Subject: libcurses, lynx & groff
To: None <current-users@NetBSD.ORG>
From: J.D.Coleman <J.D.Coleman@newcastle.ac.uk>
List: current-users
Date: 02/18/1997 15:48:09
As the curses library only supports standout and no other attributes and I
wanted to run lynx (without ncurses), I decided to add attribute support to
libcurses.  I've done that but encountered a few funnies in the original
libcurses on the way ...

Libcurses didn't a) compile and b) run with -DDEBUG defined.  The fixes
were trivial though.  Presumably no-one has done this for a long time!

Using lynx (2.6) - when I download an image, it finishes curses (calling
endwin()), launches xv on the image and then carries on using curses.  This
results in the screen display not being refreshed (as well as sending the
'te' escape to the window - although this doesn't seem to hurt).  Presumably
it should re-enable curses after launching xv?  A comment says call initscr()
'once and once only!' apart from under VMS, but it seems OK to call it under
NetBSD - and it makes the screen refresh too :)  Presumably lynx should
explicitly re-enable curses here?  Ideas, hints, the definite article?

The other funny is the PSD.doc curses documentation.  It uses numbered
footnotes using the \** -me macro.  According to my -me docs (SunOS), this
should generate an auto-incrementing footnote number but groff generates
'.nr $d 1 1' instead.  Bug in the documentation, -me macros or groff?

Thanks,

J

PS.  Anyone want to test a libcurses with underline, bold, reverse, etc in?
I can only test it against xterm, vt100 and vt52, which is a bit limited, but
it's been running fine for me for about a week.