Mutt 1.3.28 + ncurses 5.2 + xterm = blank screen

[Pavel Roskin]

Hello!

I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.

It turns out that the default configuration uses ncurses-5.2. I tried
configuring Mutt with slang-1.4.4. It works fine.

My guess is that Mutt does something with page switching that the newer
versions of ncurses don't like. I believe it's a major bug that will be a
frequently asked question as soon as mutt-1.4 is released.

I checked the lists to see if it's not already a frequently asked
question. There is a very similar description of the problem, except that
it's not specific about software versions:

http://marc.theaimsgroup.com/?l=mutt-users&m=101062129425997&w=2

[Thomas Dickey]

Post by Pavel Roskin
Hello!
I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.

what $TERM value?

Post by Pavel Roskin
It turns out that the default configuration uses ncurses-5.2. I tried
configuring Mutt with slang-1.4.4. It works fine.
My guess is that Mutt does something with page switching that the newer
versions of ncurses don't like. I believe it's a major bug that will be a
frequently asked question as soon as mutt-1.4 is released.
I checked the lists to see if it's not already a frequently asked
question. There is a very similar description of the problem, except that
http://marc.theaimsgroup.com/?l=mutt-users&m=101062129425997&w=2

That doesn't sound the same. The closest I can recall from this list is
one that says the pager comes up blank - both ncurses and slang.

[Pavel Roskin]

Hi, Thomas!

Post by Thomas Dickey

Post by Pavel Roskin
I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.

what $TERM value?

"xterm" both under xterm and rxvt. I forgot to mention that I was running
xterm from rxvt. I have found that if I run xterm from the window manager
the problem goes away!

When I run rxvt, it sets the following environment variables beginning
with "COLOR":

COLORFGBG=default;0
COLORTERM=rxvt

Both xterm and rxvt are using black background. From .Xdefaults:

XTerm*background: black
XTerm*foreground: gray85

Unsetting COLORFGBG fixes the problem.

My interpretation is that mutt uses black text on black background.
Probably ncurses interprets "default" in COLORFGBG as "black" whereas
S-Lang uses the foreground from the X resources.

Shouldn't ncurses ignore COLORFGBG if it has unsupported keywords (let's
move this discussion elsewhere if you want to continue).

Not exactly mutt problem, but may be useful thing to know if other people
ask.

[Thomas E. Dickey]

Post by Pavel Roskin
Hi, Thomas!

Post by Thomas Dickey

Post by Pavel Roskin
I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.

what $TERM value?

"xterm" both under xterm and rxvt. I forgot to mention that I was running
xterm from rxvt. I have found that if I run xterm from the window manager
the problem goes away!
When I run rxvt, it sets the following environment variables beginning
COLORFGBG=default;0
COLORTERM=rxvt
XTerm*background: black
XTerm*foreground: gray85
Unsetting COLORFGBG fixes the problem.

that's a bug that I fixed in September. The problem was that when I coded
the $COLORFGBG logic (which btw is under-documented in rxvt - you have to
read the C code to see it), it didn't occur to me that its format might
change. It happens that the format depends on whether xpm is linked in -
2 or 3 fields. The background color is the last field.

(Since it's under-documented, it's also possible that in the future
anything that relies upon that format will be broken ;-).

Post by Pavel Roskin
My interpretation is that mutt uses black text on black background.
Probably ncurses interprets "default" in COLORFGBG as "black" whereas
S-Lang uses the foreground from the X resources.
Shouldn't ncurses ignore COLORFGBG if it has unsupported keywords (let's
move this discussion elsewhere if you want to continue).

$COLORFGBG is marked as an experimental feature. I've gotten 2-3 reports
of this particular problem - but only months after I stumbled on it
myself. Apparently one or more of the rpm's last year turned that feature
on, though it was in the code almost a year.

Post by Pavel Roskin
Not exactly mutt problem, but may be useful thing to know if other people
ask.

[Jeremy Blosser]

Post by Thomas E. Dickey

Post by Pavel Roskin

Post by Pavel Roskin
I've compiled mutt-1.3.28i in the default configuration on RedHat
Linux 7.2 (i386) with all updates. If I run it in xterm (from
XFree86-4.1.0) or in rxvt-2.7.6, it shows a blank screen. I can
quit by pressing Ctrl-C and Enter. The same executable runs on the
Linux console just fine.

...

Post by Thomas E. Dickey

Post by Pavel Roskin
Unsetting COLORFGBG fixes the problem.

that's a bug that I fixed in September. The problem was that when I coded
the $COLORFGBG logic (which btw is under-documented in rxvt - you have to
read the C code to see it), it didn't occur to me that its format might
change. It happens that the format depends on whether xpm is linked in -
2 or 3 fields. The background color is the last field.

Indeed; Pavel, please see http://bugs.guug.de/db/10/1011.html for this bug
and the resolution, and ignore the mails from Cindy.

