Date: Fri, 27 Jul 2001 20:12:15 +0300
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: "Mark E." <snowball3 AT bigfoot DOT com>
Message-Id: <8296-Fri27Jul2001201215+0300-eliz@is.elta.co.il>
X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.9
CC: djgpp-workers AT delorie DOT com
In-reply-to: <3B613FAE.29303.2CF007@localhost> (snowball3@bigfoot.com)
Subject: Re: termios cursor motion support
References: <3B604325 DOT 14771 DOT 32CE14 AT localhost> (snowball3 AT bigfoot DOT com) <3B613FAE DOT 29303 DOT 2CF007 AT localhost>
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> From: "Mark E." <snowball3 AT bigfoot DOT com>
> Date: Fri, 27 Jul 2001 10:17:18 -0400
> 
> > Shouldn't the escape sequences supported by this emulation be
> > available to termcap functions such as tgetstr via the same mechanism
> > you already implemented for extended keys?
> 
> If you mean add capabilities to djgpp's termcap entry, then yes they need to 
> be added so programs that use tgetstr, etc. will know about them.  

Yes, that's what I had in mind.

> > Not necessarily: BIOS tty functions support color.  (Let me know if
> > you want more details about this.)  So this is basically a design
> > decision.
> 
> Do you mean the int 0x10 functions ah=08 and 09 to read and write a character
> plus attribute at the current cursor position?

Yes.  There's also function 0Eh, but it only allows to specify a
foreground.  However, it's more useful, since it handles special
characters such as BS, CR, and LF.  The problem with the background
color can be solved by writing a blank with the background color (with
function 9) before writing the character itself.

> True. But there are more disadvantages I want to avoid. Using conio would 
> mean that a large amount of non-posix symbols would become involved in a 
> posix function. conio's output functions are geared toward null terminated 
> strings while the termios read and write functions specify a length. And 
> conio's getch clashes with the curses function getch.  

Yes, those are all good reasons not to use conio.