No idea what the B stands for.  Alex should be the expert of J/net
interfacing. BTW there should be 4 cases
32 net call 32 J
32 net call 64 J
64 net call 32 J
64 net call 64 J
which of them failed?  What is the expect output and what was the incorrect
output it produced? May be it is a data marshalling issue. 

IIRC there is a registry setting to control which of 32/64-bit ole exe server 
being the candidate.  I'm not sure as I seldom use window.

Втр, 12 Окт 2010, Martin Pelletier писал(а):
>   On 2010-10-12 10:44, bill lam wrote:
> > Apparently the array int[] is passed as safearray. I'm not sure in J's
> > built-in olecom can handle safearray.  Did you test the code run inside
> > 32-bit but failed on 64-bit environment?
> Yes, I tested in a 64-bit environment with J6-32bit. The return value is 
> the same, and the J value is the same.
> > Or you may try fix the int[] object
> > and use getm/setm to directly read and write memory content.
> Interesting... That might be a workaround, if GetB is not intended to do 
> arrays in the first place. Incidentally, what does the B in GetB stand 
> for? Do you know if there is documentation on all COM functions?
> > You may
> > get a fairly accurate idea of the internal structure of a J array
> > from browsing the source code of the j memory map file script.
> >
> Also interesting! I will check it out!
> 
> Martin Pelletier

-- 
regards,
====================================================
GPG key 1024D/4434BAB3 2008-08-24
gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to