On Sat, Feb 28, 2004 at 01:00:27AM +0100, Branko ??ibej wrote:
> Anyway, the patch still won't handle all Unices supported by APR correctly.
>
> >However, PY_LONG_LONG is just a macro for long long.
> >
> As a matter of fact, it's not. PY_LONG_LONG is a macro fot _some_ 64-bit
> type. It may be defined to "long long" on your platform, but not on
> mine; e.g., there _is_ no "long long" on Windows.
I don't think you understand what the #if stuff is for. It just gets
rid of a WARNING. The code works fine on pretty much all platforms as
is, they just might get a WARNING.
Now as far as there being no "long long" on Windows is irrelevent. SWIG
handles that for us and converts it to __int64. Before this change you
were building python bindings with long long in the code for ages. My
patch simply changes it to give you the identical code you had *BEFORE*
the change that created the bug. While still providing the modified
code for the platforms where those types are only long.
So basically there may be some platforms that get a warning. But I've
yet to hear of anyone identify a platform that entirely doesn't work.
I'd love to eliminate warnings on all platforms. But at this point
nobody has identified a warning based on this code. For all I know the
warning I was fixing was very gcc specific (format not matching the
type) and you may not even see it on Windows.
And the original reporter of the problem has already reported that it's
fixed. I'm not getting warnings on LP64 platforms or ILP32 platforms.
So if you have a specific problem to report I'll be happy to work on it.
But please stop the handy wavy assumptions about how the change is
broken. It's less broken than the code we had in 0.37 and 1.0.0.
--
Ben Reser <ben@reser.org>
http://ben.reser.org
"Conscience is the inner voice which warns us somebody may be looking."
- H.L. Mencken
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Feb 28 01:50:29 2004