Alternatives to gcc (was Re: gcc 4.3: when will it becomestandard compiler?)

[bf]

--- On Sat, 1/31/09, Pedro F. Giffuni <giffunip at tutopia.com> wrote:

> From: Pedro F. Giffuni <giffunip at tutopia.com>
> Subject: Re: Alternatives to gcc (was Re: gcc 4.3: when will it becomestandard compiler?)
> To: "Mark Linimon" <linimon at lonesome.com>, bf2006a at yahoo.com
> Cc: freebsd-current at freebsd.org, "Sean Cavanaugh" <Millenia2000 at hotmail.com>
> Date: Saturday, January 31, 2009, 6:22 PM
> --- On Sat, 1/31/09, Mark Linimon
> <linimon at lonesome.com> wrote:
> 
> On Sat, Jan 31, 2009 at 01:08:54PM -0800, Pedro F. Giffuni
> wrote:
> > > The effort didn't go far enough. Why
> haven't we removed GNU readline ?
> >
> > Probably either because someone hasn't written a
> BSD-licensed one, or
> > someone hasn't done the work to test-compile src
> and ports on all the
> > appropriate architectures.
> 
> Wrong on both:
> 
> - libedit has a readline compatibility mode that has
> replaced GNU readline in the other BSDs.
> - If you look in the archives you will find patches.
> 
> If there really was any effort to remove GPL'd stuff
> from the tree it missed this big time: GNU readline is a
> library under the GPL (not LGPL), it should be dead long
> ago. 

The license is _a_ consideration, but not the _only_ consideration
for including some useful code.  I don't know much about the
readline case, but it was my impression that libedit was considered
and then rejected, partly because at the time it wasn't able to do
everything that readline could, and no one was willing to put forth
the effort to make it do so, and then test it, as Mark said.  I don't 
think he's wrong.  Maybe it's much better now, and it could be a drop-in
replacement, or something close to it, but it still has to be ported
to FreeBSD and tested, right?  If you've done it, or you want to do it,
where is your patchset?


> 
> Now, after running the lang/gcc43 "make check" on
> my amd64 I don't really understand why we would be in a
> hurry to update to it.
> 

Because it has a large number of bugfixes and improvements over gcc 4.2.x.
Read the changelogs for examples.  That doesn't mean that vanilla gcc 4.3.x, or the port, don't need some modifications before they could become
the base compiler. Again, I'm sure gerald@ would appreciate patches.


> Pedro.