeed, like the leading laundry soaps - will become more and more alike, all claiming to get your clothes whiter than white. Where the differences remain marked, of course, they can remain important; but it is difficult to make a balanced assessment when one is, say, an enthusiastic specialist in one package who bases much of his opinion about the other on the pages of PC Magazine. Others may wish to redress the balance a little; my own hope for NB 3.0 is that it will number printed lines (every line, every fifth line, every tenth, whatever), giving the user the choice of numbering blank lines or not, restarting numbering on each page or not, turning the numbering off and then turning it on, and so on. It's a nice feature of WP, which would be especially useful in any word processor that aims at a scholarly market. But in fact these differences will disappear. What is important when introducing new users to large and elaborate word processors (which at least in part was what that wonderful encomium on NB was about) is something that many of us in our pride of knowledge tend to forget: that most such users have no real interest - and will never have any real interest - in computers; will not develop any enthusiasm for computer software, logical or otherwise; and will never use many of the features offered by the larger packages. The true believer has a natural tendency to convert the infidel; I would rather think that in my father's house there are many mansions. Our enthusiasm might best be constrained at least by the following: a) the capacity of the user - most of the people I know who use word processors know as much about computer operating systems as they do about the internal combustion engines in their cars (after two years, one professor still cannot copy a file from one disk to another, but he did finish his book), and therefore ease of limited learning and use is essential. Here I would emphasize "limited"; b) the interest of the user - most users have no interest in the special activities of text analysis or even data manipulation: they just want to write; e.g. WordPerfect is the package of choice of blind users who are teachers and writers, in North America; c) the ambience - if everyone around is using NB, you'd be crazy to use anything else; instructional support is critical; d) inertia - if you are an adept of WP, NB, Word, FinalWord, WordStar 4.0, or whatever, you would be silly to change unless offered a significant bonus. (3) -------------------------------------------------------------------- Date: Fri, 1 Apr 88 10:59 PST From: Sterling Bjorndahl - Claremont Grad. School Subject: word processing and what makes good software good Willard has asked us, in the context of the word processing discussion, what makes good software good. One important aspect for me is that the software should offer flexibility of interface for the users. This means that the novice should have comforting prompts and menus, and the expert should be able to get where she or he wants quickly without having to fiddle with the same stuff the novice does. A second important aspect for me is that the different parts of the software package should be well integrated. My second "important aspect" means that I have little patience for word processing packages that have separate editing and formatting/printing programs. One doesn't need absolute WYSIWYG, but some aspects of WYSIWYG can save a person a lot of time and paper. From the computer's point of view, of course, editing and printing are separate functions. But this is 1988 and we should be beyond the stage of having software that is written from the computer's point of view. An integrated environment is a much more pleasant place to work, in my opinion. My first "important aspect" brings me to my favourite word processing package, hitherto not mentioned in the discussion. Lest people think that Word Perfect and Nota Bene are alone in the MS-DOS field, do not forget Microsoft Word version 4. Besides being essentially WSYIWYG and more or less integrated in all its parts (taking care of my second "important aspect"), it offers the user three ways to do everything possible: menu, function keys or macros, and mouse. The beginner can do everything possible by going through the menu. At each menu item, a one line prompt is displayed at the bottom of the screen. Pressing ALT-H at any time brings up a help screen for that menu item. The help screen itself has a menu offering suggestions to explore the help material, including an index to the help screens in case you don't know which menu item you need to choose to get the job done. Further, if you have really forgotten how to do something, the award-winning on line tutorial is completely integrated into the help menu. You can be taken directly to a 5-15 min. tutorial on the menu item you have chosen, or move to the index of the tutorial. The power user, of course, does not need all that help. For him or her, there are multiple options. Each function key has four meanings (which can be re-assigned under version 4) in combination with shift, control, and alt keys. There are macros, which can be "recorded" simply by typing what you want to do, or which can be programmed if one should have need for conditional branches or storage of values in variables while the macro executes. Finally, there is the mouse, which offers extremely easy ways to select and format text, to split windows (I often have three or more windows open), to move through the document, or even to operate the menu if you so choose. There are of course flaws. The spell checker is not all that well integrated into the package (although it is clearly better than some: at least it does not require you to go searching through the text for '#' symbols marking incorrect words, or something like that!). The printer drivers are hard to customize - they were clearly designed with programmers in mind rather than "lay" users. Fortunately, they are at least well documented. On the positive side once again, the outline processor is fabulously integrated. The thesaurus is much better integrated than the spell checker's "lookup" feature. The style sheets are a concept that I haven't seen on any other package in quite this way, and now I don't know how I could get along without them. There is no messy fiddling around with making formatting codes visible and invisible, as in Word Perfect. MS Word fits me well, with my high ratings for integration and flexibility. From what I have seen, Nota Bene will need one or two more major version releases before I would consider switching (although I do envy NB's database features). And MS Word is no worse off than Word Perfect when it comes to using non-Roman fonts (I use the Turbofonts package). The reviews I've read of MS Word agree with my experience, that any discussion which includes WP and NB must also include MS Word version 4, since its word processing features and power are comparable. Sterling Bjorndahl BJORNDAS@CLARGRAD (bitnet) ========================================================================= Date: Sat, 02 Apr 88 17:27:00 EST Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Wordprocessors & minds (108) (1) Date: 1 April 1988 (66 lines) From: Willard McCarty Subject: The withering away of the differences? (2) Date: Saturday, 2 Apr 1988 04:05:32 EST (24 lines) From: "Patrick W. Conner" Subject: Wordprocessors & minds (159) (1) -------------------------------------------------------------------- Date: 1 April 1988 From: Willard McCarty Subject: The withering away of the differences? If I understand him correctly, Norman Zacour has argued that comparing wordprocessors is not likely to get us anywhere, since as these programs develop their differences will tend to disappear. `Everything that progresses must converge.' Hmm. Whether or not this is something devoutly to be wished, my experience with other people's software suggests very strongly that it is simply not true. The whole point of talking about `the mind in the software' is that programs, especially highly complex ones, have a discernible underlying structure, which I am wanting to call a `mentality'. Ornithologists say that crows, for example, can identify particular human beings no matter what kind of disguises they put on; I'd suppose that the crow discerns a characteristic rhythm of movement that ripples through the flowing cape and floppy hat. In any case, I'd think that the good software reviewer similarly can see through the features to the program's basic assumptions; and in my experience it is very seldom true that a program changes fundamentally from version to version. There's a highly pragmatic reason for constancy of this sort: it's very expensive to rework radically the fundamentals of a complex program, whereas if the program has been well designed new features won't be terribly difficult to add. Slow routines can be rewritten in assembler, new algorithms adopted for doing this or that, but the basic ways of the program will tend to remain constant. Then, too, a successful program will have a committed group of users who for whatever reasons *like* those basic ways and would be upset to see them change. How people complain when even a single keystroke is redefined! For the software designer the principle of constancy would seem to have an important consequence. Forgive me if this is obvious, but isn't it true that initial decisions about a program are crucial to its eventual outcome? Another question springs to mind: to what extent does the language chosen for the development of the program affect its style or mentality? To my mind a software review that does not attempt to verbalize the mentality of a program is not worth reading. Should we not demand from software reviews standards comparable to those we expect of book reviews? I want to know what the author is getting at and how -- not just the topics. I realize that arguments over which wordprocessor is best are apt to add little to our knowledge about anything except each other's passions. As Zacour said, the believer is driven to convert the infidel, and few believers see any reason why they should understand the infidel's scripture. (It's the devil's work anyhow and therefore dangerous to mess with.) I'm suggesting here, however, that so much heat and so little light are produced by such arguments because the combatants don't know their weapons nor how to handle them. When they do, I suspect that they'll become more interested in the differing movements appropriate to their different implements than in fighting each other; those with truly inferior implements will eventually get discouraged and give up. So, the question remains, what makes good software good? What is a program's `mentality' anyhow? Willard McCarty mccarty@utorepas (2) -------------------------------------------------------------------- Date: Saturday, 2 Apr 1988 04:05:32 EST From: "Patrick W. Conner" Subject: Wordprocessors & minds (159) Bernard Bjorndahl's comments of the DOS version of MS-Word echo my own sentiments for that package on Macintosh. I've been using MS-Word 3.01 for the Mac for about a year now, and not in a million years would I return to earlier packages (including Macintosh WP's and WordPerfect, which I learned in the antique days of 1983 on an Osborne. The point that I would like to make is that Microsoft apparently came up with a product which could be implemented on both DOS and Mac machines, making it possible for us to share files regardless of our hardware configurations. I hope that idea is considered sufficiently important in the world of systems designers that we can get more software which is optimally executed on a choice of operating systems. Unfortunately, I don't know anyone who uses DOS MS-Word, so I cannot test how well connected the two versions of Word might be. Has anyone had any experience with moving back and forth on Mac & DOS machines with Word? Also, does anyone know whether Dragonfly has any intentions of producing NotaBene for the Mac. If not, was it not politically reprehensible for the Modern Language Association to endorse a product which could not serve a good-sized chunk of us? But then, MLA manages to be politically reprehensible a lot of the time, if I remember correctly its seesaw policies back during the good ol' Viet Nam days. ========================================================================= Date: Sun, 03 Apr 88 18:35:23 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Software mentality (44) (1) -------------------------------------------------------------------- Date: Sat, 2 Apr 88 18:15 EST From: (Wade Schuette) Subject: good software mentality As one who has done serious programming in something like 20 computer languages and a variety of operating systems and hardware, I have to agree with Willard McCarty that "good" software, whatever else it has, does indeed have a distinctive flavor and philosophy behind it. It really pays, for example, before programming in "C", to read Kernigan and Richies books and see what was in their minds when they wrote it. It is just a lot more fun to use a language with a spirit, than some of the committee efforts that are sold as software today. I'm not sure, however, that "best" is a meaningful term, as it implies a single-valued measure that satisfies us all. We could similarly argue over whether an IBM or a Mac is "best", or which of 5 good friends is one's "best" friend. At the current time different machines clearly appeal to different groups, and most large corporations are finally realizing that they are going to have to live with a mixed-vendor environment, as no one package delivers all things to all people, nor is it likely to. If people can at least agree that "best" is indeterminate, then maybe we can move on to "best for the particular purpose of ...., all other things being equal." As far as languages go, almost anything can be done in almost any one of them, with sufficient fluency and effort. The question is, what can one do easily. The good ones allow you to build up both speed and a library of higher level constructs for personalizing them, so that you can, after some time, really forget about the package and focus on the problem you are attempting to solve. I guess I'll suggest one *component* to look at in evaluating a package is how nicely the package becomes transparent once you have used it a lot. Packages with a unified philosophy and spirit are much easier to internalize and fly well than tacked-together spaghetti. But, like good friends - why does there have to be a "best"? ========================================================================= Date: Tue, 05 Apr 88 22:15:15 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Good software; wordprocessors (198) (1) Date: Tue, 05 Apr 88 10:59:11 EDT (94 lines) From: Steve DeRose Subject: Wordprocessing virtues (2) Date: Mon, 4 Apr 88 15:54:05 PDT (33 lines) From: sano@VLSI.JPL.NASA.GOV (Haj Sano) Subject: quality software (3) Date: Mon, 4 Apr 88 18:10:26 PDT (48 lines) From: tektronix!reed!johnh@uunet.UU.NET (John B. Haviland) Subject: WORD on MSDOS and Mac (44 lines) (1) -------------------------------------------------------------------- Date: Tue, 05 Apr 88 10:59:11 EDT From: Steve DeRose Subject: Wordprocessing virtues It seems to me that the discussion of word-processor (WP) virtues is mixing a number of questions. Perhaps breaking them apart will help. The first key distinction is *editing* vs. *formatting*. Although I think they should be integrated, the distinction is not artificial, nor is it (as someone seemed to imply) a new and computational distinction. Indeed, for the serious author too poor to buy a phototypesetter, they remain distinct. Editing is what authors (and their consciences, the copy editors) do; formatting is what graphic designers do. With rare and sometimes wonderful exceptions, authors are poor graphic designers, and vice versa. For letters and minor documents, just about any WP will get by, though (obviously) the "nicer" (vagueness intentional) it is, the better. For a book or other major publication, there are 2 options: (a) typeset it yourself, learning the art of graphic design (among other things) and using lots of time and expense, or (b) have someone else (presumably the publisher) take care of it. We almost always do (b) for serious books. In which case the sophistication of one's WYSIWYG display is of little importance: why does one need widowing features if none of one's page breaks will be the same in the end? Likewise for most of the high-end features which distinguish particular WPs. It seems to me we (i.e., authors) are being taken to the cleaners. We used to have publishers to do things for us; now we have to do the work. The display should be pretty enough not to impede authoring, but more is a bonus, nice but with little relevance to the actual task at hand. To which I say: Tag a paragraph as a paragraph, and so on for the other textual elements that *you as an author* find important, and leave the rest of the work to your publisher, so you can get on with scholarship. (This is what "generic" or "descriptive" markup in general, and SGML in particular, is for). If you have a system that lets you do that, great; if not, look for a new system. As for the scholarly task, namely the editing and content-production part as opposed to arranging ink in pleasing patterns, consider: 1) How easy is the interface to learn (i.e. for beginners)? On this, clearly anything with menus beats anything without, due to the established cognitive differences between recognition and recall. Let me amend that slightly to "menu or menu-like" for safety, but I mean to exclude systems which depend on memorizing 50 function keys and variants. 