Re: contrib/groffer/roff2.1.man

[Ingo Schwarze]

Hi Branden,

G. Branden Robinson wrote on Sun, Apr 19, 2020 at 07:16:58AM -0400:

> commit 762fab3e454786cc5e666f6d6324f455470b8ea4
> Author: G. Branden Robinson <address@hidden>
> AuthorDate: Sun Apr 19 20:33:39 2020 +1000
> 
>     contrib/groffer/roff2.1.man: Fix register save.
>     
>     roff2.1.man generates several nearly-identical pages from the same
>     source*.  Use register name based on the name of the generated page, so
>     they don't collide.
>     
>     Also add bug reference to changelog item.
>     
>     * This seems pretty gratituitous to me; their interfaces don't differ at
>       all (no difference in accepted options, for example), and each page
>       makes reference to all the others.  I find that a dubious choice, and
>       would just have one page with a comma-delimited set of topics in the
>       "Name" section.

Indeed, good find, that's utter stupidity.
It would be nice to have that cleaned up.

Then again, the whole contrib/groffer directory is an abomination
that i suggest should be deleted outright, without any replacement.
It's fragile wrappers on top of wrappers on top of wrappers (remember
that even groff itself is already a wrapper!), of ridiculous
complexity, that provide no additional functionality whatsoever.

Less bloat is more.

It is a landmark symptom of the absense of software design when,
instead of making the program itself small, simple, and easy to
use, people write a wrapper that clearly doesn't achieve simplicity
either, neither of the code nor of the usage.

If there is no consensus to delete groffer outright, then it would
be good to merge all the almost empty roff2*(1) manual pages into
the groffer(1) manual page, given that even though these scripts
are not short, they do almost nothing except calliing groffer in
some very contorted way.

Or at least merge all the empty roff2*(1) manual pages into a single
page as a first step.  It will still be almost empty, but at least
it will be half the size of each of the current pages - because
right now, half of each of these pages consists of references to
the others, which are just as empty.

Yours,
  Ingo