lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

lynx-dev lynx+ncurses / lynx+slang (was: ^L in popup corrupts screen)


From: Klaus Weide
Subject: lynx-dev lynx+ncurses / lynx+slang (was: ^L in popup corrupts screen)
Date: Sun, 31 Oct 1999 17:51:46 -0600 (CST)

On Sun, 31 Oct 1999, Ismael Cordeiro wrote:

> OK, Klaus, you won. I recompiled lynx2.8.3dev.13  with ncurses-5.0,

Another soul won over from the dark side :)

Just kidding.  I don't really think ncurses is better for everything.

> replicated the vt100 entry in my .termcap file, calling it "desterm", and set
> TERM to "desterm". Here are my first comments:

What's the origin of that vt100 entry (or the whole file)?

> - I got colors! It seems that vt100 capabilities, or some of them, are
> hardcoded in ncurses and that ncurses was not reading my .termcap file. Now
> it is reading the "desterm" capabilities, including the color ones.

To the best of my knowledge ncurses doesn't have anything like that
hardcoded at all.  It is slang that does that.

I think one *can* somehow build an ncurses lib that *does* have some
specific terminal info compiled in, but that should not happen by
default (if I am right that it is possible at all).

I am not familiar with all the different ways (different files) how
ncurses can get capability info (I just use terminfo not termcap).  I
assume there must be some file somewhere on your system where ncurses
found a different "vt100" entry and was using that.

> - For the first time ever with lynx compiled with ncurses I have the numeric
> keypad correctly working in NUMBERS_AS_ARROWS mode. Among other things,
> "begining of document" was 1 and "end of document" was 0 when I used lynx-nc
> (lynx+ncurses) and had TERM set to vt100. Now it's the corret 7 and 1 keys,
> respectively. All other keys also work correctly. Notice that the "desterm"
> entry in my .termcap is exactly the same as the "vt100" entry and that the
> TERMCAP variable is set to "/home/ismael/.termcap". Some more hardcoded
> things in ncurses?

I really don't think so.

Well, *lynx* has some escape sequences hardcoded in when using ncurses;
but nothing that would get triggered by just the name change.

> - Still about the numeric keypad, when I set TERM to "desterm" and use
> lynx-sl (lynx+slang) the keys on the numeric keypad don't work in
> NUMBERS_AS_ARROWS mode.

If lynx-sl is using that file at all - I assume you know that from previous
experience or can verify it be making some changes - then this may be the
same kind of problem Henry had.  If so it would be resolved with one of my
patches.

But it maybe something different - no telling without seeing what those
keys send and comparing it to your "desterm" entry.

> - I'm getting pop menus with boxes formed with "+", "|" and "-" instead of
> the ISO 8859-1 8-bit characters I have with lynx-sl.

Well that's one point were ncurses is "better".  Slang always makes
vt100-ish assumptions here, and there's no way to tell it otherwise
(last I checked) unless you don't use the relevant function at all.

> - Yes, if I open a popup menu and do a ^L the screen gets corrupted.
> 
> - I have "COLOR:4:green:black" [4 - underline - text emphasis (EM, I, B tags
> etc.)] but lynx-nc is displaying the tags in green _and_ underline instead of
> just green.
> 
> - No matter the color, bright or non-bright, I use for the foreground in
> "COLOR:1" [1 - bold - hyperlinks] lynx-nc displays it bright.

So those last two observations agree with mine, in the
   What's with the ncurses colors?
thread.

   Klaus


reply via email to

[Prev in Thread] Current Thread [Next in Thread]