2) How convenient is the interface for experienced users? Here so-called "hot-keys" seem to me the clear winner. Once you're good with it, an editor like Unix's "vi" or any of the million-key PC editors will get a lot done fast. Alternatively, a large-scale system with complex syntactic commands can do quite well (e.g. CMS XEDIT). 3) How easily can I hand off my text to a publisher and be done with it? For this, any dependency of particular formatting and layouts is a drawback. Files should be as simplistic as possible; this dictates avoiding anything which is tied to the features of your WP or printer, and leads directly back to descriptive markup. Ideally, you should not have to know what the house style *is* in order to work with a publisher (please note that I have been talking about unfair requirements being placed on authors; if an author *wants* also to be a typesetter, that's fine; but few want to). Allow me to agree emphatically that software reflects a "mind" -- See WeinBerg's **excellent** "Psychology of Computer Programming" on this, and then parts of Brook's "Mythical Man-Month". At least, some software has this property; but some is designed by committee, with the consequent cybernetic schizophrenia; much more is designed well, but dies of accretion because the original vision wasn't great enough to encompass new thoughts. In that case, the new thoughts eventually force their way in, but they result in death or decrepitude rather than growth. One might consider analogies from the history of religion, philosophy, and science. Steve DeRose Brown Univ. and the Summer Institute of Linguistics (2) -------------------------------------------------------------------- Date: Mon, 4 Apr 88 15:54:05 PDT From: sano@VLSI.JPL.NASA.GOV (Haj Sano) Subject: quality software Whenever I encounter a clumsy tool, I always wonder if it was designed to be used by a reasonable human. This applies to a hammer, heat gun, car, hockey stick, calculator, and even software. When something is well designed and well built, it might not be noticed right away, but if it is poorly designed or built, its usually apparent immediately. Good products are seldom the result of accidents. Someone (or some committee) had to think the problem through, anticipate user needs, and perform some clever designing. Then, it was tested to search for unanticipated problems and possible non-standard uses. After several iterations, a product is released. If things were planned far enough in advance, future evolutionary enhancements are possible until the fundamental design has exceeded its useful life cycle. Two examples of long lived and evolutionary design are most German cars and motorcycles, and VAX/VMS. In both of these cases, the basic design was sound, and future enhancements were taken into account. In this high tech age, not too many things are around for very long. Software is no different from any other tool. When you use a program with a good interface, it feels good right from the start, and continues to feel good as you develope expertise. The problem with much of the software available is that human factors was not integrated into the design, and the hacker mentality of "code as you go along, who cares how it looks or feels as long as it gets the job done". My philosophy is don't buy it unless its of good quality. Haj Sano sano@vlsi.jpl.nasa.gov (ARPAnet) (3) -------------------------------------------------------------------- Date: Mon, 4 Apr 88 18:10:26 PDT From: tektronix!reed!johnh@uunet.UU.NET (John B. Haviland) Subject: WORD on MSDOS and Mac (44 lines) Word on Mac and Ms/DOS I have been studiously staying out of the "which is best" fray for wordprocessors, although I agree that there is a certain delight in the meeting of the minds that goes with learning somebody else's program--whether a mere editor or a whole programming language. As one who has spent more weeks (if not months) than I care to remember writing my *own* editor (in Z80 assembler back in CP/M days), I also know only too well that insisting on meeting my *own* mind in the software isn't all it might be cracked up to be: one can waste a lot of time trying to tune a program to one's fussy desires. But I wanted to address Patrick Conner's specific questions about transportability. For serious editing, I now use a variety of tools, mostly of an EMACS flavor--generically similar editors abound on all the machines I routinely use: MSDOS, Mac, and Vax. For wordprocessing, I use MS Word--Version 4.0 under MSDOS and version 3.01 or whatever on the Mac--and here I only wish to add a footnote to Bjorndahl and Conner. I use Word for all the reasons they mention. I particularly like style-sheets, which let me totally alter the formatting parameters of a printed document with three keystrokes. I also make promiscuous use of MSDOS Vers. 4's macros, especially useful for converting documents from one formatting system to another, as well as for other more or less complex editing tasks. I will stick with these programs without regrets until someone tells me of another word processor that provides compatibility between MSDOS and Macintosh. Here my experience may be useful to others: although conversion between the Mac and MSDOS Word formats is not perfect, it is by far the best thing available of its kind. I routinely write formatted documents on my MSDOS machine at home, using font and format information designed for the Apple Laser Printer, incorporated into a variety of style sheets. I transfer the resulting binary files to the Mac in my office, fire up Word there, and it automatically converts the file perfectly, incorporating the details of the style-sheet I wish to attach. From there, I can use the superlative laser-printing capacities of the MacIntosh without having to put up with what is, for me, its relatively sluggish performance in other areas. (In reverse the process is not quite so good: a document prepared under Mac Word converts just fine to MSDOS format, but the style sheet information gets "hardwired " in the process; I can live with that until Microsoft releases a new version of Mac Word, which will doubtless have proper style sheets.) For this situation, moving daily between these two operating environments, Word has no equal. Now if I could just fine tune it to the REST of my desires... ========================================================================= Date: Wed, 06 Apr 88 20:50:35 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Notices (33) (1) Date: Wednesday, 6 April 1988 0033-EST (8 lines) From: KRAFT@PENNDRLN Subject: OFFLINE 18 now on the file-server (2) Date: Wed, 6 Apr 88 15:42:47 PDT (8 lines) From: tim@violet.Berkeley.EDU Subject: Announcement of seminars now on file-server (1) -------------------------------------------------------------------- Date: Wednesday, 6 April 1988 0033-EST From: KRAFT@PENNDRLN Subject: OFFLINE 18 now on the file-server Issue 18 of the regular column "OFFLINE", with material of interest to computing humanists, is now available on the file-server s.v. OFFLINE 18. This issue has been written by Robert A. Kraft and John J. Hughes. (2) -------------------------------------------------------------------- Date: Wed, 6 Apr 88 15:42:47 PDT From: tim@violet.Berkeley.EDU Subject: Announcement "Beyond Word Processing: A Series of Seminars on Humanities Computing" by Dr. Tim Maher, Humanities Computing Specialist, Univ. of California, Berkeley, April-May 1988, in Berkeley, Calif. A description, s.v. HUMCOMP SEMINARS, has been posted to the file-server. ========================================================================= Date: Wed, 06 Apr 88 20:53:02 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Request for information (56) (1) -------------------------------------------------------------------- Date: Wed, 6 Apr 1988 14:42 CST From: Robin C. Cover Subject: Academic Computing: Computer Labs From: Robin C. Cover At least 15% of HUMANISTS have direct responsibility for supervising institutional "academic computing," and nearly as many direct the affairs of student/faculty computer labs. May I ask for your help? I am responsible for helping design and set up a new student microcomputer lab at our graduate school. The lab will have about 20 workstations (IBM and Macintosh microcomputers), several printers and one full-time staff person. The workstations will be capable of serving as terminals on a campus network (library system; online databases services), but users will not have direct access to mainframe or minicomputer CPU, at least initially. It would be helpful if I could obtain copies of documents that describe the services of computer labs at other institutions. I suppose most labs have summary sheets for users listing hardware configurations, software support, schedules for tutorials, hours of operation, printing fees, etc. Information of this sort would be useful to me even if the computer lab contains primarily terminals connected to the campus mainframe or network. If I can count on the good will of fellow HUMANISTS to supply me with a copy of this minimal documentation (which probably exists in every lab) .. I am equally interested in longer documents which would be useful in . I am equally interested in longer documents which would be useful in thinking -- more broadly -- about computer labs as a part of campus computing services. I realize that support for academic computing varies greatly with the size of the institution, curricular offerings, administrative & financial support, etc., and that computer labs are not as essential in highly networked environments. If anyone has internal memoranda, spec-sheets or working papers that were used in determining the PURPOSES/GOALS/FUNCTIONS of the campus computer labs, these documents would be of great assistance to me. Communications by postal or email are equally welcome. Thanks to each of you who might be willing to cooperate in this request; if you cannot answer personally, perhaps you could at least forward the request to support personnel in the computer labs. Professor Robin C. Cover ZRCC1001@SMUVM1 3909 Swiss Avenue Dallas, TX 75204 (214) 296-1783 ========================================================================= Date: Wed, 06 Apr 88 20:55:30 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Software and mind (78) (1) -------------------------------------------------------------------- Date: Wed, 6 Apr 88 09:22 EST From: 1 Does software reflect the mentality of its developer? The word 'mentality' is both too weak and too strong? It is too strong in presupposing that the individual personality of the software could be reflected by a piece of software just as it could by a piece of art or literature. It is too weak in not discriminating among the various domains that guide software development. 2 The development of educational software is guided by such domains as software engineering--'structured design'--curriculum theory and design--'individualized instruction'--and cognitive psychology--'mental models', 'expert strategies'. Similarly, in the design of computer languages, theories of what a computer program is and does, guides their development. 'A program=data+algorithms' guided the development of PASCAL. 3 The more appropriate question seems to be: what theories and domains are most relevant for specific applications? In word processing, what theories have guided the development of various software systems? I have the suspicion that word processing packages, for the most part, grew out of attempts to improve clumsy main-frame line editor packages. Line editors, originally, served the function of editing program instructions rather than composing literary products. People used these program editors, beyond their original purpose and design, to write notes, and then tried to write essays. Soon designers added various tools to help them achieve these secondary uses of line editors as main goals--such as various supplementary commands for formatting hard copy. Somewhere along the line, developers decided to make 'user interface' a bit more 'friendly' by adding a 'screen' edit mode on top of the original line editor. Here is where word processors speciated from the original line-editor ancestor: the vision of editing text, not 'line-by-line', but 'screen-by-screen'. Of course, some of the precursor elements remain, such as odd commands for driving hard copy. Unlike most other software, word processing packages developed in an evolutionary manner as a by product. Theories of software design, of the nature of the writing process, only came late on the scene to guide current modifications of pre-existing structures. Indeed, the understanding of the writing process will only now be approached that we have word processors as an alternative to the technology of pen+paper. Furthermore, it is the use of full-screen functions, and of 'windowing', that will most help us to understand the underlying mental processes used in writing and thinking, and that will guide the development of word/text processing. Since nowadays, most packages provide for scholary desiderata such as creating indexes and foot notes, the leading edge will be abiltiy to produce publication quality text ('desk-top publishing'), WSIWYG, mutliple windowing, and 'importability/exportability' between text processing and graphic processing, and more generally: the distance between the intermediate process of composing/editing with a computer and the desired final product. For instance, LOTUS MANUSCRIPT has a structured edit mode where text can be created and edited by (numeric) section/sub-section. This is very close to the type of final product I mostly prefer: a text with an automatically produced table of contents for the sections and sub-sections. Furthermore, this mode of editing allows one to produce the bare outline in terms of numeric sections, and/or to zoom-in and zoom-out to various levels of detail within and between sections. The visual aid of zooming to the most abstract level of section heads, to my mind, and to cut-and-paste section/subsections is a great aid for the composition/thinking process. 