After a quick look at what you are describing... For something like this to
happen, either you've got to have some horrible heap corruption, multiple
runtimes, someone returning pointers to non-allocated memory and telling
people its really allocated memory, or someone returning pointers to memory
that its already deleted somehow. If there were something awry in the
DOMString's transcode() method, the latter could happen I guess. A quick
look at it indicates that its a non-trivial method, and I'm not sure if its
doing everything it should do.

This is also the kind of thing that happens when multiple runtimes are
used, where we allocate something (from our runtime) and give it back to
you, and you then try to delete it (from your runtime who knows nothing
about it.) Make sure that you are using the DLL runtime under VC++ when you
are building against our released binarise.

----------------------------------------
Dean Roddey
Software Weenie
IBM Center for Java Technology - Silicon Valley
[EMAIL PROTECTED]



Dean Roddey/Cupertino/[EMAIL PROTECTED] on 01/26/2000 10:37:58 AM

Please respond to [EMAIL PROTECTED]

To:   [EMAIL PROTECTED]
cc:
Subject:  Re: bug??







Andy is in a W3C meeting right now, so he can't answer. But I think that
this was already reported and is fixed now in the latest code, but if not
I'll make sure that he sees it once the meetings are over so that he can
slip it in before the new release.

----------------------------------------
Dean Roddey
Software Weenie
IBM Center for Java Technology - Silicon Valley
[EMAIL PROTECTED]



"Rupesh kumar" <[EMAIL PROTECTED]> on 01/26/2000 06:40:37 AM

Please respond to [EMAIL PROTECTED]

To:   <[EMAIL PROTECTED]>
cc:
Subject:  bug??




Version number : 1.0.1

Platform: NT4 + SP4

Compiler: MSVC6

Sample file: Attached---Its basically a cut paste of  DOMPrint sample

Problem: In function�� ostream&  operator<<(ostream& target, const
DOMString& s)

When delete is called on p it results in a Debug Assertion.

In dbgheap.c and while evaluating the expression
_CrtlsValidHeapPointer(pUserData)


Regards

Rupesh.

-------------May the "source" be with  you----------------
(See attached file: dom-fx-trav.cxx)

(See attached file: dom-fx-trav.cxx)

Attachment: dom-fx-trav.cxx
Description: Binary data

Reply via email to