>>>>> "Alexandre" == Alexandre Oliva <[EMAIL PROTECTED]> writes:
Alexandre> When --host is given, we should not just assume we've got a
Alexandre> cross compiler. Instead, we test the compiler we've got,
Alexandre> and decide whether it's a cross compiler or not.
This one, I don't get it clear. Why would someone set host != build
and not expect cross-compiling? What's wrong with
(1) cross_compiling = build != host
personally I prefer
(2) cross_compiling = --host
but for sake of compatibility, I understand why you want to delay the
debate. But what's wrong with (1)?
Alexandre> The only concessions for backward compatibility we're
Alexandre> making is that --host alone will cause a cross-compilation
Alexandre> situation to be auto-detected, even though $build will be
Alexandre> set to host, not to `config.guess'.
Hm, so you prefer (2) because it is more like what we did in CVS?
Are users really used to use --host alone? I seemed to have
understood most of the time three arguments (out of four) were given.
I fear the `maybe'.
I feel a bit stupid: I have the impression you continuously explain
why we need `maybe', and I start to understand it must be related to
the impossibility to guess reliably a fake argument to --build. But I
can't make the connection :(
Alexandre> There's no way for configure to correctly guess the build
Alexandre> platform, in general, when it's told to use a cross
Alexandre> compiler.
I don't understand this too well. Are you saying that since people
are used to CC=sun4-cc, config.guess will actually config.guess sun4
(well, some bizarre mixture composed of build_uname, and host_cc)?
Akim