4 So, what desiderata should we present to word processor developers? In general: how close is the intermediate tasks of editing text on the screen to-- 1)the various final products we want to produce (structured texts, integrated text and graphics, fancy fonts..) and; 2)the processes we use in thinking with writing (zooming to the most abstract structure and flipping among parallel texts in different 'windows'). ========================================================================= Date: Fri, 08 Apr 88 00:05:38 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Notices (73) (1) Date: Thu, 17 Mar 88 12:11:28 EST (20 lines) From: Peter.Capell@CAT.CMU.EDU Subject: Study Group on the Structure of Electronic Text (2) Date: Sun, 06 Mar 88 22:25:26 DNT (36 lines) From: Jakob Nielsen Tech Univ of Denmark Subject: RE: HyperCard stack with report on HyperTEXT workshop availa (1) -------------------------------------------------------------------- Date: Thu, 17 Mar 88 12:11:28 EST From: Peter.Capell@CAT.CMU.EDU Subject: Study Group on the Structure of Electronic Text Carnegie Mellon University's Study Group on the Structure of Electronic Text (SGSET) will sponsor a 2-day conference, May 23-24, 1988, with the theme: "The Coming of Age of Electronic Text" The conference will include discussions of the "practical aspects of coping with the difficulties of making large amounts of text available for general distribution," says William Arms, University Vice-president for Academic Services at Carnegie Mellon. "The program will consist of five parts, each of which addresses what we we feel are among the most pressing issues in moving electronic publishing forward: real-world experience in electronic electronic publishing, the capture of information, electronic text processing, implications of structuring text for retrieval, and the economics of information. SGSET's aim is to bring together researchers, librarians, publishers, and information vendors and brokers in order to facilitate the distribution and use of electronic text." [The complete posting is now on the file-server, s.v. SGSET SEMINAR.] (2) -------------------------------------------------------------------- Date: Sun, 06 Mar 88 22:25:26 DNT From: Jakob Nielsen Tech Univ of Denmark Subject: RE: HyperCard stack with report on HyperTEXT workshop available [Extracted from the IRList 4.18 with thanks.] The stack does not contain the actual papers presented at the workshop because of copyright problems. It only contains stuff written by myself. It also includes 3 earlier reports which I refer/link to from the primary report - the reason for this is to give some feel for the hypertext situation even in a situation where I can only publish my own stuff. . . . [Note: now that I have permission, and have received a copy in the mail, I am happy to recommend this - there may be some later distribution of this and related materials through ACM as part of their new Database Products series (see Feb. CACM article by P. Wegner). Meanwhile, see details below. - Ed.] Date: Mon, 29 Feb 88 15:22:49 DNT From: Jakob Nielsen Tech Univ of Denmark Subject: HyperCard stack with report on HyperTEXT workshop available My report on the recent HyperTEXT workshop is now available in a hypertext version in the form of a 400 K HyperCard stack. To read it, you will need a Macintosh and APple's HyperCard program. To get a copy of this electronic document please send two double sided Macintosh diskettes to the following address. One diskette will be returned to you with the hypertext report and the other will be kept to cover postage and handling. Jakob Nielsen Technical University of Denmark Dept. of Computer Science Building 344 DK-2800 Lyngby Copenhagen Denmark ========================================================================= Date: Fri, 08 Apr 88 00:07:54 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: What has happened to HUMANIST? Dear Colleagues: A fellow HUMANIST just this evening sent me the following comment: ------------------------------------------------------------------- I think we might have made HUMANIST too formal and stiff. Unless my mailer has been screwing up, I have noticed an extraordinary reduction in messages of all kinds. At the risk of encouraging anarchy, you might consider relaxing the content organisation and suggesting that trivial messages won't be nuked, etc. The bursts of heavy traffic -- certainly a problem -- are more than made up for with a constant and usually interesting chatter. The experiment to formalize HUMANIST has, in my opinion, not worked out all that well. ------------------------------------------------------------------- I'm not sure that I wholly agree, but I certainly know what he means. There may be a simpler and less damning explanation for the radical decline in interesting argumentation on HUMANIST, however. At this time of year, who has the time to say much of anything except "later"? Nevertheless, the higher degree of organization, the sorting of messages into categories, brings along a subtext -- which some of you have found congenial, and others have not. I have been hoping that the earlier vigour would not be lost, indeed, that all sorts of discussion would continue despite the fact that the one big room has been subdivided into several smaller ones, rather is daily subdivided into whatever rooms seem to be required. Please be assured that no messages, trivial or otherwise, have been "nuked" or even censored. In the 11 months since HUMANIST began I don't recall ever having restrained or substantially altered a single message (I have corrected the occasional typo), nor do I recall ever having thought that I should. As far as I am concerned HUMANIST is still exactly what we make it day by day, according to the principle enunciated by Blake in that poem from his Notebook, He who binds to himself a joy Doth the winged life destroy; But he who kisses the joy as it flies Lives in Eternity's sun rise. (Nos. 43 & 59, Keynes) So, let there be much kissing of joys on HUMANIST, as well as among humanists! There is no reason whatever that we cannot have both the vigorous argumentation of old and the exchange of useful information neatly classified by topic. Willard McCarty mccarty@utorepas ========================================================================= Date: Fri, 08 Apr 88 23:05:44 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Notices (46) (1) Date: 08 Apr 88 14:44 -0330 (13 lines) From: Subject: Jakob Neilsen's HyperCard report (2) Date: 30 Mar 88 07:56:38 GMT (14 lines) From: mcvax!imag!siri@uunet.UU.NET (Equipe Chiaramella) Subject: 11th ACM-SIGIR Conference (1) -------------------------------------------------------------------- Date: 08 Apr 88 14:44 -0330 From: Subject: Jakob Neilsen's HyperCard report Humanist members may be interested to know that this report is available on Bitnet from the server MACSERVE@PUCC as HYPERCARD-HYPERTEXT-WORKSHOP-PART*.HQX.1 where "*" is a wild-card character representing one of the numbers 1 through 4. (In other words, it's in four parts...) I have downloaded it and looked through it with some interest. I presume that HUMANISTs with access to the Info-Mac archives at sumex-aim could get it from there, too. David Graham dgraham@mun.bitnet (2) -------------------------------------------------------------------- Date: 30 Mar 88 07:56:38 GMT From: mcvax!imag!siri@uunet.UU.NET (Equipe Chiaramella) Subject: 11th ACM-SIGIR Conference PROGRAM OF THE 88ACM - SIGIR Conference 11th INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL JUNE 13-15, 1988 GRENOBLE FRANCE [Full announcement now on the file-server, s.v. SIGIR CONFRNCE.] ========================================================================= Date: Fri, 08 Apr 88 23:20:05 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Forum (147) [One of the new aspects of HUMANIST that seems to have provoked the most criticism and nostalgia ("longing for return") is the absence of relatively unstructured discussion. Here is an attempt to provide a subdomain for this kind of thing, to be treasured or discarded as you see fit. If you have an opinion and the time in which to voice it, please do so! -- W.M.] ------------------------------------------------------------------------- (1) Date: Fri, 8-APR-1988 11:21 EST (63 lines) From: Wade Schuette Subject: A bit of humor (2) Date: Fri, 8-APR-1988 07:15 EST (28 lines) From: Wade Schuette Subject: Whither HUMANIST (3) Date: Fri, 08 Apr 88 09:12:04 EDT (27 lines) From: Willard McCarty (indirectly) Subject: Re: What has happened to HUMANIST? (1) -------------------------------------------------------------------- Date: Fri, 8-APR-1988 11:21 EST From: Wade Schuette Subject: A bit of humor Here's some thought-provoking humor in regard to computing in general. Origin: MailCom EchoMac InfoHere, Palo Alto 415-855-9548 (Opus 1:143/444) _________________________________________________________________________ Subject: Refrigerator SIGS, anyone? Why do refrigerators NOT need user groups, pray tell? First of all, refrigerators generally WORK, and work well. Their flyback transformers do not burn out. They do not require upgrades every year, whether the basic hardware is obsolete or not. Refrigerators usually defrost themselves; their cold solder joints do not give out; they are not sensitive to power line spikes, or static electricity. Your average refrigerator has modest design goals. When you open it, it does not say "Welcome to the Refrigerator," or blink its lights. There is probably a simple control system, no megabyte of RAM. You do not need your refrigerator to be making toast at the same time it is keeping your leftovers from spoiling, controlling your VCR, or turning the front porch lights on and off. If this seems somewhat obvious consider this: By 1995 a new automobile will probably contain the equivalent of a Motorola 68000 processor, and 1 Meg of RAM. It may well speak to you, and contain several bit mapped displays. Almost every component, from the engine to the brake system, will be microprocessor controlled. And if those components fail, your vehicle may lose manuverability, or worse. Can we be so sure that there will be not be AUTOMOTIVE user groups in 1995? Or that refrigerator user groups will not some day follow suit? Here's to a 1973 Buick, and my Kenmore fridge! Disclaimer: I have no business relationship with Kenmore, or any other refrigerator manufacturer. I am a knowledgeable refrigerator user, but do not earn my living from giving refrigerator seminars, or from leading the BMUG refrigerator SIG. ------------------------------------------------------- Maybe you can help me. I just downloaded a can of Black Olives (Collosal, I think) from my refrigerator. When I tried to open it, the 'fridge told me "A utensil cannot be found to open this can." What gives? BTW, it says 'NO PITS' on the side. Thanks in advance. ---------------------------------------------------------------------------- You have to run the Olives through CanHex first. ---------------------------------------------------------------------------- Your fridge "bug" report was not specific enough. How do you expect to get answers without telling me what version refrigerator you are using? It sounds like an incompatibility with other units residing in the Fridge, or with Kenmore's new Fridge Input/Output System (FIOS). Yes, I know we did not follow the Kenmore guidelines, but it's their fault anyway. Have you tried reformatting the Fridge? -------------------- (Your turn!) (2) -------------------------------------------------------------------- Date: Fri, 8-APR-1988 07:15 EST From: Wade Schuette Subject: Whither HUMANIST (15 lines) As a member of HUMANIST all of two weeks, I can hardly speak of trends, but I can compare HUMANIST to other services. Perhaps we need a poll, or simply to pool our knowledge, on what makes a truly interesting electronic bulletin-board (BB) system. This can be as hot a debate as what makes a great Word Processor, and it is a timely topic. (I also care because I'm working on exploring running my own BB for all our school's alumni and friends, and am trying to start comparing systems.) Has anyone else used other systems with nicer features that we could possibly copy? I really enjoy COMPUSERVE (not free, alas), with its wide range of SIGS and the ability to leap in, find "threads" of interest, and follow the exchange of messages related to just those topics in the order they were created. There was also a great deal of small talk, (CB radio type chatter), that DID make the system feel a lot more relaxed and was a lot of fun to watch. Of course, that involved remote log-ins. Maybe no one here has mastered send/remote and we just need a nicer front end for the VAXEN out there as well as the IBM. (I volunteer to write it if that's the problem.) Other ideas, anyone? ... Wade (3) -------------------------------------------------------------------- Date: Fri, 08 Apr 88 09:12:04 EDT From: Willard McCarty (indirectly) Subject: Re: What has happened to HUMANIST? [The following is extracted from a note sent privately to me. Occasionally I'll do this, omitting the sender's name, when the message seems worth broadcasting and does not either damage that person's reputation or slur anyone else's. I'm quite prepared to be castigated for this practice.] It's a tough business, trying to encourage people to participate and yet preventing electronic mail overdose, which, I fear, some of us suffer from, especially when we are involved in several (somewhat unrelated) networks. The temptation simply to purge my reader when I get back from a one week absence and discover 150 files in it is very great indeed. I would suggest that it is not the case that we do not appreciate the somewhat greater formality in the latest edition of Humanist, but rather that formality has intimidated some of us who would contribute more casual remarks, but find them not substantial enough to elaborate or articulate with such reasonableness as to send them on to HUMANIST. We can't just send short notes saying "I think Word Processing is a highly overrated topic (not that I do) because the human brain is adaptable and can make most well-designed tools do what we want them to." That doesn't sound scholarly enough. ========================================================================= Date: Fri, 08 Apr 88 23:40:16 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Aims of software (49) (1) -------------------------------------------------------------------- Date: Fri, 08 Apr 88 10:15:50 EDT From: Steve DeRose Subject: Reply to S. Richmond on software > Since nowadays, most packages provide for scholary desiderata such as > creating indexes and foot notes, the leading edge will be ability to > produce publication quality text ('desk-top publishing'), WSIWYG, .. and a number of similar points. . and a number of similar points. I must disagree; it seems to me this view, while prevalent, fails to acknowledge that one can do new things with literature and computers. As Jim Coombs, Allen Renear, and I pointed out, "the dominant model [tragically] construes the author as typist or, even worse, as typesetter. Instead of enabling scholars to perform tasks that were not possible before, today's systems emulate typewriters." (CACM 11/87, p. 933). Or as Ted Nelson more colorfully put it, when asked if he was pleased