> 7!:0'' may or may not report non-J memory allocations. 7!:0 reports on space allocated/deallocated by the memory manager in the J interpreter.
----- Original Message ----- From: Oleg Kobchenko <[EMAIL PROTECTED]> Date: Monday, February 11, 2008 7:45 Subject: Re: [Jprogramming] win32 platimg To: Programming forum <[email protected]> > Thank you for testing it under Wine. > > I noticed the image location issues (same in > Test section in main file). I will try to accommodate > version differences in the next version. > > 7!:0'' may or may not report non-J memory allocations. > > > --- bill lam <[EMAIL PROTECTED]> wrote: > > > VariantClear doesn't crash J under wine. > > > > However by printing 7!:0'' before and after running test.ijs > (after amending it > > to use getimg1), the numbers are the same with or without > VariantClear so that I > > still think that VariantClear or the previous olerelease is > redundant.> > > BTW test.ijs contains /path/to/images that need fixing in J602. > > > > Oleg Kobchenko wrote: > > > No, it's OK to use VariantClear to any properly > > > created variant. VariantInit just properly zeros > > > memory and sets type to VT_EMPTY. > > > > > > VariantClear should exacly call the one and only > > > correct Release. > > > > > > Yes, obj->vtbl->Release is the definition > > > of ((*IUnknown)pobj)->Release. > > > > > > Does it not work on Wine? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
