optical scanning, cont. (78)

Willard McCarty (MCCARTY@VM.EPAS.UTORONTO.CA)
Tue, 25 Apr 89 22:44:56 EDT


Humanist Mailing List, Vol. 2, No. 891. Tuesday, 25 Apr 1989.


(1) Date: Tue, 25 Apr 89 11:22:30 EDT (13 lines)
From: jonathan@eleazar.Dartmouth.EDU (Jonathan Altman)
Subject: Re: optical scanning, cont. (150)

(2) Date: Tue, 25 Apr 89 09:30:59 -0800 (45 lines)
From: Malcolm Brown <mbb@jessica.Stanford.EDU>
Subject: comments on the Kurzweil 4000

(1) --------------------------------------------------------------------
Date: Tue, 25 Apr 89 11:22:30 EDT
From: jonathan@eleazar.Dartmouth.EDU (Jonathan Altman)
Subject: Re: optical scanning, cont. (150)

Well, I'll add another two cents to the scanning discussion, having
just seen Oxford's set of rules on scanning. The first is that
Oxford's rules and experiences are nearly identical to the Dante
Projects, and probably the best summarized explanation of how to
scan material effectively. The only change I have to add is that I
find the model 4000 relatively tolerant of type size (I find it
deals with about 9-14 point sizes of a typeface passably well).

Jonathan
(2) --------------------------------------------------------------47----
Date: Tue, 25 Apr 89 09:30:59 -0800
From: Malcolm Brown <mbb@jessica.Stanford.EDU>
Subject: comments on the Kurzweil 4000


I read the comments of the Oxford Kurzweil staff with
great interest. I agree with them on every point;
my experience with the K4000 has lead me to the same
conclusions.

I have the clear and distinct impression that the hardware
used in the K4000 is of the same vintage as the original
IBM XT. I think that accounts to some extent for how slow
it is.

I certainly hope that any future trainable machines will
not be stand-alone. This forces Kurzweil to write its
own disk operating system for the unit, a tremendous amount
of overhead for both them and us (and we pay for their
overhead, don't we?). My impression of the KDOS that runs
on the K4000 is not a good one. When our hard disk has
less than 1.5 megabytes of free space, the K4000 tends to freeze,
necessitating a restart. There are no user programs to
rebuild, reorganize or otherwise optimize the hard disk.
I can just imagine how fragmented the files must be after
a year of coming and going.

(yes, I know one could back up to floppies and then rebuild
the hard disk. Anyone who has suffered through the snail's
pace with which the K4000 performs a backup will know that
this is not an appetizing option, particularly if it has to be
done on a regular basis)

I'd be interested to hear from Oxford how reliable the
machines have been. In the year we've had our K4000, we've
had to have our terminal, hard disk and power supply replaced.
And the price for the maintenance contract is very high.

We have also come the conclusion that it is better to photocopy
and dismember books than to try to scan books directly. We
have the graphics tablet, but its use is so awkward that
we hardly ever use it.

Still, the K4000 can do the job, provided the quality of the
original is good. If not, performance rapidly goes downhill.