PATCH: Add two ObjC++ callback functions to cp/decl.c
Ziemowit Laski
zlaski@apple.com
Fri Sep 10 23:10:00 GMT 2004
On 10 Sep 2004, at 15.12, Mark Mitchell wrote:
> Ziemowit Laski wrote:
>
>>> I think you need to get these kinds of patches reviewed and
>>> approved. They're not "obvious" just because they "don't affect C++
>>> control flow."
>>
>> I guess what I meant is that they're obvious from _your_/C++
>> perspective, since C++ execution is not altered by it. No, I suppose
>> they're not obvious from the ObjC++ perspective, but I take full
>> responsibility for it as the ObjC++ maintainer. :-)
>
> These patches are occurring in the C++ front end. Therefore, a C++
> front-end maintainer should approve them. Just because they are only
> used in Objective-C++ does not mean that they have no impact on the
> C++ front end.
The patches are occurring in _files_ _belonging_ to the C++ front-end.
They affect the C++ front-end only if they prevent the file(s) in
question from compiling.
Taking your reasoning to its logical conclusion, all patches to the C++
front-end should be reviewed by the ObjC++ front-end maintainer (yours
truly) because they will almost _certainly_ have impact on ObjC++. :-)
> For example, a change to these data structures now requires changing
> your new code as well.
And, if the data structures do change, my code will break and I'll have
to fix it. When you designed the new C++ declarator data structures, I
don't recall you as much as telling me about it so that I could plan my
ObjC++ work accordingly. So what else is new? :-)
Now, I certainly don't mind your reviewing of the ObjC++-specific
stuff. The longer the process takes, the longer gcc 4.0 will stay
parked in the garage, which is probably fine with some people anyway.
:-) I just fail to see the logic of your argument, that's all.
--Zem
More information about the Gcc-patches
mailing list