Path: utzoo!utgpu!watserv1!watmath!iuvax!rutgers!att!cbnewsm!mls
From: mls@cbnewsm.ATT.COM (mike.siemon)
Newsgroups: comp.sys.mac
Subject: Re: Text file madness: diagnosis & prescription.
Summary: trying Brandon Allbery's experiment
Message-ID: <8454@cbnewsm.ATT.COM>
Date: 11 Jan 90 04:59:20 GMT
References: <2706@aecom.yu.edu> <5900@ncar.ucar.edu> <1998@eric.mpr.ca> <1990Jan7.172731.12580@NCoast.ORG>
Distribution: na
Organization: AT&T Bell Laboratories
Lines: 72



In article <1990Jan7.172731.12580@NCoast.ORG>, allbery@NCoast.ORG
(Brandon S. Allbery) writes:

> +---------------
> | In article <8315@cbnewsm.ATT.COM> mls@cbnewsm.ATT.COM (mike.siemon) writes:
> | >The example of reading an arbitrary file is just that, an example of a
> | >much larger general problem.  If I have a file I may want to use it in
> | >a number of different ways, totally unconstrained by the file's origin
> | >or by my intentions when I create the file.
> +---------------

> No -- you can override it very easily:  select the application and the file
> you want opened, then select "Open" from the File menu.

This is a bit obscure.  If I try to select, e.g., the file generated by my
scanner software *and* a word processor, I find that the Finder doesn't *let*
me extend selections across different folders.  (surely you *don't* mean for
me to keep everything in the *same* folder? :-))

I can of course drag one of these to the *other's* folder, or both onto the
desktop (and then must remember to drag them back where I really want them
:-))  Now if I click and shift click (or whatever) to select them both, a
double click may launch the application *with* the file I want (at least,
Word did this for me; I really don't know if this is general or not.  One
*never* knows whether something on the Mac is general or not.)  Thus, I do
not need the File menu's Open line -- at least in this case.  I was unaware
of the possibility until you mentioned it; are you suggesting this is known
to ordinary Mac users?  It is not mentioned in Apple documentation that I
have read.  Or are you suggesting that ordinary Mac users will divine this
incantation from ordinary use, or that *all* combinations of selections and
menu choices give some reasonable result?  Or that I am supposed to spend
hours "playing" with this box to "see what happens" in all the potential
situations that could be generated by mouse operations? (And what then; must
I write my own documentation to record these experimental findings?)

My scanner software tries to by "nice" to me by allowing its output to be
in Word or MacWrite format.  So does my communications software, but neither
of them knows about Nisus, for example.  (And why *should* they?)

But suppose I want to add the numbers that happen to be in the third column
of this text file I just scanned in.  A spreadsheet program might or might
not be able to cope with this file.  If WordPerfect on the Mac has the same
facilities as on PCs, *it* might be able to do the add (after some silliness
or other; for all I know Word may have buried arithmetic capabilites, too --
possibly hanging off the KitchenSink menu.)

Do you begin to see what I'm talking about?  Either everything has to *know*
about everything else -- impossible unless you forbid innovation, as in the
Lisa -- or else one has to have general combinators or constructors for the
complex operations that actually describe what people *want* to do.  It is
no answer to say that I can do anything by step-by-step elemetary operations.
Theoretically, the only operations editors *need* are deletion and insertion.
I would *not* recommend an editor that used only the elementary operations.

UNIX has major deficiencies, particularly in a visual interface.  I am not
aruging that the Mac should be "like" UNIX in any obvious way (and UNIX too
has the problem of one not knowing what will generalize -- one of the BUGS
sections in an early UNIX command notes that its syntax is "rebarbative.")
But Ritchie and Thompson's original ACM article about UNIX makes one very
good point that is almost always lost in these later discussions.  The idea
of *all* files being byte sequences, connected via stdin/stdout pipes in a
*general* combinator, has all the power that algebra does in mathematics.

The Mac draws beautiful Roman numerals; one is almost seduced into thinking
that the art of reckoning with these numerals is all there is to mathematics.
-- 
Michael L. Siemon		In so far as people think they can see the
cucard!dasys1!mls		"limits of human understanding", they think
att!sfbat!mls			of course that they can see beyond these.
standard disclaimer				-- Ludwig Wittgenstein