[Top][All Lists]

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

Re: A doubt about my participation in ncurses project.

From: daniel silva ferreira bruno
Subject: Re: A doubt about my participation in ncurses project.
Date: Sat, 9 Mar 2013 05:38:45 -0800 (PST)

I worked at these questions and I noticed that:

a) The ncurses port don't initialize in Console2 software because the initscr() function of the ncurses library fails.

b) I couldn't test it completely because any ncurses application can't to be initialized at Console2 and Console2 works only on MS Windows OS.

c) It depends on b).

d) I tested it. By
, I noticed that:
|- For the "C1 Controls and Latin-1 Supplement" (UTF-8 (hex.) from 0x80 to 0xFF)
   |- no errors.

Should I do more tests? How should I do them?

e) I didn't test them yet.

Best Regards

Daniel Bruno.

--- Em seg, 28/1/13, Thomas Dickey <address@hidden> escreveu:

De: Thomas Dickey <address@hidden>
Assunto: Re: A doubt about my participation in ncurses project.
Para: "daniel silva ferreira bruno" <address@hidden>
Cc: address@hidden
Data: Segunda-feira, 28 de Janeiro de 2013, 21:58

On Mon, Jan 28, 2013 at 03:24:56PM -0800, daniel silva ferreira bruno wrote:
> Hi, Thomas.
> I liked to know, please, how I can to help the ncurses project at MS Windows
> (or Mingw, or MS Visual Studio, if possible!) programming, if I can.

well, Erwin Waterlander reported several issues (and I've been working
on those, but have not investigated all).  Here are a few:

    a) the ncurses port works in a regular Windows console window,
       and in ConEmu - but does not initialize in Console2.

       It would be very useful to identify the source of the problem.
       PDCurses works in Console2, though it has its own share of
       problems :-)

    b) there are differences between those three (as well as with
       mintty) for the ability to display double-cell characters.

       My current understanding is that a console application cannot
       tell much about the font in use (aside from its name), and
       so cannot tell whether a given font uses double- or single-cells
       for some problematic line-drawing characters.  (It would be nice
       if I'm wrong, and someone shows how to do this :-)

    c) I started some experimental code to improve screen-painting
       by reducing the amount of output.  It doesn't handle the
       double-width example which Erwin gave me.  (Solving this
       may actually depend on (b) mentioned above).

    d) Input of Unicode (wide) characters is incompletely tested.

       As I did for output, some changes are probably needed.
       MinGW's wide-character support builds, but does not work
       well enough for use.

    e) Erwin also had comments about surrogate pairs, screen flicker.
       I'm unsure if these are addressed.

(as usual, my progress on these depends on time available after doing
normal maintenance...)

Thomas E. Dickey <address@hidden>

reply via email to

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