Re: Reverting various commits

[G. Branden Robinson]

Following up on an issue from last month...

At 2021-05-21T19:28:17+1000, G. Branden Robinson wrote:
> At 2021-05-20T16:09:59+0200, Ingo Schwarze wrote:
> > Hi Branden,
> > G. Branden Robinson wrote on Thu, May 20, 2021 at 01:23:18AM -0400:
> > 
> > > commit bf4b3dde3ba442a0cf52e986d2549f1dc47f43c5
> > > Author: G. Branden Robinson <g.branden.robinson@gmail.com>
> > > AuthorDate: Thu May 20 11:45:10 2021 +1000
> > > 
> > >     [mdoc]: Align header/footer spacing with man(7).
> > >     
> > >     * tmac/mdoc/doc-common-u (doc-end-macro): When continuously
> > >       rendering, increase page length by same amount we vertically
> > >       space after flushing a pending output line, for symmetry
> > >       with other spacing requests (and to prevent nasty surprises
> > >       analogous to those in Savannah #60611).
> > >     
> > >       (doc-header): Put 3 vees of space after the header in
> > >       continuous rendering mode, not 1 (and increase page length
> > >       accordingly).
> > 
> > I very strongly object.  Please revert all recent commits
> > manipulating mdoc(7) vertical spacing before and after header and
> > footer lines (i think there was at least one other during the last
> > one or two days).
> 
> I can't take this literally.  As part of the above quoted material
> there is the following.
> 
> > >     * tmac/mdoc/doc-common-u (doc-end-macro): When continuously
> > >       rendering, increase page length by same amount we vertically
> > >       space after flushing a pending output line, for symmetry
> > >       with other spacing requests (and to prevent nasty surprises
> > >       analogous to those in Savannah #60611).
> 
> If you'll take the time to read Savannah #60611[1] I believe you will
> agree that I found and fixed a rendering problem.  Maybe you'll need
> to do some experimentation on your own to be convinced.  I sliced my
> changes very finely while tracking it down and I'm as confident as I
> can be without contriving an mdoc document exhibiting it--to which I
> did give some consideration, before deciding that groff_mdoc(7) is
> sufficiently lengthy and exhibits sufficient exercise of the mdoc(7)
> feature set that I'd catch regressions in my usual process[2].
[...]
> Traditional man(7) never had continuous rendering mode.  As far as I
> know, in the context of man(7) documents, it's groff's innovation, and
> we can therefore change it.
> 
> I have no objection to doing so; the main reason I didn't was that I
> didn't want to disturb existing practice.  I get the feeling you'll
> tell me I respect all the wrong traditions.  ;-)

I've now taken care of this.  Here is a diff between the relevant macros
just prior to the commit quoted above, and now.  As I hope is obvious, a
straight reversion was not appropriate.

diff --git a/tmac/mdoc/doc-common-u b/tmac/mdoc/doc-common-u
index 3d6cfb57..bd8e88c4 100644
--- a/tmac/mdoc/doc-common-u
+++ b/tmac/mdoc/doc-common-u
@@ -934,8 +934,10 @@
 .de doc-header
 .  ev doc-env-dh
 .  doc-setup-page-layout
-.  if !\n[cR] \
+.  if !\n[cR] \{\
+.    \" Re-use doc-header-space between page top and header.
 .    sp \n[doc-header-space]u
+.  \}
 .  nr doc-reg-dh \w'\*[doc-caption-font]\*[doc-header-string]\f[]'
 .  nr doc-reg-dh1 \w'\*[doc-caption-font2]\*[doc-volume]\f[]'
 .  if (\n[doc-reg-dh] + \n[doc-reg-dh1] + \n[doc-reg-dh] >= \n[.lt]) \{\
@@ -955,8 +957,10 @@
 .  tl 
\*[doc-caption-font]\*[doc-header-string]\f[]\*[doc-caption-font2]\*[doc-volume]\f[]\*[doc-caption-font]\*[doc-header-string]\f[]
 .  ie !\n[cR] \
 .    sp \n[doc-header-space]u
-.  el \
+.  el \{\
+.    pl +1v
 .    sp 1v
+.  \}
 .  ev
 .  ns
 ..
@@ -1017,14 +1021,12 @@
 .  \" footer.  So we fudge the page length to make sure that the last
 .  \" page is never ejected until we want it to be.
 .
-.  if \n[cR] \
-.    pl +3v
 .  fl
-.
 .  doc-check-depth
 .
 .  if \n[cR] \{\
-.    sp 3v
+.    pl +1v
+.    sp 1v
 .    tl 
\*[doc-caption-font]\*[doc-operating-system]\f[]\*[doc-caption-font2]\*[doc-date-string]\f[]\*[doc-caption-font]\*[doc-operating-system]\f[]
 .    \" suppress empty lines after the footer
 .    pl \n[nl]u

> Before we (temporarily) depart historical matters, is it your claim that
> mdoc/BSD nroff innovated "continuous rendering mode"?  I ask simply
> because I'm curious.

I remain curious about this.

Regards,
Branden

signature.asc
Description: PGP signature