Re: Help wanted with bkgd() rookie mistake

[G. Branden Robinson]

At 2025-01-05T13:33:02-0500, Thomas Dickey wrote:
> On Sun, Jan 05, 2025 at 12:17:15PM -0600, G. Branden Robinson wrote:
> > At 2025-01-05T13:06:31-0500, Bill Gray wrote:
> > > On 1/5/25 11:23, G. Branden Robinson wrote:
> > > > In both, I expect the background character to be set to an
> > > > underlined @ sign.
> > > 
> > > I see the background as underlined @ signs with both programs.
> > > I'm using (default,  system-installed,  now getting to be elderly)
> > > ncurses 6.2.20200212.  So perhaps a behavior change between then
> > > and now?
> > 
> > Hmm.  That may be.  I'm on the bleeding edge with 6.5.20250104.
> 
> Both programs look fine to me, but I don't replace the system
> (n)curses, but rather use the test-packages whose build-scripts are in
> the "packages" directory.

I'm not replacing the system ncurses, but installing to
"$HOME/ncurses-HEAD/".

> bkgd:
>       linux-vdso.so.1 (0x00007ffc5b99a000)
>       libncursesw6.so.6 => /lib/x86_64-linux-gnu/libncursesw6.so.6 
> (0x00007f43130c7000)
>       libtinfow6.so.6 => /lib/x86_64-linux-gnu/libtinfow6.so.6 
> (0x00007f431308c000)
>       libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f4312eab000)
>       /lib64/ld-linux-x86-64.so.2 (0x00007f431313b000)

Hmm.  ldd is a red herring for me.  When I reinstall ncurses "HEAD", my
ldd output doesn't change.

(I seem to remember reading a cautionary note that ldd itself could be a
vector for security attacks due to the fact that it tries to dynamically
resolve which ELF objects to load.  But I seem to be getting less
dynamism than I expected.)

LD_LIBRARY_PATH seems to work in general, though.  I know because when I
let my ordinary environment, which preferentially looks in
$HOME/ncurses-HEAD, prevail, neomutt core dumps with modern ncurses.  I
have to defeat my LD_LIBRARY_PATH just for that program.

$ make
cc -D_DEFAULT_SOURCE -D_XOPEN_SOURCE=600 bkgd.c -o bkgd -lncurses -ltinfo
cc -D_DEFAULT_SOURCE -D_XOPEN_SOURCE=600 
-I/home/branden/ncurses-HEAD/include/ncursesw 
-I/home/branden/ncurses-HEAD/include bkgrnd.c -o bkgrnd 
-L/home/branden/ncurses-HEAD/lib -lncursesw -ltinfow -ldb

Any suggestions for where to look for the cause of the trouble?

> and defining NCURSES_WIDECHAR when compiling.

Hmm, I use "./configure --enable-widec".  Is that different?

I'm attaching my scripts for configuring and rebuilding ncurses; let me
know if anything looks stupid.

While I'm at it I'll include my Makefile for my examples.  Very simple.
Very crude.

> I have seen comments that ncurses-examples aren't simple...

People may have noticed that I contribute a lot of patches to the man
pages, so often I need to write something micro-targeted at a specific
API call.  Often, I want to see what the library does with invalid
inputs.  Your test programs and examples have value, but they're more
like integration tests and demos.

Regards,
Branden

remake-ncurses.sh
Description: remake-ncurses.sh

config-ncurses
Description: config-ncurses

Makefile
Description: Makefile

signature.asc
Description: PGP signature