with how new WYSIWYG systems fulfill all his dreams about hypertext and the online docuverse: "NO! WYSIWY*G* -- GET, where? On *paper*! The Macintosh is a *paper simulator*. Millions of virtual trees cut down; it's the deforestation of the American mind." (Keynote address at Hypertext '87 conference, UNC Chapel Hill). I believe the cutting edge is not desktop publishing, but desktop access to literature. Writing brought knowledge to those not alive at the same time; alphabetic scripts brought widespread literacy; the printing press brought standardization and affordability of single books; optical storage and hypertext can bring affordability and the ability to navigate effectively to entire *libraries*. One can cram over 1000 books on a single disk; that means I could have a major research library (say 2 million or so volumes to start) on the shelves of one small room, at a media cost (not counting publishers' profit, I'm afraid) of perhaps $10,000. At that point, I'll care about page breaks and footnote placement about as much as I now care about paper selection for my printer (i.e. a little). Comments? Steve DeRose ========================================================================= Date: Sun, 10 Apr 88 18:31:12 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Report on NB 3.0 (21) (1) -------------------------------------------------------------------- Date: 10 April 1988 From: Itamar Even-Zohar I received Nota Bene 3.0 some time ago and have written a review (distributed via HUMANIST and NOTABENE LIST) of its achievements and problems. This is a corrected updated version of that document, based on a longer experimentation with the program. In this version, Hebrew works all right the way I customized the Beta version before, but we still expect the more advanced Nota Bene version. So I will NOT refer to any specific problems with Hebrew in this document. [Now available on the file-server, s.v. NOTABENE REPORT2. It is more that 600 lines long.] ========================================================================= Date: Sun, 10 Apr 88 18:34:16 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Forum (1) Date: Sat, 9 Apr 88 16:55 EST (29 lines) From: PROF NORM COOMBS Subject: New Humanist format (2) Date: 09 Apr 88 12:06:21 EST (13 lines) From: Malcolm Hayward Subject: A Cold Day in the Kitchen (1) -------------------------------------------------------------------- Date: Sat, 9 Apr 88 16:55 EST From: PROF NORM COOMBS Subject: New Humanist format Like many humanist members, I have mixed reacations to the recent change in format. In general, I appreciate not getting 20 to 40 messages a day. However, I think it is possible that I am very quick to throw away a series of connected messages on the basis of the title or firfst one in the series. It certainly takes less of my time. On the other hand, I keep feeling like I may well be missing something. I do not know how much others are disturbed by one feature of the mass mailing as it now exists. Each item in the mailing still contains the whole screen=full of bitnet garbage. All I need is the name and node etc. of the sender. All those other lines just interfere. Of course, because I "read" with a speech synthesizer, I must "endure" all of the garbage. Perhaps the rest of you can visually skip it and pick out the line of interest without becoming annoyed with the REST! I wonder if the software could not identify the meaningful line and throw the rest away before mailing us the package. In general I vote for the change. I do get another digest without all the electronic trivia. Whether it is done by machine or hand I do not know. I would think programming that would not be too difficult. Norman Coombs Rochester Institute of Technology NRCGSH@RITVAX.bitney (2) -------------------------------------------------------------------- Date: 09 Apr 88 12:06:21 EST From: Malcolm Hayward Subject: A Cold Day in the Kitchen An addendum to the last message. KNOCK KNOCK. Who's there? KENMORE. Kenmore who? KENMORE BE SAID ON THE SUBJECT OF REFRIGERATORS? Last time I went to my Kenmore for a byte (too obvious), I opened the freezer door too quickly and that stupid cheap plastic bar that holds the frozen orange juice cans and stuff broke right off. It was the last remaining one that had not broken off. Anyone who has a Kenmore will know what I mean. I thought, maybe that BAR was too COLD. I could go no farther. ========================================================================= Date: Sun, 10 Apr 88 18:36:17 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Software, mostly wordprocessing (102) (1) Date: Sat, 9 Apr 88 11:37 PST (52 lines) From: Sterling Bjorndahl - Claremont Grad. School Subject: aims of software: pro WYSIWYG, con DeRose (2) Date: 09 Apr 88 11:50:01 EST (32 lines) From: Malcolm Hayward Subject: On Authors as Typesetters and so on (1) -------------------------------------------------------------------- Date: Sat, 9 Apr 88 11:37 PST From: Sterling Bjorndahl - Claremont Grad. School Subject: aims of software: pro WYSIWYG, con DeRose I must disagree with Steve DeRose and others who seem to denigrate WYSIWYG. I agree that electronic publishing is in the process of making paper publishing obsolete, but will we not care how our electronic documents look? I do hope we won't be cursed with text-only, 80 column screens forever. If so, give me paper!! Preparing a document for publication means taking some care with how it looks. Not that I want to control every last detail - but my typescript provides one rather effective way of communicating my ideas to the publisher's graphic design department, with whom I can subsequently negotiate. Furthermore, most people in the humanities do a lot more than publish books. Most of them also teach courses. Having a WYSIWYG word processor makes the production of nice-looking handouts much more pleasant (especially since my free-hand drawing capabilities are quite poor). And even if paper hand-outs should one day be replaced by one networked work-station per student, I hope that my electronic handouts will not be limited to text-only, 80 column screens. Here too, I'll vote for WYSIWYG. In addition, many folks around here distribute papers in seminars. These papers may never be published, or may go through several versions before being published. WSYIWYG helps me prepare this kind of document so that its visual appearance reflects the high quality of its contents :-). Remember, neatness *always* counts, like it or not. One doesn't need WYSIWYG in order to be neat and tidy, but I find that it helps me. So, to come back to DeRose versus Richmond, one "cutting edge" of software where I live, is how to make the things I produce look good more easily. I am interested in the latest advances in word processing software, graphics software, printers. Since I am not a computer professional, these things are very important to me. DeRose is of course right when he says that the technological "cutting edge" for humanists will be the electronic library. I do in fact use texts on CD-ROMS - but I find that this is still at a very primitive state of development compared to what it will be ten years from now. I have no way to make notes in the margins of the CD-ROM texts I use. There are no accompanying illustrations. There is no way to break the 80-column barrier. And it will be several years before this technology is widespread enough to affect most of the scholarly community. Thus, the small advances in WYSIWYG-on-paper are not to be sneered at, at least for the medium term. This is something that is very relevant to us. Sterling Bjorndahl BJORNDAS@CLARGRAD.BITNET (2) -------------------------------------------------------------------- Date: 09 Apr 88 11:50:01 EST From: Malcolm Hayward Subject: On Authors as Typesetters and so on I think Steve DeRose is making some good points here about the kind of overkill of features available in editing/wordprocessing/desk-top publishing packages. I edit a journal and have recently been getting some submissions--nicely laid out, laser-printed--that honest-to-God look better than they will in print, just about. But those looks don't really matter because the papers have to be read for content anyway, and if they are going to be published, they will have to be reformatted for the journal. Another factor enters in here: in the last year or two the software for typesetting machines has gotten much better at interfacing with ordinary word-processing packages. Whereas two years ago I had to recode material for the typesetter (,, etc.), now the typesetting program will automatically convert my WordPerfect files, picking up paragraph indents, italics, and so on. All I need to do is specify type fonts, point sizes, and indicate where a new font is to be used (a title or works cited). I guess the point is, that as long as research continues to be distributed in this more or less traditional way, wordprocessing packages have gone about as far as they need to go in their formatting capabilities and typesetting packages have come about as close as they need to come to wordprocessing software to make it easy to go from an author's ideas keyed in to his or her machine to a copy of a journal in the hands of a reader (or the back stacks of a library). A more complex (for me) issue is whether to plan to continue publishing this way. I seem to remember there was a plan afoot in the Canadian Parliament a few years back to force Canadian scholarly journals to convert from a paper to an electronic medium. Was that right? Is that a live issue? ========================================================================= Date: Mon, 11 Apr 88 18:28:44 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: One person's junk (20) The following arrived this afternoon as a private note to me. I pass it on for your consideration. `I vote for a little "censorship" in the sense that there ought to be a "junk" mail folder with subject list so one can pick and choose. The first refrige note was cute, but by the second one I was already wishing it hadn't gotten started.' If you think that this is worth discussing, please do. Willard McCarty mccarty@utorepas ========================================================================= Date: Mon, 11 Apr 88 18:37:47 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Interfaces: the appeal of the mouse (78) (1) -------------------------------------------------------------------- Date: Mon, 11 Apr 88 15:51:03 EDT From: Jeffrey William Gillette Subject: Of Mice and Men A colleague and I recently discussed word processors, their interfaces, and the future. In the course of the discussion a rather stereotyped (if vigorous) debate ensued on the merits of a mouse-and-menu interface (what I call the "MacWindows" interface) vs. a more conventional command-oriented interface (I believe the test case was Nota Bene). I make no secret of my infidelity towards my IBM PC clone. I feel that the advent of the Macintosh was the single most important event of the 1980s, and that the salvation of the PC is in Windows, the Presentation Manager, and similar products. In days past I have condescendingly dismissed my colleague (and other friends with a similar point of view) as one who, perhaps, had little interest in the technology beyond getting a particular job done, or, perhaps, had never used a superior MacWindows program, or, perhaps, for some other reason did not realize that the ascendancy of the Mouse Age is inevitable. After many such conversations, I am almost prepared to own that there are intelligent, informed, computationally savvy users who have tried the MacWindows approach, and found it unsatisfying. The questions I should like to pose to humanists are two: 1) Why is it that some people (myself included) swallow the MacWindows interface hook, line and sinker, while others find such an approach unconducive to their work? What is the difference in temperament that provokes such opposite responses? 2) Is it possible to please both classes with a single product? What type(s) of interface(s) would such a word processor have? ----------------------------------------------------------------- I should like to suggest an observation on the subject (which will, I hope, occasion some debate and refutation). When the Mac was first introduced, it was touted as the computer "for the rest of us." This slogan notwithstanding, the first, largest, and most loyal group of Mac users are the computational sophisticates I frequently refer to as "hackers." At the risk of pressing my point beyond proper bounds, I will add that, although I know people who have purchased Macs as their first computer, I cannot think of a one who was not a "quick study" with computers. I know many technologically "average" people who use PCs, but it seems to me that it is a technologically elite group that gravitates toward the computer "for the rest of us." To complete the thought, my impression is that the same situation holds with respect to the Microsoft Windows product. Would anyone care to venture an explanation as to why it is largely engineers and hackers who are pushing forward the MacWindows standards, frequently against the objections of the very users (both "power users" and neophytes) whose cause they purport to champion? ========================================================================= Date: Mon, 11 Apr 88 18:41:41 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: w-p, typesetting, & the price of technology (48) (1) -------------------------------------------------------------------- Date: Mon, 11 Apr 88 22:41:49 IST From: Ron Zweig In the ongoing debate on w/p, there have been a number of references to the increasing ease with which new w/p software allows authors and editors to have text typeset directly from disk. A number of discussants have been using these techniques, as I have, for the past few years. I have an observation to offer on this new task, and wonder if others might have reached the same (or other) conclusions. Five years ago, the direct interfacing with typesetters from a w/p disk was cumbersome, and it usually required a lot of mutual patience and practice, not to mention the entering of typesetting codes into the w/p file. But it was worth it because it (i) offered significant savings in cost - 40% (my facts are drawn from my experience in Israel, but I think that are probably valid elsewhere too) (ii) it was, or at least promised to be, much quicker, (iii) it was definitely hi-tech. Since then, the techniques have become widespread, much simpler and far more likely to work successfully. BUT ... the price advantage has disappeared as typesetting costs have crept back up. So much so, in fact, that this novel use of w/p technology has simply redefined the division of labor between typesetter and editor/publisher/author. The latter do more of the typesetters work and have little gain to show for it. True, the quicker turnaround and the lessened aggrevation are worth something. But has the introduction of computing transformed our relations with typesetters to our benefit or theirs? As desktop publishing intrudes more and more into our work, I begin to feel that the situation will repeat itself. We will not only do all the keyboarding and introduce the major codes, but we will also end up doing the page-makeup, design etc, with little to show for it in reduced production costs. I enjoy playing with Ventura as much as anyone, but I begin to wonder whether scholarly publishing or the printing trade has most to gain by the new technical possibilities. Ron Zweig Tel Aviv University H27@TAUNIVM ========================================================================= Date: Mon, 11 Apr 88 21:25:07 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Typesetting