[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Success cross-compiling ncurses-5.3-20030719 -- but can't install
From: |
Dan Kegel |
Subject: |
Re: Success cross-compiling ncurses-5.3-20030719 -- but can't install |
Date: |
Thu, 28 Aug 2003 19:39:54 -0700 |
User-agent: |
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030624 |
Thomas Dickey wrote:
ncurses-5.3 seems to be smart about not using strstream,
but this fails when you specify a cross-compiler with --build-cc ...
that part I'm curious about - didn't get a chance to follow up on it,
but don't know offhand why build-cc would interfere with the C++ compiler
checks. build-cc is supposed to be used only for the couple of build-time
executables that generate some source files under the ncurses directory.
Then I misunderstood it. Here's what "configure --help" says:
--with-build-cc=XXX if cross-compiling, the build C compiler (\$BUILD_CC)
I misread that as "the cross-compiler", believe it or not.
Not sure what you can do about users who can't read,
but maybe you could change the text to read
--with-build-cc=XXX C compiler used to generate temporary executables
used during build (\$BUILD_CC)
or something like that.
Is the output of 'tic' byte-order independent? If so, I could
compile ncurses a second time, for the build system, and use the
'tic' it produces to install the terminfo entries.
It is supposed to be byte-order independent, yes.
Cool!
If I understand you right, all we have to do is arrange
for the 'tic' used during build time to be built with $BUILD_CC.
Would that be an acceptable change?
In the meantime, I'll probably try some horrible kludge like
using my workstation's copy of tic; it's from a slightly older
release of ncurses, but maybe it'll work anyway.
- Dan
--
Dan Kegel
http://www.kegel.com
http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045