Re: File problem....

["M.W. Poirier" poirmw@xxxxxxxx]

Let me begin by stating that I have resolved the problem I was
having with my footnotes (at least for the moment, or until
someone suggests a better solution) by creating two files, one
for text and one for notes. The notes will be merged at the end
of the text file when the need to print arises.

This said, let me reply to some of Robert Holmgren's questions
from this morning:

1. I checked the position of the problem footnote, and it is at
the 61172 point.

2. I deleted a longish footnote above the problem footnote, and,
as expected, the problem disppeared.

3. I've noted that when I cursor from the top of the file to
the bottom of the file using the down arrow key, the problem foot-
note appears in the main text, preceded by the notation FN1.
However, if I jump to the end of the file, and curson to the top
using the up arrow key, the note is normal in all respects but
one. That is, the footnote number of the problem footnote is
that of the footnote immediately above the problem note. I
usually can change this to the correct number by entering the note,
as if to edit it, and then existing the footnote.

4. I have also noticed that if I split my long notes into three or
four notes, the problem disappears. I don't know if this is a
short duration phenomenon, ...or whether it could be the solution.

Robert Holmgren may have had another test for me to perform, but
I have yet to get to that one. This leads me to a broader issue.

Like many academics, early on in the preparation of a book, I
tend to keep the main body of the text as devoid of tangential
remarks and other notes as is possible. This allows me to keep
the main line of the argument that I want to make as clean and
direct as it can be, and it forces me to use footnotes to explore
ideas that are tangential to the main line of the argument.
Eventually, these tangential remarks will be draw into the main
text, if they fit, deleted if they do not seem appropriate, or
left as they are. What this means is that, at the beginning of
work on a new piece, I usually end up having a large number of long
notes, ...hence, the problem I had. Now, for two questions: 1)
How do the rest of you handle these matters, and 2) would I still
be faced with the same problem if I were to use a combination of
footnotes (set #1, #, #3), endnotes and labels to house my
tangential remarks?

M.W. Poirier


============================================================================
On Sun, 28 Jul 2002, Robert Holmgren wrote:

> ** Reply to message from "M.W. Poirier"  on Sat, 27
> Jul 2002 08:36:50 -0400 (EDT)
>
> OK, next several suggestions. First, I'd try Tom Hawley's idea, of adding or
> subtracting some dummy text -- a good chunk, say 500-1000 characters -- before
> the errant footnote, and see if the specific problem disappears. It won't solve
> your problem, but it will be diagnostic. This usually happens (when it happens,
> which isn't often) because text, such as your footnote or a programming frame,
> straddles a segment barrier in memory. Right around 64K, for example -- also
> around 57-58K, for reasons I've never understood. I'd be curious to know what
> the beginning and ending character positions of the footnote are (if you have U2
> installed, just command "pos"). Visual symptoms are dropped
> characters, invisible characters, wierd formatting, truncated files (sometimes
> the first pages are apparently missing). The underlying (filed) text remains
> fine, but the misleading/erroneous visual formatting leads you make unnecessary
> edits (in an effort to repair what isn't really broken except on-screen).
>
> If this is the problem, just add a dummy, non-printing NoTe to your text:  This is a NoTe> That will move the memory positions around. It usually happens
> that, in the course of normal editing, the memory positions get moved anyway,
> and the problem disappears.
>
> Second, you may have too many, or too long, footnotes. The size of the
> footnote buffer -- total size of all footnotes -- is 63-64K, and that's an
> absolute limitation, for which chain printing is the only solution. Try
> eliminating altogether one of the earlier footnotes in the text (a lengthy one,
> preferably). If the errant footnote suddenly reappears, and numbering is
> correct (minus one, of course), then you've found your problem: split the text.
>
> The third thing to do is to look at your footnote, and also right around the
> edges of it, at the DOS level (with a file LISTer). Sometimes a three-byte
> character will creep into text, for various reasons. And occasionally, that
> 3-byter will contain a guillemet or a screen control character (such as
> Ascii-253) as the second or third byte in the "character". That can wreak
> havoc. Somehow Editor "sees" that byte individually, instead of as part of the
> 3-byter, and responds to it -- even though it is never displayed.
>
> > I've noticed that it insists on having the same note
> > number as the immediately preceding note, which seems to mean that it
> > is not recognizing the previous note.
>
> Inspect the previous note as well at the DOS level. One thing to try, is put
> it in Draft mode, go to TOF, and then page down manually, one page at a time,
> until you reach your problem area, and see if the symptoms persist. Sometimes,
> if you jump directly to a position late in a heavily formatted file, Editor has
> trouble counting up all the previous footnotes (which, obviously, it has to do
> -- it has to "read" the whole file up to that point, in order to render the
> correct FN numbers, MoDes, fonts, well everything). It gets confused. It is,
> after all, only human.
>
> I don't think, with 291 guillemets, that you've reached the "too many formats"
> problem, which IIRC is 873, and which generally tells you that on the PRompt
> line anyway...
>
> If it's any comfort, after you've fiddled with it long enough, you'll solve it
> -- and it would be interesting to know exactly what the solution was.
>
> -----------------------------
> Robert Holmgren
> holmgren@xxxxxxxx
> -----------------------------
>
>