Mingw-w64 has finally been approved by Red Hat legal!

Erik van Pienbroek erik at vanpienbroek.nl
Wed Feb 22 22:37:47 UTC 2012


Hi everybody,

Today we received word from Red Hat Legal that the mingw-w64 toolchain
is no longer subject to any known legal blocks. This means that (after a
year of waiting) we finally can continue with the introduction of the
mingw-w64 toolchain in Fedora!

The timing of this legal approval is a bit too late to get the mingw-w64
toolchain introduced in Fedora 17 as the guidelines state that no new
features should be added to the distro anymore after the Feature Freeze
(which was two weeks ago). However, we (Kalev, Kai and me) think that it
is realistic to get the mingw-w64 toolchain introduced in Fedora 17
before the Beta development freeze (March 20). After all, it would be a
shame if we'd have to wait until October/November while we already have
everything ready. The use of mingw-w64 has already been tested and
improved for over a year thanks to various contributors (you rock!) so
we think it's safe enough to introduce mingw-w64 even though we missed
the Fedora 17 Feature Freeze.

So where do we stand now?

We currently have 114 mingw packages in Fedora rawhide. Out of these 114
packages 17 packages are already using the new naming scheme (with the
package name prefix 'mingw-' instead of 'mingw32-').

In the testing repository there currently are 75 packages (out of the
114 packages in Fedora) which are already ported to the new packaging
guidelines and have both win32 and win64 support. That's about 2/3rd of
all mingw packages which are currently in Fedora

What's the current short-term plan?

Kalev and I have come up with a list of tasks which we need to get done
before March 20 to make the introduction of mingw-w64 happen in Fedora
17:

- Get a separate F17 buildroot from rel-eng
- Finish the core toolchain package reviews
- Import the new packages in git (both master and f17 branches) and
  build them in f17-mingw koji buildroot
- Perform a mass rebuild of all mingw packages in f17-mingw
- Check if everything builds/works fine
- Create an bodhi update ticket which includes the 5 toolchain packages
  and all rebuilt packages (more than 100)
- Rawhide automatically inherits the new builds here
- (Optionally) Get the blessing of FESCO to allow this late feature in
Fedora 17
- File a ticket to have the FPC activate the new packaging guidelines
- Profit!

Kalev already started with reviewing the mingw-headers package. There
currently are three other packages which still need reviewing:
mingw-binutils: https://bugzilla.redhat.com/show_bug.cgi?id=673786
mingw-gcc: https://bugzilla.redhat.com/show_bug.cgi?id=673788
mingw-crt: https://bugzilla.redhat.com/show_bug.cgi?id=673792
Any help with those remaining three package reviews is really
appreciated!!

Once these toolchain packages are reviewed and approved the rest should
be a pretty easy flight which should only take a couple of days. The
package reviews is the most critical part here so we need to get that
resolved as soon as possible.

Okay, so what about the longer-term plan?

Once the toolchain packages are available in Fedora 17 we can start with
merging all the changes from the testing repository back to Fedora. For
the 17 packages which are currently in Fedora and are already using the
new package naming scheme this is a straightforward commit and build.

The other remaining 97 packages need to be renamed first before win64
support can be added. In order to get a package renamed in Fedora one
has to file a review request and have somebody else perform a complete
review of it. As we all know that this is really time consuming job,
especially for 97 different packages. We could try to ask FESCO for an
exception to this package rename procedure for all these 97 packages or
a subset of them (based on what the maintainers of the packages in
question desire). My recommendation would be to file a mass rename
request to FESCO for the 75 packages which are already ported to the new
packaging guidelines in the testing repo. Nothing is carved in stone
about this subject yet, but we can already start a discussion about this
subject now. The earliest moment when this all will be relevant is when
all short-term goals are achieved before March 20.


I would also like to take a moment to thank all the people who have been
involved in the development and testing of the mingw-w64 based packages
in the testing repository up to now. Thanks to you guys the mingw-w64
based packages in the testing repository are in a great shape right now
and we can make the transition relatively painless. In alphabetical
order here are the people who I would like to thank for their help in
testing and development:

Andreas Bierfert
Boszormenyi Zoltan
Christophe Fergeau
Dominik Schmidt
Dov Grobgeld
Jay Higley
Kai Tietz
Kalev Lember
Levente Farkas
Marc-André Lureau
Michael Cronenworth
Mikkel Kruse Johnsen
Paweł Forysiuk
Richard Jones
Tim Mayberry

I hope I didn't forget any one :)

Kind regards,

Erik van Pienbroek







More information about the mingw mailing list