On Sep 3, 2008, at 9:24 AM, Charles Lepple wrote:
> On Wed, Sep 3, 2008 at 10:22 AM, William G. Scott
> <[EMAIL PROTECTED]> wrote:
>>
>>
>> On Sep 3, 2008, at 4:54 AM, Martin Costabel wrote:
>>
>>> William G. Scott wrote:
>>>
Apparently gmp compiled on my G5 ppc is not compatible with a
>>
On Wed, Sep 3, 2008 at 10:22 AM, William G. Scott
<[EMAIL PROTECTED]> wrote:
>
>
> On Sep 3, 2008, at 4:54 AM, Martin Costabel wrote:
>
>> William G. Scott wrote:
>>
>>> Apparently gmp compiled on my G5 ppc is not compatible with a G4
>>
>> Yes, this is an old problem. Often it gives bus errors
William G. Scott wrote:
>
>
> On Sep 3, 2008, at 4:54 AM, Martin Costabel wrote:
>
>> William G. Scott wrote:
>>
>>> Apparently gmp compiled on my G5 ppc is not compatible with a G4
>> Yes, this is an old problem. Often it gives bus errors or illegal
>> instruction errors. The G5 build app
On Sep 3, 2008, at 4:54 AM, Martin Costabel wrote:
> William G. Scott wrote:
>
>> Apparently gmp compiled on my G5 ppc is not compatible with a G4
>
> Yes, this is an old problem. Often it gives bus errors or illegal
> instruction errors. The G5 build apparently uses ppc machine
> instru
During the last re-build of KDE , I saw several messages
> /sw/bin/dpkg - warning: downgrading kdelibs3-unified-dev from
> 3.5.9-3 to 3.5.9-2.
and also, at least, for kdebase-unified-dev ..
Those came right before a new build, when fink checks for (b)deps and
conflicts;
nevertheless, reinsta
William G. Scott wrote:
> Apparently gmp compiled on my G5 ppc is not compatible with a G4
Yes, this is an old problem. Often it gives bus errors or illegal
instruction errors. The G5 build apparently uses ppc machine
instructions that don't exist on G4.
In principle, gmp has some code to
Am 02.09.2008 um 21:17 schrieb Alexander Hansen:
> Are you saying that you made such a local modification to the package?
Yes. Since TclTk is not that important for me I can live without the
newest version. So I am using a local variant of the original package.
> If so, why didn't you mention
Peter Dyballa wrote:
[]
> Is $exec_prefix set correctly when using the original Fink package?
Yes. Fink sets --prefix=/sw automatically, and exec_prefix is set from
$prefix by configure, if you don't set it explicitly. That is, unless
you fiddle with the tcl stuff. The configure script has the f
