[Fwd: GCC 4.3.0 Status Report (2007-08-09)]

[François-Xavier Coudert]

Hi all,

> We entered Stage 2 on July 6th.  I plan to put us into Stage 3 on
> September 10th.  At that point, we will accept only bug-fixes -- no
> more new features until Stage 1 for GCC 4.4.
>
> Are there any folks out there who have projects for Stage 1 or Stage 2
> that they are having trouble getting reviewed?  Any comments
> re. timing for Stage 3?

I think it raises a few questions about the Fortran projects that
haven't been included yet. Projects that don't make it into 4.3 will
have to wait until branching, which is still a long time away (based
on previous releases, we're looking Dec 2007/Jan 2008). Patches will
have to be maintained against bitrot during that time.

 To me, the issues that need to be addressed are:

  ** The "procedure pointers and type-bound procedures" by Janus Weil. **

Janus, I now we pinged you recently, but I guess we now really need to
move forward quickly. IIUC, your SoC patches need to be reviewed and
approved before entering stage 3. Since you have posted only few
patches in the past, we can expect a bit of going back and forth
before final approval (this is not a comment on the quality of your
patches in any way; just that there always is some adaptation required
for the first patches, there are so many requirements for inclusion!).
Thus, it is probably good that you start preparing your current
patches for formal review (testcases, ChangeLog entries, formatting,
explanation, etc.) while finishing your work on the procedure pointers
themselves. In any case, we probably will need patches before the end
of August.

Of course, this is only my own opinion, and I don't know how much you
have been in touch with other gfortran maintainers and Steven
Bosscher. Maybe there's been a lot more going on that what I've seen.


  ** Formal/actual argument checking for same file procedures, by Paul Thomas **
  ** Revamp of read_module, same usual suspect **

I know there hasn't been much comment on your proposed patch for
file-wide structures, but I really think it was based on lack of
enough technical understanding of what you proposed than from lack of
interest. I, for one, think it's a really great way of doing things,
but I couldn't actually comment on the technical merit of your patch;
although, since you wrote it, I'm sure it's great! ;-)

So, the question now is, do you think the timeframe of doing this is
compatible with 4.3? After stage 3 begins, we'll have time to fix
bugs, but we won't be able to introduce new features any more... It's
your call, obviously.



  ** IEEE module, by Asher Langton **

Asher, you mentioned recently that you were interested in implementing
the IEEE modules. From your mail, I take it that your project is still
an idea and not advanced enough for a 4.3 inclusion, am I right?



PS: After that, well, we'll be fixing bugs and making plans for the
new features in 4.4! I'd like to propose asynchronous I/O, IEEE
modules (I can help Asher there) and non-ASCII character kinds.

-- 
FX