Re: bug in start_color() affecting extended_pair_content() in version 6.1

[Thomas Dickey]

On Mon, May 20, 2019 at 02:47:56PM -0700, Jeffrey Kintscher wrote:
> I initially tested with ncurses 6.1 installed by brew. I switched to the
> sourcecode distribution for more detailed debugging.

I build a copy of your test program, turning on the debug-trace and
collecting the output logs, etc. in the attached, but don't see a
problem.  I made sure to use the terminfo database which was built
with ncurses 6.1:

Archive:  /tmp/testcolors.zip
  Length      Date    Time    Name
---------  ---------- -----   ----
     8377  2019-05-21 19:19   trace
    42613  2019-05-21 19:18   Makefile
      431  2019-05-21 19:15   foo.c
      130  2019-05-21 19:12   testit
      267  2019-05-21 19:06   doit
    53808  2019-05-21 19:08   config.status
   105450  2019-05-21 19:08   config.log
    85617  2019-05-21 19:10   make.log
    18283  2019-05-21 19:08   run.log
      120  2019-05-21 19:20   foo.log
---------                     -------
   315096                     10 files

recap -

        + "doit" is the script with configure options,
        + I edited the makefile to get the test-program compiled quickly.
        + "testit" is the environment tweaks I used to eliminate
          other paths/databases.
        + The "trace" file is what I normally use for testing, but
        + foo.log is the stdout from your example.

-- 
Thomas E. Dickey <address@hidden>
https://invisible-island.net
ftp://ftp.invisible-island.net

testcolors.zip
Description: Zip archive

signature.asc
Description: Digital signature