costs (26) (1) -------------------------------------------------------------------- Date: 11 Apr 88 20:10:59 EST From: Malcolm Hayward Another issue with typesetting costs: obviously as it becomes easier for an editor to move from an electronic text to typeset galleys, so too it is easier for a typesetter to do so, in theory driving down the cost there. Pretty quickly it will be seen that the only advantage to doing typesetting will be the small time saving that might accrue to putting in codes at the point of assemblage rather than first assembling a text in one place (the editor's computer) and coding it in another. I find right now, however, that typesetting only runs about one third of my total production costs; since production costs are about equal (rule of thumb) to other editorial and distribution costs, typesetting is only one sixth of the total cost of producing a journal, book, or what have you. Thus even if I had total control of my typesetting the savings would not be of much moment. ========================================================================= Date: Tue, 12 Apr 88 19:25:15 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Interfaces (168) (1) Date: Tuesday, 12 April 1988 1019-EST (58 lines) From: JACKA@PENNDRLS (Jack Abercrombie & Todd Kraft) Subject: Comments on MS Windows (2) Date: Tue, 12 Apr 88 10:11 EST (23 lines) From: Subject: wsiwyg (19 lines) (3) Date: Tuesday, 12 Apr 1988 07:44:07 EDT (16 lines) From: "Patrick W. Conner" Subject: Interfaces: the appeal of the mouse (78) (4) Date: Mon, 11 Apr 88 20:27:03 EDT (24 lines) From: Diane Balestri Subject: of mice and students (5) Date: Mon, 11 Apr 88 22:38 EST (29 lines) From: Subject: Interface for the blind (1) -------------------------------------------------------------------- Date: Tuesday, 12 April 1988 1019-EST From: JACKA@PENNDRLS (Jack Abercrombie & Todd Kraft) Subject: Comments on MS Windows Under an IBM technology transfer project, we have been working with MS Windows (version 2.03) for some three months now. We have made it through the mounds of technicial information accompanying the package as well as the software itself. We feel that we now can share with you some of our preliminary observations. Most important, MS Windows is a forward albeit diagonal step in the right direction. We can understand why Apple is sueing Microsoft. (Question: Why didn't Xerox sue Apple?) MS Windows 2.03 seems to be a polished and relatively stable product with no significant bugs. List of Positive Comments: 1. The Graphics interface is well-designed and powerful. We have been running Windows on a PS/2 (model 60), and speed of display is comparable to early Macintosh. 2. It is good that they have given us a full diskette of sample programs to use and study since programming in the Windows environment is unlike normal DOS programming. (See commment below.) 3. Windows extensive libraries, GDI, User, and Kernel, certainly cut down on programming development especially in designing user interface. List of Negative Comments: 1. Although Windows provides for powerful functions for Latin fonts, it lacks sufficient development for non-Latin scripts such as Arabic or Hindi. From our perspective, foreign font development would include additional information (left offset and movement) on each character in a font. This information is lacking in the current version of Windows we are using. A good example of what we are advocating can be found in the font descriptions used by the HP LaserJet. Since Windows lacks these important pieces of information, again in our view, we have been forced to work around the problem by creating an add-on resource. 2. Windows is gigantic by DOS standards, and as we move to OS/2 such will not be the case. Also, the entire package runs adequately on a PS/2 50. (We haven't tried it on anything smaller.) We tend however to feel that the PS/2 80 will become the low-end machine for real functional use of Windows/Presentation Manager. 3. Programming with Windows is not the easiest task. Because of its interactive and multitasking nature, programming problems that were more difficult to solve are now simplier, but some of the simplier things have become difficult. One example of the latter is trying to execute a fscanf (PASCAL: readln(filein,line) from a file. (2) -------------------------------------------------------------------- Date: Tue, 12 Apr 88 10:11 EST From: Subject: wsiwyg (19 lines) What should we "get" with WSIWYG? "WSIWYG" encapsulates the desideratum of no distance between the intermediate process of composing/editing with a computer and the desired final product." Steve DeRose points out that when we emphasize getting paper products as the output of wordprocessing the leading edge becomes the cutting edge of deforestation. However, as others point out, the current demand on those of us who use word processing is to produce paper products. For instance, a recent conference announcement on computing and philosophy stipulates that submissions be in paper rather than through electronic mail or floppy disk. Perhaps, if we shift our desired end product from paper output to monitor output, we can change the criteria of WSIWYG to monitor-ready quality, such as: split screen, colour, graphics, multiple-layered screen, and mutiple-windowed screen as the "get" part of the formula. Realistically, we want wordprocessors to perform different functions--as e-mail text generators, desk-top publication systems.... So, WSIWYG must, to day, include a "G" which equals paper. When "G" shifts to mainly monitor products, WSIWYG will change its meaning. (3) -------------------------------------------------------------------- Date: Tuesday, 12 Apr 1988 07:44:07 EDT From: "Patrick W. Conner" Subject: Interfaces: the appeal of the mouse (78) I have to run to teach a class, so this will be brief and ill proof-read. The Mac interface appeals to technological sophistocates because they appre- ciate its potential not only in word processing, but in developing the union of man and machine. That's really what engineering is all about. The folks who only want to use a single application are served by any computer they take the trouble to learn (and to them it is trouble); they aren't interested in the machine's potential, and they don't have the sort of imaginations which conjure up better versions of their applications. I think that the notion of bicameral dominances may also be involved, but that may also be a bunch of once-fashionable hogwash. Does anyone know whether right-brained people prefer one sort of machine and left-brained another? (4) -------------------------------------------------------------------- Date: Mon, 11 Apr 88 20:27:03 EDT From: Diane Balestri Subject: of mice and students I read Jeffrey Gillette's comment about Macs and hackers with interest, and a little surprise. At Princeton, where I keep an eye on the way computers are penetrating the population and the curriculum, the Mac is the overwhelming choice of the student body, only a few of whom could by any stretch of the imagination be called hackers. We initiated a student discount purchase plan (with easily obtained loan plan to cover it) this fall; over 75% of machines purchased have been Macs. At this point most of the public machines available for students in labs or wordprocessing clusters are IBMs, by the way, thanks in most part to the major grant we've enjoyed from Big Blue--so it's not that we are encouraging the Mac with classroom application. On the contrary, student (and increasing faculty) preference for the Mac is beginning to drive our planning for new clusters and facilities. My sense is that the kids are finding the mouse/menu interface very intuitive and visually appealing; they also like the ease of formatting a paper and the quality of the laseroutput. (Needless to say, most of what they are doing is wordprocessing.) In other words, I don't see the Mac as a hackers' heaven at all, though it may be that too. To me, it's the machine that's converting the doubters and making the amateurs feel that the computer is a pretty friendly and useful tool after all. (5) -------------------------------------------------------------------- Date: Mon, 11 Apr 88 22:38 EST From: Edward Friedman Subject: Interface for the blind Regarding computer interface concerns - I recently visited the research group of the National Foundation for the Blind in NYC. Various voice and dynamic braille interface devices make most computing activity possible for the blind. However, the more recent windowing, use of icons and other intrinsically visual interface devices ( like the mouse ) have created a crisis for those concerned with use of computers by the blind. It may not be possible for substitute proceedures to be developed. I wonder if some of the Humanist members have thoughts on this issue. Are there any blind members in Humanist? Ed Friedman ========================================================================= Date: Tue, 12 Apr 88 19:27:21 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Typesetting (32) (1) -------------------------------------------------------------------- Date: Mon, 11 Apr 88 20:56:44 MST From: Mark Olsen There are a number of hidden costs that occur when a journal or author wants or is required to perform typesetting. I was on the staff of an academic journal that the SSHRC required -- for funding purposes -- to do all its typesetting in-house. Very quickly, the editor and I, the only two computer "wizards" who had sufficient knowledge and computers, functioned as copy-editors, typists, layout artists and everything else. Needless to say, the journal did not appear at a regular rate. Even worse, the learning curve for editorial assistants, graduate students, was so long that few had learned all the ins and outs before they had graduated or moved to more "rewarding" work. The required level of expertise will necessitate a half-time permanent person who can learn the job and provide continuity. While this was a few years ago, and I doubt that one would have to write a typeset simulator to check the formatting on dot matrix before setting it to press (every error cost money), I am still VERY suspicious of anyone who wants to get their texts camera ready. The benefits are minimal and the costs in terms of time and effort that could be expended elsewhere are too large. Writers and editors should write and edit! Mark ========================================================================= Date: Tue, 12 Apr 88 19:58:01 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Forum: improving HUMANIST (82) (1) Date: Tue, 12-APR-1988 06:30 EST (18 lines) From: Wade Schuette Subject: HUMANIST FORMAT;filter items with subject keywords (2) Date: 12 April 1988 (46 lines) From: Willard McCarty Subject: Good ideas for improving HUMANIST (1) -------------------------------------------------------------------- Date: Tue, 12-APR-1988 06:30 EST From: Wade Schuette Subject: HUMANIST FORMAT;filter items with subject keywords (10 lines) An important property of a bulletin board system seems to be a workable facility for either automatically filtering out items you don't want to see (due to topic or source), or for telling at a very fast glance from one line whether to read or delete something (or maybe explore it further.) I'm not sure what HUMANIST's technical capacities are. I'd rather have 15 unbundled messages, with one *good* key line for each, than 5 bundled message groups. How about the idea of forcing messages to have the subject line in the format: KEYWORD:(# lines):details, where we agree on 10-20 short, fixed KEYWORDS, such as WP (word processing), HUMOR, HELP, CONF (conference notice), LIB (new library item), etc.? On some systems you can do that, and automatically set a filter so you never even see messages that have keyword areas you don't want to see. (2) -------------------------------------------------------------------- Date: 12 April 1988 From: Willard McCarty Subject: Good ideas for improving HUMANIST As `editor' of HUMANIST I am very grateful indeed for the time, energy, and ingenuity members expend on making suggestions for the improvement of our discussion group. Often it must seem that these simply get ignored. Often, it seems to me, suggestions are good but are either technically impossible with the present software or would mean a significantly greater amount of work for me. In part HUMANIST's success depends on the donation of my time, the postmaster's time, some computer cycles and storage space, the ListServ software, indeed, the networks by means of which all of you receive its messages. Resources in Toronto are sufficient to run HUMANIST as it now is, but we haven't the programming power to improve ListServ even if we were permitted to do so, and since the author gives it away to all, we cannot complain much about what we have from him. A few HUMANISTs have contributed VM/CMS software to make my job easier and to help with the biographies, and I am most grateful; two or three are working on resorting and tagging the biographies; another has taken on the job of writing the periodic summaries of activity; another volunteered to be a software review editor, but that initiative came to nought through no fault of his (the world isn't ready for co-publishing electronically). We have, that is, the beginnings of an editorial staff. Its membership is open. All you need to join is to work for the common good. Speaking of which, I detect that the next piece of real work that needs doing for some HUMANISTs is the writing of software to take apart bundled messages. I seem to recall that bundling is a problem for some of you. Is anyone willing and able to write clever code to handle this problem on whatever systems it may exist? Some of you like bundling; the rest of you, I'm sorry to say, will just have to accept it as a fact of life if for no other reason than it radically reduces the amount of time I have to spend processing the mail. Some of you will remember what HUMANIST was occasionally like before editorial intervention was imposed, and if your memory is clear you'll not want to return to those bad old days. `Electronic Chernobyl' was an evocative phrase at the time. So, thanks for the suggestions. I'll be even more happy to see the volunteers. Willard McCarty mccarty@utorepas ========================================================================= Date: Wed, 13 Apr 88 22:08:50 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Interfaces (191) (1) -------------------------------------------------------------------- Date: 13-APR-1988 18:06:50 GMT From: Philip Taylor Subject: Of mice and men Jeffrey William Gillette's recent contribution ("Of mice and men") so aroused my innate biases that I feel obliged to put aside my lethargy and respond; unfortunately, in so doing, I realise that I am about to open (re-open ?) the very same can of worms that INFO-VAX seems unable to close. Jeffrey suggests that computer users may be classified into "pro-mouse" and "anti-mouse" groups; INFO-VAX (frequently) suggests that computer users may be classified into "pro-VMS and pro-UNIX" groups. I wonder if the two groupings are orthogonal or related ? I should start by confessing my own prejudices: I am anti-mouse, anti-UNIX, and pro-VMS. My reasons are probably not well understood, even by me, but introspection suggests that they include: A liking for (natural) language; a dislike for the current tendency to represent as many concepts as possible as ideographs, particularly where the potential audience does not share a common first language (c.f. the safety instructions now universal among airlines, and which make as little sense to me as would the same instructions