[Gridflow-dev] osx ruby/ruby configure

[james tittle]

On Mar 21, 2006, at 5:31 PM, Mathieu Bouchard wrote:

> On Tue, 21 Mar 2006, james tittle wrote:
>
>> ...well, my first thought is that this is being included  
>> incorrectly, so
>> I did a recompile of that file with "g++ -E", and indeed this  
>> seems to
>> be the case: I'm thinking that we need to compile this as "gem.cpp"
>> instead of "gem.c"?
>
> For historical reasons, the suffix for all C++ files is .c, and  
> then I use
> gcc's -xc++ option to say it's really C++. The file "gem.c" is recent
> because I didn't want to even think about changing the makefile.

...I see:  I went ahead and change it to gem.cpp and added a target  
to the makefile, and while everything recompiled fine, I still wound  
up with the original problem :-(

> I still want you to quote me that OpenTransport file.

...I'm sending the whole file, cuz it's large and deprecated and  
shouldn't even be showing up; but, here's the relevent lines causing  
problems:

/System/Library/Frameworks/CoreServices.framework/Frameworks/ 
OSServices.framework/Headers/OpenTransport.h:723: error: expected  
identifier before numeric constant
/System/Library/Frameworks/CoreServices.framework/Frameworks/ 
OSServices.framework/Headers/OpenTransport.h:723: error: expected `}'  
before numeric constant
/System/Library/Frameworks/CoreServices.framework/Frameworks/ 
OSServices.framework/Headers/OpenTransport.h:723: error: expected  
unqualified-id before numeric constant
/System/Library/Frameworks/CoreServices.framework/Frameworks/ 
OSServices.framework/Headers/OpenTransport.h:4238: error: expected  
declaration before '}' token

enum {
   T_LISTEN                      = 0x0001, /* An connection request  
is available     */
   T_CONNECT                     = 0x0002, /* Confirmation of a  
connect request  */
   T_DATA                        = 0x0004, /* Standard data is  
available        */

...the T_DATA line is #723, so it's probably already defined  
elsewhere in gridflow...

/* Not used frequently enough to justify inlining.*/
/* OTLock is just a convenience type with some convenient macros.*/


typedef UInt8                           OTLock;

#define OTClearLock(lockPtr)   *(lockPtr) = 0
#define OTAcquireLock(lockPtr)   (OTAtomicSetBit(lockPtr, 0) == 0)

/ 
************************************************************************ 
*******
**
** FROM HERE ON DOWN ARE THE C++ Interfaces to Open Transport
**
************************************************************************ 
********/

#ifdef __cplusplus
}                   // Terminate C definitions

...the last bracket above is line 4238...

...I looked thru gem to see where carbon.h is included, and will try  
to clean up and see if that helps...I don't know much about reading  
the output of "g++ -E", but it seems to have lots of clues...but,  
it's 4+MB.s for gem.c!

jamie