Re: [groff] 02/04: grotty.1.man: Make editorial fixes.

[G. Branden Robinson]

[coming back to this after a while]

At 2019-06-30T17:38:16+0200, Ingo Schwarze wrote:
> Thanks; overall, these seem to be impovements.

Thanks, Ingo!

> There is one detail that should really be removed though, i think:
[...]
> The -R option is already scary.  The -r option is even worse,
> introducing substantial security risks *and* usually resulting in
> broken output, too.  We should absolutely not recommend it.  Mentioning
> it causes the totally misleading impression that it might provide
> any benefit at all in the groff context.
> 
> People who are reckless enough to consider using it for other reasons
> can see easily enough in the less(1) manual that it implies -R.

I'm updating grotty(1) to remove mention of less's -r.  I don't share
your analysis of the problem; I think it's the terminal emulator's
responsibility to expose controls to the user to enable "dangerous"
features if it is going to implement them at all.  For example, xterm
has menu items for "Allow Font Ops", "Allow Termcap Ops", and "Allow
Window Ops", but they all default off.  By contrast, its menu items for
"Allow Color Ops", "Allow Mouse Ops", and "Allow Title Ops" default on
(on my Debian Buster system).

However, I am disappointed in less's man page for not documenting these
issues in greater detail, and I'm a bit exasperated at its terribly
inaccurate description of ISO 6429/ECMA-48 SGR escape sequences.  It
calls all of them "ANSI color escapes" whether they have to do with
changing the device's color attributes or not, and it ignores the CSI
alternative form of the escapes, which compliant devices have to
support.

So I decided to say as little about "less" as I can get away with
without making life harder for groff users, and that happens to coincide
with your preferences as I understand them.

Regards,
Branden

signature.asc
Description: PGP signature