[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Mingw-cross-env-list] Re: About porting of GCC and possible redunda
Re: [Mingw-cross-env-list] Re: About porting of GCC and possible redundancy
Mon, 26 Apr 2010 12:40:54 +0200
Kārlis Repsons <address@hidden> schrieb:
> On Sunday 25 April 2010 21:45:09 Volker Grabsch wrote:
> > However, for historical reasons, mingw-cross-env uses mingwrt and w32api
> > from MinGW, but the GCC from TDM, a small fork of MinGW.
> So MinGW and TDM (sorry, I see for first time, who are they?) have some
> parallel efforts on GCC? Why don't they just cooperate to produce one, if you
> have some idea?
I'm not part of either project, and I didn't bother to ask them,
so I simply don't know.
I guess that the MinGW people considered the first GCC 4.x versions
to be too unstable to officially support them. And the TDM people
had a different opinion, or had a daunting need for GCC 4.x, or
whatever, so they did it on their own.
This is a quite natural flow in free software projects. Stuff like
that happens from time to time. It isn't necessarily a bad thing,
and it doesn't automatically mean that people don't cooperate.
Also note that TDM didn't release anything after GCC 4.4.1, while
MinGW provides GCC 4.5.0. So it seems that TDM was just a temporary
> > Also note that patches from MinGW wander
> > back to the GCC project, so some GCC versions seem to be usable for
> > win32 cross compiling without any extra patches of the MinGW project,
> > because those are already included. However, we haven't yet checked in
> > how far that really works.
> Is it supposed, that in time the original GCC will support cross-compiling to
> some platforms and just their runtime environments and APIs will need to be
That would be the ideal case. And that might be the future.
But I guess the current split has organizational advantages, because
that way, the MinGW people can work independenty from the GCC project.
However, this is really a question should should ask to the MinGW
and/or GCC people.
> > BTW, in addition to GCC/mingwrt/w32api there is a fourth important
> > package: Binutils.
> > So when we
> > speak about porting, Binutils is a non-issue.
> Just still curious: how much work there currently is just to port GCC and why
> (basically) it has to be done?
This is also a question you should ask to the MinGW people.