It is theoretical and I found a workaround:
using renamed j dlls - one per thread: j1.dll, j2.dll etc.

This returns different callbacks, and as I believe, any
global variables are separate, so there is yet less
chance of multithreaded issues as Raul pointed out.

see multi3.ijs at
  http://www.jsoftware.com/jwiki/OlegKobchenko/Threads

--- Eric Iverson <[EMAIL PROTECTED]> wrote:

> Your observation is correct. The callback implementation does not allow 
> for multiple instances in the same task. This is a solvable problem but 
> would take some effort. Is your concern currently theoretical or do you 
> have a serious application in mind that requires this?
> 
> ----- Original Message ----- 
> From: "Oleg Kobchenko" <[EMAIL PROTECTED]>
> To: "General forum" <[email protected]>
> Sent: Sunday, December 31, 2006 8:21 AM
> Subject: [Jgeneral] multithreaded callback
> 
> 
> > It turns out, cdcb returns the same address
> > for different J instances in the same process,
> > making it impossible to use callbacks for
> > slave J instances.
> >
> > Any suggestions?


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to