indexing redux

[Carlo Caballero Carlo.Caballero@xxxxxxxx]

On Sun, 14 Dec 1997, Bruce L. Felknor wrote:

>     However, I am thinking (page proofs won't arrive for a week or two)
> of putting terse entries-cum-page-number on a lined pad instead of cards,
> then alphing them by computer, and elaborating and consolidating as
> appropriate.
>     Or does Xy4 (DOS) also limit sorting to 640 k?

Dear Bruce,
   I think that this way of sorting would work perfectly. As for memory
usage, my manual for Xy4 (DOS) indicates that 640K is still the base, but
"if you have expanded memory, it automatically uses up to 4MB of it" (p.
1-1).
   *(Parenthetical question for others on the list: Will XyWrite use
expanded memory under DOS 3.3, or do you have to have DOS 5 or higher?)
   It seems to me that XyWrite would probably be able to handle the
sorting task you've described with absolutely no problem. Sorting, as an
independent task, is something that XyWrite seems to do extremely well and
quickly. Let us know how it goes when you try it (number of items, size
of file, etc.); I, at least, would be curious to see how XyWrite responds,
particularly if your list requires a high sorting key. The reason I think
the chapters of my dissertation were too much for it is that I couldn't
reduce the sorting key very low, because I had "hand-made" index markers
 with entries such as:

Faur?-Fremiet, Emmanuel
Faur?-Fremiet, Philippe

Institut de Paris
Institut Catholique

And worse. So reducing the sorting key to six or seven, which might have
allowed XyWrite to extract an index without running out of memory,
defeated my index. And I was trying to do get XyWrite to do this complex
indexing across six, large files in a TYS@ job. I briefly thought of
making up coded abbreviations for longer entries; I might try doing that
one day, but I ended up forgoing the index altogether (it wasn't REQUIRED
for the dissertation, I just thought it would be useful).

Actually, I don't know that I WOULD ever try the abbreviating solution.
When I finish the book based on the dissertation, the publisher will send
me, in the last stage, page proofs, which will have no relation to the
pagination in my XyWrite files. The index will have to be done the old
fashioned way from the page proofs. The indexing functions could still be
helpful, however, to authors producing "camera-ready" copy. Bruce's
method is a good way of comprimising by taking advantage of XyWrite as a
sorter and formatter. If one could get comfortable enough, one could even
skip the note pad and type the entries directly into the computer. But
this is the sort of work I like to do on planes, comfy sofas, etc., and
the notepad is light, tiny, and easy.

Reading about Xy4's use of expanded memory makes me want to try putting my
dissertation chapters through its engine, unless someone here can tell me
that it will make no difference under DOS 3.3. And if it DOES make a
difference, you can be sure that I'll get a higher version of DOS (or
OpenDOS), because I use work with large files, and Xy4 could use some help
there. I started my dissertation under Xy3 and finished it (last year)
that way; the formatting was complex, and I didn't want to switch in
midstream. That's my main excuse for never trying indexing in Xy4. Also,
I didn't imagine that a program that created temporary files (that little
"X" in the corner, you know) for a mere directory (!) was going to be too
thrilled with chaining and indexing five chapters of about 100,000 bytes
each.

CC

#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#__#

Carlo Caballero
Research Associate
The Graduate School and College of Music
University of Colorado, Boulder

thyrsus@xxxxxxxx