expressed in hieratic or demotic scripts); a dislike for slang, and a dislike for (unnecessary) abbreviation ("cuppa" for "cup of tea"; "Toys Us for "Toys are Us"; "Pick 'n' match" for "Pick and match"); a heretical belief in proscriptive grammar, as opposed to the (supposedly old-fashioned) prescriptive and (currently acceptable) descriptive grammars; the belief that, when typing, the fingers rest naturally on the `home keys', and should move as little as possible from those keys. (there may well be others). So, given these prejudices, what is my ideal man/machine interface ? (I apologise for the use of sexist terms, and assure any reader(s) that no slur is intended). A command-line interface, in which the vast majority of the characters used are alphabetic, and in which these characters form (natural-language) words. Where natural-language words are used, their meaning should be as close possible to their meaning in natural language. Where alphabetic characters are used, their case should not matter (I justify this by suggesting that, in the English language at least, it is extremely difficult to construct a sentence in which the meaning would change by changing the case of the letters, and by asking the question that students invariably ask when encountering a case-sensitive computer for the first time: "But WHY doesn't the computer understand that `DELETE' means the same as `delete' ?") Where non-alphabetic characters are used, their meaning should not conflict with their most common non-computer-related meaning. For example delete myfile.text /confirm (or Delete MYFILE.TEXT /confirm) meets these requirements; rm myfile.text -c and do not. (The examples are somewhat artificial; I am not sufficiently familiar with UNIX or MAC/Windows to contruct real analogues of the VMS command above). My objections to the UNIX-like command are: that "rm" is not a natural-language word, and therefore has no inherent meaning; that "-" typically indicates negation, whereas in this context it indicates assertion; that "c" has no unambiguous meaning; and that "C" should mean the same as "c". and my objections to the MAC/Windows-like command (sequence) are: that it requires removal of the hand from the home keys; that it requires good visual-motor co-ordination to move a pseudo-object using an essentially `detached' right-hand; (I have far less objection to a light-pen approach, where the pseudo- object is actually pointed to by an object held in the hand). that it requires an understanding of the various ideographs used to represent the concepts being manipulated; and that it requires a time-dependent response, which is atypical for computer applications (I can spend an hour entering a single command to VMS, breaking off for coffee if I like, even between the characters of a single word, but once I start to select an object using a mouse, I must complete the second button-depression within a (seemingly very) short time. But even the suggested interface is not sufficient; the most proscriptive of proscriptive grammarians would accept that there is sometimes a case for abbreviation, as, for example, when a term is frequently used. It is therefore necessary to allow commands and qualifiers to be abbreviated, but clearly undesirable that they should be abbreviable to the point where they become ambiguous (does "d" mean "delete" or "define" ?). VMS adopts the convention that all commands and qualifiers are abbreviable to the first four characters, and are guaranteed unique when so abbreviated (as the VMS command set is dynamically extensible by the user, some constraints must be placed on the definition of new commands to ensure that they do not clash with existing commands after abbreviation). Abbreviation to four characters allows the knowledgeable user to speed up his/her entry of commands, but still leads to a fairly verbose command string; VMS therefore allows any command (or any parameter, or any qualifier, or any combination thereof: in fact, any arbitrary string) to be bound to a key. Such keys are not a part of the main keyboard, and therefore the home keys plus their neighbours retain their canonical meanings; only the function keys may be re-bound. It is important to realise that the option to abbreviate a command, or to bind a command-string to a key are just that: options; it is never necessary to abbreviate a command, or to use a function key instead of a command, and thus the naive user need never learn obscure commands or function-key sequences; he or she may continue to use full natural-language words for as long and as often as they like. But even VMS has its limitations: for example, if a user has entered the command Delete MYFILE.TEXT and wishes to know what the keyword is which will allow him or her to change their mind before the deletion actually takes place, there is currently no (simple) way of asking VMS what options may follow a given command, once the command has been part-entered. (One can always ask for help on a command before using it, but, once it has been entered, one cannot branch to the help system without cancelling the command). Stan Rabinowitz's "WHAT" program overcomes this limitation. As each command is entered, an incremental syntactic analysis takes place; at any point during command entry, it is possible to request help (by entering a question-mark): the syntax analyser `knows' what may legitimately follow the part-command entered, and (using nested windows, to which I have no objection whatsoever) displays the legitimate continuations of the part-command. Furthermore, for those who are slow typists, but who still prefer to see commands and qualifiers expressed in full, any part-command may be automatically completed by the system; on entering , the syntax analyser will automatically complete the part-command, provided that it is already unambiguous, or will indicate that the part-command is ambiguous and requires further specification before it can be completed. (I would accept that the key is not mnemonic in this context, whereas I believe that the earlier use of the question-mark for incremental help is mnemonic.) So, is the combination of the VMS and "WHAT" interfaces ideal ? I believe that it is, and that it leaves the MAC/Windows interface miles behind (and leaves the UNIX interface light-years behind, but that's not really the point at issue). OK, I've opened the can of worms; over to you ..... (Philip Taylor; RHBNC, University of London; U.K.) ========================================================================= Date: Wed, 13 Apr 88 22:14:38 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Forum: expanding eyes (67) (1) Date: Tue, 12 Apr 88 22:51:50 EDT (13 lines) From: Charlan@CONU1 Subject: The new format (2) Date: Tue, 12 Apr 88 20:59 CDT (6 lines) From: Wayne Tosh / English--SCSU Subject: re: (What's happened to HUMANIST?) (3) Date: Wed, 13 Apr 88 13:07:04 BST (8 lines) From: "prof.s.r.l.clark" Subject: Re: One person's junk (20) (4) Date: Wed, 13 Apr 88 18:50:48 BST (12 lines) From: AYI004@IBM.SOUTHAMPTON.AC.UK Subject: Censorship (1) -------------------------------------------------------------------- Date: Tue, 12 Apr 88 22:51:50 EDT From: Charlan@CONU1 Subject: The new format I must say I approve of the new compressed format for HUMANIST. I access netnorth through the UMass maile operating on a CYBER system. I am connected through a 1200 baud line. I would not have access to software packages that would do a triage, nor can I scan through quickly. The current configuration permits me to disregard technical discussions that do not interest me (eg. the computer-based analysis of text files) and focus on those subjects that I find of interest. My thanks to Willard for his efforts to make the system manageable. Maurice Charland (2) -------------------------------------------------------------------- Date: Tue, 12 Apr 88 20:59 CDT From: Wayne Tosh / English--SCSU / St Cloud, MN 56301 Subject: re: (What's happened to HUMANIST?) Keep up the good work, categorizing as you have been. HUMANIST is working fine. (3) -------------------------------------------------------------------- Date: Wed, 13 Apr 88 13:07:04 BST From: "prof.s.r.l.clark" Subject: Re: One person's junk (20) I endorse the plea for a junk-mail folder: I only joined the hotline a few days ago, and my reader is already clogged up every morning. Open discussion is fin e, but it would be nice if it were about something significant. Question: is anyone working on interactive texts? (4) -------------------------------------------------------------------- Date: Wed, 13 Apr 88 18:50:48 BST From: AYI004@IBM.SOUTHAMPTON.AC.UK Subject: Censorship The call for various forms of censorship, either editor-administered (one person's junk) or through some filtering device, destroys one of the positive aspects of HUMANIST - the unexpected. One could even say that learning requires the unexpected - not to expand one's edifice, but to change it. William Blake would probably say something like "The eye sees more than the heart knows". Brian Molyneaux (AYI004@uk.ac.soton.ibm) ========================================================================= Date: Wed, 13 Apr 88 22:17:48 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: The new OED (89) (1) -------------------------------------------------------------------- Date: 13-APR-1988 13:36:55 GMT From: Geoffrey Wall From official ( and semi-unofficial) statements about the progress of the NEW OED, I have put together a brief account of what is on the way in the next five years A new integrated OED (original text of 1928, along with Supplements and some newly compiled material) is to be published in book form in 1989. Meanwhile, the old OED, without material from the supplements, was published on CD-ROM at the beginning of 1988. OUP sees this as a dummy run for an integrated version on CD-ROM, due in the early 1990's. The current CD-ROM version will help get the design right. It's just 'the first step on a sharply rising learning-curve'. Various technical problems have limited the current CD-ROM version. Firstly, the typographical intricacies of the original printed text could not easily be transferred to the electronic version, because of the limitations of most current VDU screens. OUP decided to do away with some of the more obscure language fonts (esp the Greek and the Old English) and with the profusion of diacritical marks. 'A more purist approach would have put the publication beyond the reach of most of its potential users.' Secondly, limits are imposed by the size of the work. One CD would hold the entire OED, with just a little space left over on the disk. But obviously the disk only becomes useful as indexes are added to the raw text. Given the complex structure of the OED text (some forty fields in the database) indexes could easily bulk larger than text. OUP's solution has been to put out several CD versions. First, the 'complete' version. This fills three disks. And, to do any useful work, it needs three linked CD-ROM drives, so as to search all three disks simultaneously. (Most current CD-ROM editions require only one CD drive.) Then there will be two reduced versions: a linguistic disk (OED minus the illustrative quotations) and a literary disk (quotations only). Facilities on this version include: search whole corpus for any word or phrase; or generate complex specialised lists. (For example: lists of words supported by quotation from Shakespeare; words from any chosen register (eg slang) used by Walter Scott; words from Hindi that occur in English before 1750. Thus, a whole realm of information hitherto buried in the printed version of the OED will become accessible. Simple queries, for example: -What interjections were in common use in the period 1670-1720? -List all mineral names with the dates when they were named -What meanings of words does Milton follow Spenser in using? Or, more complex queries, for instance -investigate historical shifts in thinking about relations between body and mind by scrutinising a set of key terms (mind, body, emotion etc) The electronic OED will not, they think, be in competition with the New OED in printed form. It will be used in different ways, as a 'list-maker' rather than a 'page-turner'. ( Who would want to read a 60,000 word entry on screen?) OUP say that an on-line database is not (not yet? not ever?) appropriate for a dictionary of this kind. OED on CD-ROM will sell mainly to the major research libraries, rather than to individual users. OUP currently predict that there will be no significant income from the sale of electronic versions in the first five years. Personally, not being able to afford a PC clone, let alone a dedicated three-drive CD player, I'd like to have on-line access to the OED as soon as possible. Along the lines promised by the Institut National de la Langue Francaise, for general access to their lexicographical database at Nancy. Meanwhile, perhaps the NEW OED Centre at Waterloo would consider some kind of access? Geoffrey Wall ========================================================================= Date: Thu, 14 Apr 88 20:07:50 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Interfaces (132) (1) Date: 14 Apr 88 10:24 -0330 (76 lines) From: David Graham Subject: interfaces (2) Date: Thu, 14 Apr 88 03:58 EST (12 lines) From: PROF NORM COOMBS Subject: Mouse or keyboards (3) Date: Wednesday, 13 Apr 1988 23:07:19 EDT (22 lines) From: "Patrick W. Conner" Subject: MacSonnet (1) -------------------------------------------------------------------- Date: 14 Apr 88 10:24 -0330 From: Subject: interfaces First of all, apologies for what will probably be a long posting... As someone whose first experience with computers involved VMS, whose second was UNIX, and whose third was (and is) Macintosh, I really feel that I cannot let Philip Taylor's comments go by without replying. I frankly find the assertion that the VMS command-line interface is a "natural language" interface ludicrous. How well I remember my first attempts to create a new directory, and my fre- quent trips to the on-line help and the manuals. Not to mention trying to navigate through several levels of directories once I had managed to create them. And as for sorting or merging something... Not to mention the fact that often this "natural language" interface tends to be completely discarded once the command-line has been left behind. Just now I wanted to include a copy of Philip's message in this file as a buffer (Yes, I'm logged onto a VMS system from home at the moment). So inside my mail file I entered something like the following series of commands to view the first lines of the file to make sure I had spelled his name correctly: CTRL-Z * inc humanist.bmail;1 =hum * sh buf * ty 1 to 10 * % Unexpected characters at end of command [or something similar] * ty 1:10 [I try again, having forgotten how to do this] * ty 1:20 [first try didn't show enough of the file] I submit that for me at any rate this is neither natural nor economical. I'm sure there are far better ways to do this in VMS, but for me VMS will always be something one puts up with rather than something one loves. Contrary to Philip's experience, I find Unix much more congenial than VMS. When learning Unix (which I will admit I do not manipulate much better than VMS), I found the basic command set easy to learn an d remember *because* it was abbreviated and because it was (in the main) mnemonic [please don't quote 'cat' at me--I always use 'more' :-)]. My main objection to Unix is that in order to get it to do anything _really_ worthwhile, one has to learn *much* more than the basic command set. I'm sure that with 'vi' and 'troff' one can perform wonders, but the amount of time I have to expend to get anything remotely resembling what I want is such that I simply cannot be bothered--there's more to life than memorizing command languages, and I want to be *doing* something rather than figuring out *how* to make the /*&??