Post by Thomas E. Dickey
$COLORFGBG is marked as an experimental feature. I've gotten 2-3 reports
of this particular problem - but only months after I stumbled on it
myself. Apparently one or more of the rpm's last year turned that feature
on, though it was in the code almost a year.

Well, I think it was more the other bug where it would get turned on if
other development features like hard-tabs were turned on. It was
apparently a combination of these two.

[Jeremy Blosser]

Post by Jeremy Blosser
Well, I think it was more the other bug where it would get turned on if
other development features like hard-tabs were turned on. It was
apparently a combination of these two.

Sorry, I mean a combination of the colorfgbg bug, and the bug where
colorfgbg was enabled when unrelated dev features were enabled.

[Thomas Dickey]

Post by Jeremy Blosser

Post by Thomas E. Dickey
$COLORFGBG is marked as an experimental feature. I've gotten 2-3 reports
of this particular problem - but only months after I stumbled on it
myself. Apparently one or more of the rpm's last year turned that feature
on, though it was in the code almost a year.

Well, I think it was more the other bug where it would get turned on if
other development features like hard-tabs were turned on. It was
apparently a combination of these two.

I suppose so - though both features are in the same category (not part of
the default options):

Development Code:
--with-develop enable all development options
--enable-colorfgbg compile with $COLORFGBG code
--enable-hard-tabs compile with hard-tabs code
--disable-root-environ limit environment when running as root
--enable-xmc-glitch compile with limited support for xmc

(noting now that colorfgbg is "development" rather than "experimental",
I guess my reasoning at the time was that it was a small feature that
wouldn't break much ;-).

[Thomas E. Dickey]

Post by Pavel Roskin
Hi, Thomas!

Post by Pavel Roskin
I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.

the term 'blank screen' was misleading (to me). "Totally black" might
have jogged my memory regarding $COLORFGBG (though the unrelated report
of the pager which does not display anything was what I was thinking
about).

It's not a mutt problem, though as you noted. When I first saw it (early
September or late August), it was from running the ncurses test program.

[Ken Wahl]

Post by Pavel Roskin
Hello!
I've compiled mutt-1.3.28i in the default configuration on RedHat Linux
7.2 (i386) with all updates. If I run it in xterm (from XFree86-4.1.0) or
in rxvt-2.7.6, it shows a blank screen. I can quit by pressing Ctrl-C and
Enter. The same executable runs on the Linux console just fine.
It turns out that the default configuration uses ncurses-5.2. I tried
configuring Mutt with slang-1.4.4. It works fine.
My guess is that Mutt does something with page switching that the newer
versions of ncurses don't like. I believe it's a major bug that will be a
frequently asked question as soon as mutt-1.4 is released.
I checked the lists to see if it's not already a frequently asked
question. There is a very similar description of the problem, except that
http://marc.theaimsgroup.com/?l=mutt-users&m=101062129425997&w=2

I compiled 1.3.28 on RH 7.2 with ncurses-5.2.12 and XFree86-4.1.0-3 and
mutt works fine in xterm, gnome-terminal and Konsole. My $TERM is set
to xterm. My compile options are:

Mutt 1.3.28i (2002-03-13)
Copyright (C) 1996-2001 Michael R. Elkins and others.
Mutt comes with ABSOLUTELY NO WARRANTY; for details type `mutt -vv'.
Mutt is free software, and you are welcome to redistribute it
under certain conditions; type `mutt -vv' for details.

System: Linux 2.4.9-31 (i586) [using ncurses 5.2]
Compile options:
-DOMAIN
-DEBUG
-HOMESPOOL +USE_SETGID +USE_DOTLOCK +DL_STANDALONE
+USE_FCNTL -USE_FLOCK
-USE_POP -USE_IMAP -USE_GSS -USE_SSL -USE_SASL
+HAVE_REGCOMP -USE_GNU_REGEX
+HAVE_COLOR +HAVE_START_COLOR +HAVE_TYPEAHEAD +HAVE_BKGDSET
+HAVE_CURS_SET +HAVE_META +HAVE_RESIZETERM
+HAVE_PGP -BUFFY_SIZE -EXACT_ADDRESS -SUN_ATTACHMENT
+ENABLE_NLS -LOCALES_HACK +HAVE_WC_FUNCS +HAVE_LANGINFO_CODESET +HAVE_LANGINFO_YESEXPR
+HAVE_ICONV -ICONV_NONTRANS +HAVE_GETSID -HAVE_GETADDRINFO
ISPELL="/usr/bin/ispell"
SENDMAIL="/usr/sbin/sendmail"
MAILPATH="/var/spool/mail"
PKGDATADIR="/usr/share/mutt"
SYSCONFDIR="/etc"
EXECSHELL="/bin/sh"
MIXMASTER="/home/ken/Mix"