Date: Mon, 21 Oct 2002 10:40:56 +0300 From: Ruslan Ermilov <ru@FreeBSD.ORG> To: Terry Lambert <tlambert2@mindspring.com> Cc: "Andrey A. Chernov" <ache@nagual.pp.ru>, "M. Warner Losh" <imp@bsdimp.com>, arch@FreeBSD.ORG Subject: Re: color, again, in grotty Message-ID: <20021021074056.GE14584@sunbay.com> In-Reply-To: <3DB3A962.9F26FCA0@mindspring.com> References: <20021017.101833.110719994.imp@bsdimp.com> <20021018095026.GA3386@sunbay.com> <20021018.094801.123456703.imp@bsdimp.com> <3DB06A8B.E40B3004@mindspring.com> <20021018201919.GA15100@nagual.pp.ru> <20021021065156.GB14584@sunbay.com> <3DB3A962.9F26FCA0@mindspring.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--f61P+fpdnY2FZS1u
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Mon, Oct 21, 2002 at 12:14:42AM -0700, Terry Lambert wrote:
> Ruslan Ermilov wrote:
> > > > > : So we must either fix syscons(4) or to somehow tell grotty(1) to
> > > > > : emit old sequences to print bold and italic characters.
> > > >
> > > > Fixing Grotty to use the termcap ("TermCap" -- Terminal Capabilitie=
s)
> > > > database is the correct approach, in almost all cases.
> > >
> > > I agree. Grotty should get termcap color capabilities using TERM env.
> > > variable, if output isatty(). That is termcap purpose.
> >=20
> > It is almost never isatty(), because the most typical scenario is
> > to pass the output through a pipe to ${PAGER}, or to a compressor.
>=20
> Pagers can't tell the difference between an escape sequence that
> results in characters which are not displayable without the
> escape sequence (and therefore take a character cell on the screen)
> and those which manipulate display attributes (and therefore do not
> take any screen space) from those which do cursor positioning in
> place of "CR/LF" (e.g. "delete line" at the top of the page is
> equivalent to scrolling) vs. those which do nothing apparent.
>=20
Look at the -R option of less(1). The point I was trying to make is
that the use of isatty(3) is no help here, because even if grotty(1)
took the libtermcap(3) approach, the output would almost never be a
TTY.
> Therefore, it's unreasonable to run processed data through a
> pager, if the results of that processing include terminal
> implementation dependent escape sequences..
>=20
> The way this is normally dealt with is to use "_^HX" for underscore,
> "X^HX" for bold, etc., utilizing "overstrike", and have the pager
> convert it based on the pager's knowledge of the terminal display
> sequences from the termcap.
>=20
This only covers _attributes_, but not color sequences, which are
the Subject: here. While we can still use an old scheme to produce
device-independent sequences for attributes, we lose the color
support completely: FreeBSD's man(1) will be black-n-white forever.
> Note that, so long as ASCII <BS> (0x08) is backspace, and moves
> the cursor one to the left, in the absence of additional terminal
> interpretation, the resulting data will be displayed as readable
> text, minus the attributes (e.g. even if the second character
> replaces, rather than overstriking the first, the text will end
> up being the same, without any display attributes at all).
>=20
>=20
> I guess a lot of kids these days have grown up without having to
> use real tty or terminal hardware, so they don't understand all
> the work that has already gone into finding the correct way to
> deal with these issues...
>=20
Definitely not me. I have been involved into the development of
the at386 compatible terminal back in early 90's. I also worked
on Esprit 400 terminals which were vt100 based.
So far nobody told me what would be the correct alternative way to
support colors in grotty(1). I tried the isatty() + libtermcap
approach first, but it died horribly because the output is almost
never a TTY. It became apparent to me that grotty(1) cannot use
libtermcap, because it never knows what TERM it will be displayed
on -- quite often the grotty(1) output is saved for future reference
(e.g., in the form of catpages). Such an alternative would be to
write a filter (similar to or extending the functionality of ul(1))
that would be capable of parsing the ANSI SGR attribute/color
sequences and translate them to the right sequences for a given
output device, or to strip them off completely.
Cheers,
--=20
Ruslan Ermilov Sysadmin and DBA,
ru@sunbay.com Sunbay Software AG,
ru@FreeBSD.org FreeBSD committer,
+380.652.512.251 Simferopol, Ukraine
http://www.FreeBSD.org The Power To Serve
http://www.oracle.com Enabling The Information Age
--f61P+fpdnY2FZS1u
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (FreeBSD)
iD8DBQE9s6+IUkv4P6juNwoRAhFdAJ0dzOt13VcOpXOyQA49pHyP0cuNoQCfYzuX
Ad5Rf79O8bOLs52m6JRtasU=
=VQ0N
-----END PGP SIGNATURE-----
--f61P+fpdnY2FZS1u--
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021021074056.GE14584>