
Signature
Chris Mullins, MCSD.NET, MCPD:Enterprise, Microsoft C# MVP
http://www.coversant.com/blogs/cmullins
That shouln't cause a particular object to not be collected though.
My interpretation of "memory leak" could have been different than yours or
Ward's. Mine what that a particular object wasn't collected... I don't pay
attention to usable space as an indiciation of memory leaks. Due, as you
say, to fragmentation...

Signature
Browse http://connect.microsoft.com/VisualStudio/feedback/ and vote.
http://www.peterRitchie.com/blog/
Microsoft MVP, Visual Developer - Visual C#
> > I can't think of any reason why the GC won't collect all inaccessible
> > memory.
[quoted text clipped - 4 lines]
> When the heap is fragmented, that compaction doesn't work very well, and you
> can end up without much "free" memory...
Chris Mullins [MVP] - 19 Apr 2007 22:26 GMT
I agree - this wouldn't at all cause something to not be collected.
The original poster should be able to use one of the existing tools
(Scitech's Memory Profiler, Red Gate's Memory Profiler, heck - even WinDbg
with SOS can do this), walk the root paths, and see why his object wasn't
collected. It should be pretty easy.
The tools have recently started getting so very good. Red Gate & SciTech now
have x64 support, SciTech can load in MiniDump files so you can do
post-crash memory analysis - it makes debugging so much easier than it used
to be.

Signature
Chris Mullins, MCSD.NET, MCPD:Enterprise, Microsoft C# MVP
http://www.coversant.com/blogs/cmullins
> That shouln't cause a particular object to not be collected though.
>
[quoted text clipped - 13 lines]
>> you
>> can end up without much "free" memory...
Ward Bekker - 20 Apr 2007 13:00 GMT
Thanks guys for your help,
More discussion of this thread can be found here:
news://news.microsoft.com:119/e0ZNz6mgHHA.4704@TK2MSFTNGP06.phx.gbl