nanogui: Cross-compilation of microwindows
Subject:
Re: [nanogui] Cross-compilation of microwindows
From:
Gil Glass ####@####.####
Date:
22 Aug 2005 21:57:13 +0100
Message-Id: <OFBBEB9ED0.63DF483B-ON85257065.0072A0AC-85257065.00730DEE@acterna.com>
Wow, that's pretty ambitious there, Peter. I'm impressed!
I did try working with AF_INET sockets on Nano-X, so that I could display
on my host PC while executing on my target, but I couldn't get it to work
and I don't remember why. It's been a few weeks and I've slept since
then. ;-) It sounds like you might have a better handle on it.
And while we're on the subject, I have a question that maybe you or one of
the more experienced Nano-Xers around here can answer. I thought that
"Microwindows" was just the old name of Nano-X and that the change came
about because Bill Gates had a hissy fit. Now it seems, based on your
algorithm below and some other things that I've read in the code and the
documentation, that Microwindows is another entity that is separate, but
related to, Nano-X. What's the deal?
Cheers,
Gil Glass
Telecom Field Services
JDSU
Germantown, MD, USA
+1-240-404-2551
Peter Barada ####@####.####
08/22/2005 04:42 PM
To
####@####.####
cc
####@####.####
Subject
Re: [nanogui] Cross-compilation of microwindows
>Based on my limited experience with cross-compiling, it looks like what
>you are doing should work.
Yeah, that change seems to work for me.
>With regard to the X11 headers, I'm assuming that your target is running
X
>Window and you're not using NXLib. If that's the case, then you should
be
>able to use the X11 header files from your i686. This is what I did when
>building FLTK (a widget library) and linking against NXLib.
I'm trying a pretty bizarre method. I have:
On the target:
Fluid application making X11 calls
--> nxlib converting X11 calls into microwindows calls
--> microwindows converting microwindows calls into nano-X calls
--> nanox client code pushing nano-x calls over a socket
On the host:
Nano-X server litening to the socket
--> converting nano-X calls to X11.
--> X11 server displaying results on host display.
This is so I can test cross-compilation on the target without having a
complete X11 implementation(cross-compiling X looks like a giant
pain; anyone have a cheat sheet?), but using the X11 on the host to
see the results. Then after I see that working, I'll implement an LCD
driver for microwindows on the target and use that when I have working
hardware.
To compile the Fluid app, I had to steal the host X11 headers and put
them into the toolchain(hopefully that won't bite me, or will it?).
I'm trying to convince nano-x to not used the named socket since that
seems to only work for sockets on the target, and instead am trying to
us the AF_INET method in src/nano-X/client.c and srvnet so I can
specify NXDISPLAY.
Does anyone see any glaring holes in my approach?
--
Peter Barada
####@####.####