@Section
   @Title { Alternative conventions for white space }
   @Tag { white }
@Begin
@PP
As Section {@NumberOf spaces} explains, when two objects are separated
by one or more white space characters (spaces, tabs, and newlines), this
same amount of white space will separate the two objects in the output.
@PP
Two other conventions for interpreting these white spaces have been
used in other document formatting systems.  Roughly, they are:
@ID @Tab
    @Fmta { @Col A ! @Col B }
{
@Rowa
    A { troff }
    B { Like Lout, except that at every point where a sentence ends at
the end of an input line, add one extra space in the output. }
@Rowa
    A { @TeX }
    B { Replace all sequences of two or more white spaces by one.  Then,
at every point where a sentence ends, whether or not it is at the end
of a line, add one extra space in the output. }
}
Lout offers these two alternative conventions by means of the
initialspace. @Index { @Code "@InitialSpace" option }
@Code "@InitialSpace" option.  This is similar to the
@Code "@InitialFont" option described at the end of Section
{@NumberOf fonts}, in that you can set it at the beginning of your
document, like this:
@ID @OneRow @Code {
"@SysInclude { doc }"
"@Document"
"    @InitialSpace { lout }"
"//"
"@Text @Begin"
"..."
"@End @Text"
}
or you can set it in the setup file.  The above example shows the
default value, {@Code lout}, which produces Lout's usual spacing;
lout.space @Index { @Code lout spacing }
troff.space @Index { @Code troff spacing }
tex.space @Index { @Code tex (@LaTeX) spacing }
the alternative values are @Code "troff" and {@Code "tex"}.
@PP
How to tell whether a sentence has ended is a vexed question.  For
the @Code "troff" method, Lout looks for a word at the end of a line
ending in one of `.', `:', `?', or `!' optionally followed by either
a right quote character or a right parenthesis.  Actually, this depends
on the current language (Section {@NumberOf languages}); the rule just
given is for English, and other languages may differ.
@PP
The @Code "tex" rule for where a sentence ends is slightly more
complicated.  Lout looks for a word, not necessarily at the end
of an input line, which ends as described for @Code "troff" but
in addition has a lower-case letter preceding that.
@PP
You must use a paragraph symbol, such as @Code "@PP" or {@Code "@LP"},
to separate paragraphs.  The common convention of other systems,
that a blank line marks a paragraph, is never true of Lout.
@PP
Whatever rule is adopted, there are occasional exceptions where you
will have to indicate explicitly whether you want an ordinary space
or a between-sentences space.  For this there are two symbols,
@Code "~" (ordinary space) and {@Code "~~"} (between-sentences
space).  For example,
@ID @Code "Dr.~Kingston"
will produce an ordinary space between the two words, even with
@Code "tex" which would otherwise consider that spot to be the end
of a sentence.  Spaces adjacent to these two symbols have no effect on
the result.  However, @Code "~" produces an unbreakable space (that
is, one that will never be replaced by the end of a line) in contrast
to just leaving a space, which is breakable.
@End @Section