Re: CVS commit: src/share/man/man9

["Valeriy E. Ushakov" <uwe%stderr.spb.ru@localhost>]

On Mon, Mar 16, 2009 at 18:39:27 +0100, Joerg Sonnenberger wrote:

> On Mon, Mar 16, 2009 at 05:30:33PM +0000, Valeriy E. Ushakov wrote:
> > Log Message:
> > Don't restrict column width of the last column of the methods table.
> > nroff does the right thing and properly wrap descriptions that don't fit.
> 
> Please revert the first part. This is invalid markup as it doesn't list
> all columns.

As far as I can tell (my roff is rusty) effective troff semantic of

  -column ... <stringN> ...

is

1) to restrict the width of column N
2) to set the tab position and indent to where column N+1 starts

that is, Nth spec affects both columns N and N+1, or in other words,
Nth column gets its width from spec N, but its position from spec N-1.

When the spec for the last column is omitted from -column args the
effective semantic is (again, as far as I can tell)

1) make the rest of the page width available to this last column

2) since this column is the last one, there's no need to set the tab
   position and indent for (nonexistent) column N+1

In other words, for the last column we get its position from previous
spec and we don't want to restrict the width, so all is fine and
consistent.  I.e. from roff standpoint omitting last column spec is
fine and does the right thing formatting-wise by using the rest of the
page width.

So i think that when you do the exercise of reinterpreting mandoc
markup as semantic markup you should assign you abstract semantic
accordingly.

Intuitively, there must be a way to specify the width of the last
column to be "the rest of page width".  The man page in question
demonstrates the need for this quite convincingly, I think.

So please, fix the abstract semantic you assign to mdoc instead. :)

SY, Uwe
-- 
uwe%stderr.spb.ru@localhost                       |       Zu Grunde kommen
http://snark.ptc.spbu.ru/~uwe/          |       Ist zu Grunde gehen