$ machine do something it clearly doesn't sympathize with. I am not going to indulge in a rhapsodic description of the wonders of the "Mac/Windows" interface, but I must say that Philip's description of mouse use is misleading in that after the first few minutes, there is no conscious thought involved in clicking on a file and dragging it to the trash. [Mac mice have only one button, not "three apparently identical" ones] And contrary to the talk about ideographs, the great bulk of Macintosh use (in my experience) does *not* involve icons. The interface is graphic-based, true, but the icons are largely restricted to the "desktop" or top-level directory; once an application is launched, very little actual icon use usually occurs. The advantage of the mouse/window/pull-down (or pop-up) menu interface, in my view, is that it permits the beginner to use the menus, where all the commands are laid out in full view, and progress to using the keyboard equivalents, of which a multipli- city are often provided, as experience is gained. This means a sharply reduced reliance on the manual, which cuts the loss of time which entailed by going to the manual, finding the appropriate passage, and trying to understand what the manual-writer has written. I think Philip's comments, however, point up something said earlier by Jeffrey Gillette (I think), which is that there is a class of sophisticated, knowledge- able, intelligent computer users who simply do not find the mouse-based inter- face attractive or useful. Often these people, when I encounter them, seem to be computer professionals: system managers, Unix gurus, people with a storehouse of proprietary knowledge which they have spent time acquiring and which is valuable to them and to others. Often, I think, the "Mac/Windows" interface is uncongenial to them *because* it divorces them from the command line and from the heart of the machine which they know so intimately. As our local "micro- computer specialist" said, "I do not think faculty will want Macintosh computers, because on the Macintosh you cannot have direct access to the oper- ating system." I must emphasize that I most surely do not want to start some pointless flame war about VMS/Unix/Mac/DOS interfaces: isn't the interesting question here precisely the debate about why some of us prefer one interface to such a high degree, and the subsidiary question of why we feel so strongly about it? Why do I feel impelled to answer Philip's comments? Is it just consumer loyalty? I don't think so... And yes, I still use VMS and Unix :-) David Graham (2) -------------------------------------------------------------------- Date: Thu, 14 Apr 88 03:58 EST From: PROF NORM COOMBS Subject: Mouse or keyboards Could not resist throwing in another perspective on using a mouse vs using a keyboard. It seems to me that the keyboard is a more linear, rational process while the mouse is more visual and intuitive. I would love to see a psychological study on right brain, left brain and mouse or keyboard preference. Or is this way off the wall? Norman Coombs (3) -------------------------------------------------------------------- Date: Wednesday, 13 Apr 1988 23:07:19 EDT From: "Patrick W. Conner" Subject: MacSonnet The Poet Composes a Sonnet on His Mac Shall I compare thee to an IBM PC? Thou are more friendly and more temperate: Rough DOS doth encrypt the darling disks A &B, A hind'rance to my work I've learnt to hate: Sometimes in code the IBM can shine, But often is his grey complexion dimm'd By graphic Mac from which he graphically decline, By fate, or Apple's interface, untrimm'd; But thy unclon.ed excellence shall not fade, Nor lose possession of the users thou ow'st; Nor shall Amiga brag thou wander'st in his shade, When in increasing bytes to pow'r thou grow'st: So long as men can breathe, or eyes can see, So long lives Mac, and this gives life to me. ========================================================================= Date: Thu, 14 Apr 88 20:16:47 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Visit report: Knowledge Warehouse Project (22) (1) -------------------------------------------------------------------- Date: 14-APR-1988 09:23:24 GMT From: Lou Burnard A company called Mandarin Communications has for the last year been running the Knowledge Warehouse project, a pilot investigation into the feasibility of archiving publishers' typesetting tapes as a quasi- commercial, semi-philanthropic venture. Despite the billing for this one day conference organised by Mandarin at SOAS ("To review the next steps in establishing the National Electronic Archive"), I did not come away with the impression that the interests of all three parties were being (or were likely to be) equally well-served by the proposed Archive, or Warehouse. [The full report is now available on the file-server, s.v. WAREHOUS REPORT.] ========================================================================= Date: Thu, 14 Apr 88 20:20:03 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: The New OED (25) (1) -------------------------------------------------------------------- Date: Wednesday, 13 Apr 1988 22:52:40 EDT From: "Patrick W. Conner" I agree with Geoffrey Wall. I'd like to have the OED available as soon as possible, too, and I won't be able to afford a personal set-up. Could we on Humanist create a petition to BRS or Dialog or another vendor (or vendors, if those are not internationally accessible) to urge them to supply the OED at an hourly rate? If a large number of HUMANIST members indicated that they would like to have access to the OED on-line, and that they might even subscribe to a commercial database vendor to get it, I'll bet it would be made available fairly quickly, assuming that Oxford UP were willing. How should we go about this? I'll be glad to keep a file of short endorsements to be forwarded to a vendor once we've built up enough to be persuasive. Patrick Conner VM47C2@WVNVM West Virginia University ========================================================================= Date: Thu, 14 Apr 88 20:23:50 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Classification of Topics on HUMANIST (22) (1) -------------------------------------------------------------------- Date: Thu, 14 Apr 88 10:36:49 GMT From: Max Wood TOPIC : TOPIC I currently recieve HUMANIST into the dark plumbing of Primos Mail via the none too gentle auspices of ISOCEPT. I already run software that sorts and auto-edits my mail so as to remove header junk and topics that don't interest me. Unfortunately I have to rely upon the Subject field of the header for these choices. Therefore I heartily agree with the TOPIC classification that has been proposed for HUMANIST and should it be introduced I would certainly be willing to pass on my Prime based software for pre-processing mail to any that might also be in a Primos environment. ========================================================================= Date: Fri, 15 Apr 88 23:54:04 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Interfaces (145) (1) Date: Fri, 15 Apr 88 19:38:45 EDT (54 lines) From: Jeffrey William Gillette Subject: Of Mouse and Men (pt. 2) (2) Date: Fri, 15 Apr 88 15:24:13 PDT (20 lines) From: cbf%faulhaber.Berkeley.EDU@jade.berkeley.edu (Charles Faulhaber) Subject: Re: Interfaces (3) Date: Fri, 15 Apr 88 15:25:38 EDT (48 lines) From: Willard McCarty Subject: Interfaces (1) -------------------------------------------------------------------- Date: Fri, 15 Apr 88 19:38:45 EDT From: Jeffrey William Gillette Subject: Of Mouse and Men (pt. 2) Yes, my first note on interfaces was intended to be somewhat incendiary. I hoped to provoke a response like that of Philip Taylor, and I suspect that he is not the only one with similar feelings. No, I am not interested in theoretical disputation over: Is the MacWindows interface better than a command line approach, or vice versa. As a sometime developer of computer applications I am more interested in the questions: What types of people prefer the MacWindows interface, and what types of things do they want their computers to do? Similarly, what types of people prefer a command-line interface, and what types of tasks do they want their computers to assist them with? I am rapidly coming to the conclusion that most of the "common wisdom" regarding the merits of one style interface over the other are, in fact, nothing more than marketing slogans and old wives' tales (apologies to any old wives who happen to be reading this!). The Mac is "simpler and more intuitive." People who prefer a command-oriented approach have "vested interests in preserving the status quo." The real issue is efficiency - keyboard vs. mouse (this last myth will, I suspect, be put to rest by Microsoft, which has mistakenly redesigned Windows 2 to make the keyboard behave like a mouse). On the other hand, personal observation has convinced me that there are significant distinctions in personality and temperament between those who are attracted to the MacWindows style, and those who prefer a command-oriented interface. The earliest and most loyal converts to the Mac were hackers (which does not mean, as Diane Balestri correctly points up, that everyone who buys a Mac today is a hacker). Over the last 3 years, almost all genuine innovation in computer software has originated on the Mac and later been ported to the PC (e.g. desktop publishing, hypertext). While I do not think it true that Mac users are more creative than their PC counterparts (although this myth is popular in some circles), I think I can observe a certain pattern of creativity, or of curiosity in the technology for its own sake, that seems much more common in Mac users than in users of command-oriented systems. In particular, both Philip Taylor and Norman Coombs echo an intuition I have come to, and can almost articulate: computer users who prefer the MacWindows interface tend to think spatially while users who prefer a command-oriented approach tend to think verbally. Obviously I am not foolish enough to think this generalization true of all users of Macs or PCs, but I do suspect it to be true of a large number of people who have their choice of environments. Any comments on this theory? (2) -------------------------------------------------------------------- Date: Fri, 15 Apr 88 15:24:13 PDT From: cbf%faulhaber.Berkeley.EDU@jade.berkeley.edu (Charles Faulhaber) Subject: Re: Interfaces (132) I am not a technical computing specialist (i.e., I couldn't program my way out of a paper bag), but I started with UNIX, got a PC when they first came out, have worked with CMS, and now have a Sun (running suntools) which is an icon-based system similar to the Mac. Suntools wins hands down. And my experience is that mouse-based systems are the tool of choice for people who are interested in getting a piece of work done rather than in hacking. Charles B. Faulhaber Department of Spanish UC Berkeley CA 94720 bitnet: ked@ucbgarne internet: cbf@faulhaber.berkeley.edu (3) -------------------------------------------------------------------- Date: Fri, 15 Apr 88 15:25:38 EDT From: Willard McCarty Subject: Interfaces Somebody recently observed that the point of machine interfaces is to humanize the technology. I think that this is not just true of computers but is the general human tendency to transform nature into art, to put a human face on (or find it in) a very inhuman world, i.e., to create. What face can we put on (or find in) something except our own? Perhaps it's true that computers are now complex enough to begin to manifest aspects of human personality. If this is so, then we should look to the cultural and social contexts in which certain computers were developed. The Macintosh makes a good study in this regard because the company that developed it (yes, I know about Apple's intellectual debt to Xerox PARC) has kept the borrowed design successfully to itself. It is interesting to look back on the original article in Byte, for example, in which the Mac was announced, and to notice the people involved. It's easy to make silly generalizations about this fairly coherent group, but they did come out of a particular historical moment that some of us have shared and will recognize instantly, whether or not we feel sympathy with it. The corresponding case of the IBM design is more difficult to bring into focus, at least for me, and I wonder why. I don't think we'll get anywhere, or anywhere very interesting, by putting value judgments on these machines. We may get somewhere and contribute substantially to the advance of what is loosely called "interface science," however, by probing more deeply into the reactions people have, not just psychologically but culturally. To return to the Mac, we could ask about the relationship joining the apocalyptic dreams of ca. 1965-1975, their translation into the technical subculture, particularly in California, and the sort of concerns and aims of the designers of this really revolutionary machine. Could we then see in the machine a reflection not only of the designers' personae but also of what has happened to the social revolution in which they participated? I am not name-calling, rather trying to account for the rather amazing degree to which the inert micro-boxes of both major kinds call forth passionate zeal and seem to demand unreasoning belief. Even if no significant contribution can be made to this point directly, at least it seems possible that we might come to understand how building better programs is tied to more accurate professional self-knowledge. Willard McCarty mccarty@utorepas ========================================================================= Date: Fri, 15 Apr 88 23:56:29 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Electronic OED (63) (1) Date: Thu, 14 Apr 88 21:10:18 MST (16 lines) From: Mark Olsen Subject: OED NOW (2) Date: Fri, 15 Apr 88 14:26:56 EDT (30 lines) From: Darrell Raymond Subject: access to the electronic OED (1) -------------------------------------------------------------------- Date: Thu, 14 Apr 88 21:10:18 MST From: Mark Olsen Subject: OED NOW You might be better off trying to get your library to access the OED on-line. The ASU library has already set up the Grollier's Encyclopedia on the main card-cat. computer. A user can access it from any terminal in the library and (real soon now) on dial-up from any remote computer. The library is currently exploring putting the OED on the same machine. This is far better than having it on DIALOG since a) there is no connect charge and b) you can use it at home, the office or the library, and c) it is free, just another library service. The Grollier's Encyclopedia works well and I am looking forward to having the OED. Any guesses at how long it would take to download the OED at 2400 baud???? Mark (2) -------------------------------------------------------------------- Date: Fri, 15 Apr 88 14:26:56 EDT From: Darrell Raymond Subject: access to the electronic OED In response to Geoffrey Wall : >The electronic OED will not, they think, be in competition with the New >OED in printed form. It will be used in different ways, as a 'list-maker' >rather than a 'page-turner'. ( Who would want to read a 60,000 word entry >on screen?) Given that you have a workstation running X.10, it's much easier to read a large entry on screen than on paper, because you can quickly display lots of text and dynamically alter its format or elide various parts of the entry. Our display software supports all the fonts and special characters necessary to produce