simulavr-devel
[Top][All Lists]
Advanced

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

Re: [Simulavr-devel] simulavr on 'bsd


From: ken restivo
Subject: Re: [Simulavr-devel] simulavr on 'bsd
Date: Tue, 8 Jan 2002 07:51:35 -0800
User-agent: Mutt/1.3.25i

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, Jan 07, 2002 at 11:29:21PM -0700, Theodore A. Roth wrote:
> On Mon, 7 Jan 2002, address@hidden wrote:
> 
> :)just tried to compile simulavr on netbsd 1.5.1 again. 
> :)
> :)it does pretty well on portability. but a couple things appear broken:
> :)
> :)1) ../simulavr/src/disp/disp.c:541: undefined reference to `resizeterm'
> :)    this is an ncurses thing, and netbsd doesn't grok ncurses that well.
> :)    it's theoretically possible to install ncurses, but i've never been 
> able to.
> :)    i'd suggest replacing it with some curses least-common-denominator, 
> :)    if such a thing exists.
> 
> I've looked into this and it doesn't look like there's a way to implement
> terminal resizing without using resizeterm(). I will admit that it could
> be done if you worked with the internals of the window struct but I ain't
> gonna do that since it would be even less portable than using resizeterm.

i've done it with SIGWINCH handlers, and then getting the new window size 
coordinates through curses, but it's a pain in the ass. i agree it's not worth 
it.

> 
> I can do a little hack to make it compile, but I don't think it's really
> worth it. If a system doesn't have the resizeterm(), you just have to
> disable building of disp. If the user really needs a display, there's
> always the option of writing a custom app for that system. That's why disp
> is a separate program and there's a comm protocol which is documented in 
> the user manual. I really didn't intend disp to be an end all display 
> tool, but it is useful for debugging. I wonder if slang or newt is more 
> portable? Wouldn't hurt to have another tool as a add-on in the future. I 
> would like to keep the lib dependency to a minimum though.
> 
> Ken, you might be interested in this:
> 
>   http://www.netbsd.org/Changes/changes-1.6.html#curses(3)

oh, cool. thanks!  i'll update to netbsd 1.6 and see what happens then.

i'd suggest something like AC_CHECK_FUNC(resizeterm, , AC_MSG_ERROR("your 
curses implementation is too old. please update."))

> 
> :)
> :)2)  on_exit again. i'd hacked main.c to use atexit instead as a workaround.
> :)    on_exit appears to be a gnu thing, neither netbsd nor freebsd 4.3 have 
> it.
> :)    atexit semantics are different too, there's no way to pass a void * arg.
> 
> Done and checked in. Try it and see if it breaks.
> 
> I'll look into these tomorrow. No gaurantees though.
> 
> :)3)  texi is a NOP on netbsd. probably makes sense to do an AC_PROGRAM for 
> it,
> :)    and fail to ./configure without it, or allow an option to build w/o it.
> :)
> :)4)  the use of times() in utils.c:get_program_time is broken on netbsd.
> :)    compiles fine, but causes segfaults. urk.
> :)    the netbsd docs recommend using gettimeofday() instead.
> 

some time ago, i sent you a quick patch that fixed this . it wasn't perfect, 
but it worked for me, and might just need some cleaning up. i can send it again 
if it is of any use.

thanks again!

- -ken
- -- 
- ------------------
One world. Many gods. Plenty for everyone.


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE8OxWGe8HF+6xeOIcRApSHAJ9kWPR83K09bRNBWrwGLot/+tdGLACgtfx/
0IWAk7eyqnjeXxTKd6dchGw=
=ccqz
-----END PGP SIGNATURE-----



reply via email to

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