On Wed, Jan 28, 2026 at 08:21:41PM -0500, Thomas Dickey wrote:
> On Wed, Jan 28, 2026 at 06:29:43PM -0600, Chris Adams wrote:
> > I'm not sure if this is so much a bug report as it is a request for
> > which is the best option with (near-)current software.
> > 
> > When I upgraded to Fedora 43, I got screen-5.0.1 and ncurses-6.5.  Per
> > screen bug https://savannah.gnu.org/bugs/?67436 screen 5.0 implements
> > italics now, and the terminfo entry for "screen" has the italics
> > sequence for standout mode, so things wanting standout get italics
> > instead.
> > 
> > I use real xterm (401 on Fedora 43), and for ages I've used TERM=xterm,
> > which means when I run screen I get TERM=screen (this was in part
> > because for work I'd often be logging into network gear with very
> > limited terminal info, but they had xterm and screen, although that's
> > not really an issue for me anymore).  That's where I see the problem
> > with standout vs. italics.
> > 
> > Trying combinations, I see:
> > 
> > "TERM=xterm" (or xterm-new or xterm-256color which is Fedora xterm default):
> > - smso gets standout
> > - sitm gets italics
> > 
> > "TERM=xterm screen" gets TERM=screen:
> > - smso gets italics
> > - sitm does nothing
> > 
> > "TERM=xterm-new screen" gets TERM=screen.xterm-new:
> > - smso gets standout
> > - sitm does nothing
> > 
> > "TERM=xterm-256color screen" gets TERM=screen.xterm-256color:
> > - smso gets standout
> > - sitm does nothing
> 
> the above looks consistent
>  
> > "screen -T screen5" gets TERM=screen5:
> > - smso gets standout
> > - sitm gets italics
> 
> hmm - thanks for the reminder - I see this in my Fedora rawhide:
> 
> screen-5.0.1-6.fc44.x86_64
> 
> I haven't looked recently at screen version 5
> (which was unreleased for several years).
> 
> infocmp has
> 
> comparing screen to screen5.
>     comparing booleans.
>     comparing numbers.
>     comparing strings.
>       ritm: NULL, '\E[23m'.
>       rmso: '\E[23m', '\E[27m'.
>       sgr: 
> '\E[0%?%p6%t;1%;%?%p1%t;3%;%?%p2%t;4%;%?%p3%t;7%;%?%p4%t;5%;%?%p5%t;2%;m%?%p9%t\016%e\017%;',
>  
> '\E[0%?%p6%t;1%;%?%p2%t;4%;%?%p1%p3%|%t;7%;%?%p4%t;5%;%?%p5%t;2%;m%?%p9%t\016%e\017%;'.
>       sitm: NULL, '\E[3m'.
>       smso: '\E[3m', '\E[7m'.
>  
> ...which as I mentioned, I haven't retested for quite a while.
> 
> > So should I just use screen5 for screen (which then is a separate issue
> 
> yes... until "screen version 5" actually becomes prevalent.
> Debian testing still has 4.9; OpenSUSE also (from last week anyway).
> Other systems will take longer.
> 
> (we're talking about a change to a 30-year-old bug which doesn't improve much)
> 
> > because I also use RHEL systems that don't have screen5, sigh)?  Do the
> > screen.xterm* variants add more than the screen5 entry?

The screen.xterm* variants don't add anything.  Basically "screen" detects
something like screen.$TERM, and uses that.  Its reuse/extension/whatever
isn't perfect, so my customized entries tweak things in the screen.xterm*
to correspond with test results.

As a result, none of those screen.xterm* have italics as done in screen5.

I expect that I'll have to add something like screen4.* entries to act as
a bridge, and then at some point change this

screen|VT 100/ANSI X3.64 virtual terminal,
        use=screen4,

to

screen|VT 100/ANSI X3.64 virtual terminal,
        use=screen5,

If you have only one machine, and it doesn't have a $HOME/.terminfo,
then compiling that chunk with tic should allow you to use TERM=screen
from that machine -- but I expect it to not be that easy.

(I'd have left the 30-year-old bug alone and added some new/unused code
to get italics, but then there are lots of hard-coders who wouldn't agree)
 
> > Thanks for tips.
> > 
> > Side note: I do enjoy seeing the history embedded in the terminfo
> > file... I wrote the "ADM 31 DIP Switches" documentation chunk back in
> > 1993, emailed to Richard Shuford who was then the keeper of terminal
> > history.
> > https://www.vt100.net/shuford/terminal/lear_siegler_adm_setup.txt
> > 
> > -- 
> > Chris Adams <[email protected]>
> > 
> > 
> 
> -- 
> Thomas E. Dickey <[email protected]>
> https://invisible-island.net



-- 
Thomas E. Dickey <[email protected]>
https://invisible-island.net

Attachment: signature.asc
Description: PGP signature

Reply via email to