an accurate proof of OED text, and does it in real time. >Personally, not being able to afford a PC clone, let alone a dedicated >three-drive CD player, I'd like to have on-line access to the OED as soon >as possible. Along the lines promised by the Institut National de la >Langue Francaise, for general access to their lexicographical database at >Nancy. Meanwhile, perhaps the NEW OED Centre at Waterloo would consider >some kind of access? Ownership of the content of the OED is retained by Oxford University Press, hence access permission is granted by OUP and not by the University of Waterloo. Scholars are most welcome to visit and access the online OED, but we cannot unilaterally extend this to general online access. ========================================================================= Date: Fri, 15 Apr 88 23:59:18 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: ALLC/ICCH Conference (33) (1) -------------------------------------------------------------------- Date: 15 April 1988 From: Ian Lancashire Subject: Call for papers, ALLC/ICCH Conference CALL FOR PAPERS Association for Computers and the Humanities Association for Literary and Linguistic Computing 16th International ALLC Conference -- 9th ICCH Conference 6--10 June 1989 University of Toronto, Toronto Ontario, Canada The 16th International ALLC Conference and 9th International Conference on Computing and the Humanities will be held conjointly at the University of Toronto from June 6th to 10th, 1989. Papers on all aspects of computing in linguistics, ancient and modern languages and literatures, history, philosophy, art, archaeology, and music are invited for presentation at the conference. [The full announcement is now available on the file-server, s.v. ALLCICCH CONFRNCE.] ========================================================================= Date: Sun, 17 Apr 88 11:35:03 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Hiatus until Wednesday &c. Dear Colleagues: I will be away until Wednesday of this week, and so HUMANIST will be silent until that evening. ListServ, being mechanical and rooted to the spot, will not sleep nor leave the country with me, so you are welcome to send in contributions to HUMANIST, which I'll bundle up and send out on my return. I will also be away from HUMANIST for a much longer period, approximately two months, beginning the middle of May. Rather than allow HUMANIST to run automatically (and so court the floods of electronic junk mail some of us will remember), or to shut it down for this period, I have arranged for a local HUMANIST and good friend, Abigail Young, to take over as editor temporarily. For various reasons she'll do this as , so you should notice no interruption in service. The temporary change may provide a test of my hypothesis that we impress our personalities on the things that we do with computers. So if HUMANIST's editorial touch becomes more graceful, witty, incisive, and good humoured after 15 May, you'll know why. In any case, I am very grateful to Abby for agreeing to take on the job. Willard McCarty mccarty@utorepas ========================================================================= Date: Wed, 20 Apr 88 20:56:04 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Notices (67) (1) Date: Tue, 19 Apr 88 15:32:59 +0200 (17 lines) From: Prof. Choueka Yaacov Subject: ALLC/AIBI Conference (updated posting) (2) Date: Wed, 20 Apr 88 12:36:44 est (38 lines) From: munnari!fac.anu.oz.au!nasdling@uunet.UU.NET (DAVID NASH) Subject: Update on Australian non-link... (1) -------------------------------------------------------------------- Date: Tue, 19 Apr 88 15:32:59 +0200 From: Prof. Choueka Yaacov Subject: ALLC/AIBI Conference (updated posting) ALLC/AIBI 1988 Joint Conferences: Fifteenth International Conference on Literary and Linguistic Computing and Second International Conference on Computers and Biblical Studies. June 5-13, 1988, Jerusalem * Sponsored by: ALLC - Association for Literary and Linguistic Computing AIBI - Association Internationale Bible et Informatique * With the participation of: ACH - Association for Computing in the HUMANITIES [An updated posting is available on the file-server, s.v. ALLCAIBI CONFRNCE.] (2) -------------------------------------------------------------------- Date: Wed, 20 Apr 88 12:36:44 est From: munnari!fac.anu.oz.au!nasdling@uunet.UU.NET (DAVID NASH) Subject: Update on Australian non-link... [The following will interest those of you who remember the loss of our members from New Zealand due to the excessive cost to them of receiving international e-mail. For those who don't, the following is more or less self-explanatory. It was sent to me by David Nash, formerly of MIT and now in Australia. While at MIT he was investigating ways of getting HUMANIST to our colleagues Down Under; what he has subsequently discovered confirms what the New Zealanders told us about the cause of their resigning. -- W.M.] Here is an exchange which you find informative: 20-APR-1988 09:46:51 To: NASDLING Subject: Re: News -- sci.lang > I've been acquainting myself with the news available here, which is a > nice selection, but only part of Usenet, right? What's involved in > adding something like sci.lang ... You are right. The newsgroups you mention come under the category of "privately imported", ie you don't get them unless you or someone else pays for them. The cost is around $250 per megabyte of news. You (or ANU -- we're not picky) pay in advance, preferably for at least 2 Mb of news, and we arrange to bring it in. Everyone in Oz will get it, BTW, which might have a side effect of encouraging others to pay for it (when it is due to vanish!!!!) ========================================================================= Date: Wed, 20 Apr 88 20:59:34 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Query: citing e-text (20) (1) -------------------------------------------------------------------- Date: Wed, 20-APR-1988 18:14 EST From: Wade Schuette Subject: Legal/practical question: How cite w/o page #? (10 lines) Today's New York Times (4/20/88 p D1) has an article on the lawsuit between West Publishing Co. and Mead Data Central, Inc. (Westlaw vs Lexis) over Lexis' use of West's page numbers as standard for citing legal decisions. This seems a far more general questions, with many issues related to the protection of intellectual property. Question: does anyone know of good ways to cite text once it becomes stored electonically and "page #" becomes a rather meaningless term? Who does what now? ========================================================================= Date: Wed, 20 Apr 88 21:02:25 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Software, mostly w-p (70) (1) -------------------------------------------------------------------- Date: Mon, 18 Apr 88 12:21:57 GMT From: Sebastian Rahtz I just read 62 pages of HUMANIST and have finally caught up with the WP/software minds debate, concerning which I wish to make four observations; but before I do so, may I in passing refer to some nonsense about refrigerators I skimmed over? On our departmental machine (named Queen Victoria, don't knock her..), our refrigerator has an account; he is called Vincent, and performs a sterling job in keeping the coffee milk cold, and cooling bottles of wine for special occasions. Vincent (vf@uk.ac.soton.cm) often contributes to the departmental bulletin board, sends occasional mail, and generally keeps a lively look out on our affairs. Vince is not yet on the payrol, but we are working on getting him elected to the University Senate. (This is all true). But as for WP: a) The frenetic NotaBene adulation neglects the fact that, in this country at least, the *support* for Nota Bene from the distributor is laughable. I chose my washing machine on the reputation of the supplier for backup, not for features - if I had to choose a word-processor (god forbid!), I'd buy Microsoft Word because I trust Microsoft to stand by me. OK, I know on that argument I'd buy IBM mainframes, but I hope the point is taken about support. b) The mind-in-the-software - Willard says that software, once designed, remains faithful to the initial concept. Yes, but what about software designed to meet a prescribed task? Take databases - if (when?) the ISO standard for SQL is adopted, will we be able to tell the difference between Oracle, DB2 and Ingres? Or C compilers - as someone said, the spirit of C is a personality thing, but the difference between Microsoft and Borland isn't. To pursue that, once one accepts the idea of the integrated compiler a la Borland, you can't really do much about it. I suspect also that Willard is not a MacMan; I take his point about the devious mind of the NotaBene creator being visible still, but would it be on a Mac? I don't see much personality in Mac programs (with the honourable exception of Illustrator). Maybe others do. c) De Rose states the case against WYSIWYG more eloquently than I can; I'd go all the way with him and the others, and further, to say that academic word-processing is a dead-end which we will soon see the death of, thank god. There's text-processing, preparing material for electronic publication (from BBs and mail to generic markup for databases and printing) and there's computer-assisted design (of which 'DTP' design of leaflets for paper printing is a a subset). The use of a computer as typerwriter for letters, memos and the like (which is all that WP programs are any good for) should wither away, I hope. Even letters fall better into the generic markup game - a reversion to the 'take a letter, Miss Jones, and don't bother me with the details' system, which I am sure we will all agree is *much* better. d) I am editing a conference proceedings at the moment; all the contributors so far have sent disks or e-mail files. Not *one* of them has marked-up his/her text in a *complete* form for sensible editing by me (and yes, I did send out guidelines). Even the one who used troff with a sensible macro package did not understand about heading macros, but put in his own spacing and fonts. As for bibliographies.... The moral is, as I sit over a hot PC-Write reformatting them (forget about all those monster packages, PC-Write is *wonderful* for daily editing), that however many features you give people in their WP package, they probably won't use them! Thank god no-one has sent me newspaper column formatted text. Next year I'll send them all Coombs et al.'s CACM article on generic coding. Sebastian Rahtz, Computer Science, University, Soutampton, UK PS back to the 'humour': the TeXhax mailing list produced a very witty April 1st issue; why nothing on HUMANIST? Are we all that boring? ========================================================================= Date: Wed, 20 Apr 88 21:12:39 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: Job posting (17) (1) -------------------------------------------------------------------- Date: 20-APR-1988 16:21:55 GMT From: Sarah Rees Jones Subject: Advert for Information Officer, Computing Service U N I V E R S I T Y O F Y O R K COMPUTING SERVICE [Job posting for an "Information Officer" will be found on the file-server, s.v. INFOFFER JOB.] ========================================================================= Date: Wed, 20 Apr 88 21:15:59 EDT Reply-To: Willard McCarty Sender: HUMANIST Discussion From: Willard McCarty Subject: New OED, a brief history (83) (1) -------------------------------------------------------------------- Date: 20-APR-1988 14:23:10 GMT From: Grace Logan New OED Project In 1984, for the 100th anniversary of the publication of the letter "A," Oxford University Press announced its intention to computerize the OED, putting the 12-volume text and the 4-volume Supplement into machine-readable form by the end of 1986, and then publishing an integrated OED by 1989. At the time, this seemed like science fiction, but in fact, at the moment, the project is right on schedule. The fourth volume of the Supplement was published in 1986, and almost at the same time the machine-readable text of the 12 volumes of the OED and the 4-volume Supplement had been inputted. This was accomplished by the International Computaprint Corporation (ICC) of Fort Washington, Pa., between January 1985 and June 1986. The text contains almost half a million definitions illustrated by two and a half million quotations. The ICC had inputted some 350 million printed characters at less than the contractual rate of 7 errors or less in 10,000 keystrokes--in the latter batches a rate of 4 errors or less was regularly achieved. Since then, the text has been proofread by OUP, and Murray's notation of pronunciation has been converted into the International Phonetic Alphabet. By last summer, OUP had completed the automatic integration of the 4-volume Supplement into the body of the l2-volume OED and made the first pass of the cross-referencing system. The editors are now working on the manual integration of the Supplement into the OED, that is, the 20 per cent by volume that could not be done automatically. To assist the editors in the integration of the text, the OUP computer group has developed a computerized editing system, whose name tells it all: Oxford English Dictionary Integration, Proofing, and Updating System, or OEDIPUS. So far there are no signs of a tragic conclusion. Indeed, the New OED computing system has already won the British Computer Society Applications Award for 1987 for its innovative use of the computer. OEDIPUS as part of this system provides a means of examining and correcting the text of the Dictionary on computer screens. It allows the editors not only to insert new words into the dictionary but to enter the editors' additions, corrections, insertions, deletions at any point within an entry. As this is being done, it is sent to the typesetters and proofread again, after which further corrections are made. In December 1987, OUP and Bowker and Tri-Star made available the text of the unintegrated 12-volume OED produced by ICC on two CD-ROM disks, A-N and O-P, costing about $1250 for the set with documentation. The entire dictionary could have fitted onto one disk, but adding indexes and inverted files required another disk to be used. Bowker and Tri-Star had devised a system that allows the user to make inquiries concerning the data stored on the CD- ROM. The program divides the screen into three windows: the top for query and field, the middle for viewing the results of a search, and the bottom for identifying function keys. The text can be searched very quickly and easily on eight fields derived from the approximately forty fields found in the dictionary entries. These include LEMMA, ETYMOLOGY, SENSE, and LABEL (part of speech, usage, specialized language), and, from the quotation field, DATE, AUTHOR, WORK, and QUOTATION TEXT. The results may be viewed in the saved queries window. The system is still being developed and is fairly rudimentary at the moment. For instance, since the earliest date of a word's occurrence is not marked in the data, one cannot find out what words in a particular field first entered the language. Similarly, it is impossible to search on the ends of words or within a word. No doubt this will be corrected in time. The integration of the OED and its four Supplements is expected to be completed and published in book form in the Spring of 1989, extending to 20 volumes. The second edition of the OED will then be available on CD-ROM. In the meantime, the University of Waterloo is developing two inquiry tools, Generalized OED Extracting Language (GOEDEL) and PAT, which together form a very powerful means of searching and retrieving information from the database. =========================================================================