On Mar 24, 2007, at 12:49 AM, Norman Palardy wrote:

>
> On Mar 23, 2007, at 8:52 PM, Guyren Howe wrote:
>
>> On Mar 23, 2007, at 8:48 PM, Guyren Howe wrote:
>>
>>> Why should returning a single value be easy but multiple values be a
>>> pain in the proverbial? I don't see anything special about returning
>>> one value.

If REAL is to eventually provide a means to create DLL's / plugins  
for other tools (like Photoshop) they'll need to adhere to standard  
calling conventions for the various architectures they compile for.
That includes adhering to the mechanisms already in place for  
returning values from called functions.
I'd bet this includes returning one result in a register (which  
mostly means returning one result)

Having to make it so a function can "know" what language called it so  
it can adhere to certain calling conventions could be problematic  
(ie/ IF called by RB then I can return values in one way and if  
called by a C function then call it another way)

This seems especially true when using an RB function in a callback  
where the calling conventions MUST be adhered to and the RB function  
has no real way to determine if it was called by an external C  
function or an internal RB function (as that same function could be  
called by either means)

But without some kind of confirmation by Mars as to whether this  
suspicion is indeed true it's just speculation.

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>

Reply via email to