Experiment: computing the number of characters per line (DR 11-0011)

[MURATA Makoto]

Folks,

I tried to find the rule for computing the number of characters per
line by experiment.  I changed the value of the charSpace attribute of
the docGrid element using oXygen.  (I did not use MS Office to change
the value of this attribute, since it also changes other values thus
hampering my intention).  Then, I invoked MS Office to see how many
characters fit in a line and also to examine the page layout setting
view to see the number of characters per line setting.

Here is the result.

charSpace  the number of characters
    0       18
 1502       18
 1503       17
 4121       17
 4122       16
 7065       16
 7066       15
10403       15
10404       14
14219       14

The description of this attribute in 29500-1 provides a rule for
computing the desired character pitch.   By multiplying the
number of characters and the computed desired character pitch, I
computed twips per line.

twips per line
3780
3912.011719
3694.760742
3912.075195
3682.03125
3911.953125
3667.529297
3911.938477
3651.210938
3912.001953

But what is the column width?  Here goes.

pgSz/@w (twips)           11906
pgMar/@right(twips)        3997
pgMar/@left (twips)        3997
col width (twips)          3912 = 11906 - (3997 + 3997)

Twips per line computed above does not exceed the col width (twips).
Although 3912.011719, 3912.075195, and 3912.001953 are exceptions, the
difference is very small.  So far so good.

But why can't we have just one more character per line?
In other words, does
  (1 + the number of characters per line) * desired character pitch
exceed the column width?

Here is the computed values.

3990
4129.345703
3912.099609
4142.197266
3912.158203
4156.450195
3912.03125
4172.734375
3912.011719
4191.430664

I am afraid that 3912.099609, 3912.158203, and 3912.011719 are not so
large.  They are even smaller than one of the exception values shown
above.  Why can't we have one more character in line?

I am confused.  Probably, to understand subtle deviations, we have to
read the source code.


Regards
Makoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20150808/0e45910a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: experiment-2015-08-08.xlsx
Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
Size: 12508 bytes
Desc: not available
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20150808/0e45910a/attachment.xlsx>