This pisses me off

This is what pisses me off. We have an application that is a very
large commercial application. The application has several COM dlls and
MFC code
Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
(passed to free/delete is invalid).It is not reproducible,

What is frustrating is that there is no help from MS to dectect such
problems. We already wasted 2 months tring different approaches.

PageHeap that every Microsoft Employee so proudly talks about is the
peice of crap.

Can you imagine allocating 1 page after every allocation. Maybe it
will work for test apps that Microsofties write for their own testing
but it is a piece of crap for a big application ...


I am so sick of this problem. MS keeps talking about .NET LongHorn,
bullshit etc but never tries to research ways to help us fixing the
existing problems.

I wish all developers and companies out there buycott the ,NET and all
other crap Microsofties keep pushing at us.

Many of you may sense frustration in my mail and I really am
frustrated .

Rob
Sun Jun 27 19:31:01 CDT 2004

So why don't you go complain to the Microsoft.Public.vc.mfc group or the
platform SDK group?
And what does your MFC rant have to do with .NET?
Frankly, this group is for people using .NET, not for rants. Get a blog.
Lots of devs have more legitimate problems than your ranting and come here
to get answers, so what makes you so special? Grow up please.

-Rob Teixeira [MVP]

"Humber Consumer" <consumer62000@yahoo.com> wrote in message
news:5e1334d1.0406271341.16710c83@posting.google.com...
> This is what pisses me off. We have an application that is a very
> large commercial application. The application has several COM dlls and
> MFC code
> Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
> (passed to free/delete is invalid).It is not reproducible,
>
> What is frustrating is that there is no help from MS to dectect such
> problems. We already wasted 2 months tring different approaches.
>
> PageHeap that every Microsoft Employee so proudly talks about is the
> peice of crap.
>
> Can you imagine allocating 1 page after every allocation. Maybe it
> will work for test apps that Microsofties write for their own testing
> but it is a piece of crap for a big application ...
>
>
> I am so sick of this problem. MS keeps talking about .NET LongHorn,
> bullshit etc but never tries to research ways to help us fixing the
> existing problems.
>
> I wish all developers and companies out there buycott the ,NET and all
> other crap Microsofties keep pushing at us.
>
> Many of you may sense frustration in my mail and I really am
> frustrated .

Jerry
Sun Jun 27 21:33:34 CDT 2004

Good for you that you can always find somebody to blame, even when you have
absolutely no idea what you're talking about. Let me guess, you work as a
consultant, don't you?

Just FYI: an MFC application does not use .Net. If it was a .Net application
it wouldn't be able to free an invalid memory block.

Jerry

"Humber Consumer" <consumer62000@yahoo.com> wrote in message
news:5e1334d1.0406271341.16710c83@posting.google.com...
> This is what pisses me off. We have an application that is a very
> large commercial application. The application has several COM dlls and
> MFC code
> Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
> (passed to free/delete is invalid).It is not reproducible,
>
> What is frustrating is that there is no help from MS to dectect such
> problems. We already wasted 2 months tring different approaches.
>
> PageHeap that every Microsoft Employee so proudly talks about is the
> peice of crap.
>
> Can you imagine allocating 1 page after every allocation. Maybe it
> will work for test apps that Microsofties write for their own testing
> but it is a piece of crap for a big application ...
>
>
> I am so sick of this problem. MS keeps talking about .NET LongHorn,
> bullshit etc but never tries to research ways to help us fixing the
> existing problems.
>
> I wish all developers and companies out there buycott the ,NET and all
> other crap Microsofties keep pushing at us.
>
> Many of you may sense frustration in my mail and I really am
> frustrated .

Niki
Mon Jun 28 11:28:58 CDT 2004

Hey buddy, you made those errors, you fix them. MS provides tools, which do
get better in time, but are of course not perfect. Would you like them to
stop enhance their tools until they're perfect? I surely wouldn't.

So, take a deep breath, and go back to your debugger!

Niki

"Humber Consumer" <consumer62000@yahoo.com> wrote in
news:5e1334d1.0406271341.16710c83@posting.google.com...
> This is what pisses me off. We have an application that is a very
> large commercial application. The application has several COM dlls and
> MFC code
> Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
> (passed to free/delete is invalid).It is not reproducible,
>
> What is frustrating is that there is no help from MS to dectect such
> problems. We already wasted 2 months tring different approaches.
>
> PageHeap that every Microsoft Employee so proudly talks about is the
> peice of crap.
>
> Can you imagine allocating 1 page after every allocation. Maybe it
> will work for test apps that Microsofties write for their own testing
> but it is a piece of crap for a big application ...
>
>
> I am so sick of this problem. MS keeps talking about .NET LongHorn,
> bullshit etc but never tries to research ways to help us fixing the
> existing problems.
>
> I wish all developers and companies out there buycott the ,NET and all
> other crap Microsofties keep pushing at us.
>
> Many of you may sense frustration in my mail and I really am
> frustrated .

Jason
Tue Jun 29 09:32:01 CDT 2004

I would try BoundsChecker or another tool to watch your allocations and
frees. You would be amazed at the small leaks that add up to the
unreproducable types of things you are talking about.

OR Switch to Dot Net. Those memory problems have gone away.




"Humber Consumer" <consumer62000@yahoo.com> wrote in message
news:5e1334d1.0406271341.16710c83@posting.google.com...
> This is what pisses me off. We have an application that is a very
> large commercial application. The application has several COM dlls and
> MFC code
> Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
> (passed to free/delete is invalid).It is not reproducible,
>
> What is frustrating is that there is no help from MS to dectect such
> problems. We already wasted 2 months tring different approaches.
>
> PageHeap that every Microsoft Employee so proudly talks about is the
> peice of crap.
>
> Can you imagine allocating 1 page after every allocation. Maybe it
> will work for test apps that Microsofties write for their own testing
> but it is a piece of crap for a big application ...
>
>
> I am so sick of this problem. MS keeps talking about .NET LongHorn,
> bullshit etc but never tries to research ways to help us fixing the
> existing problems.
>
> I wish all developers and companies out there buycott the ,NET and all
> other crap Microsofties keep pushing at us.
>
> Many of you may sense frustration in my mail and I really am
> frustrated .

Ed
Sat Jul 03 19:04:14 CDT 2004

Have you tried a product like BoundsChecker
(http://www.compuware.com/products/devpartner/bounds.htm) to figure out
exactly what's happening?

"Humber Consumer" <consumer62000@yahoo.com> wrote in message
news:5e1334d1.0406271341.16710c83@posting.google.com...
> This is what pisses me off. We have an application that is a very
> large commercial application. The application has several COM dlls and
> MFC code
> Occassionally we get assert on 1040 line in dbgheap.c,heap pointer
> (passed to free/delete is invalid).It is not reproducible,
>
> What is frustrating is that there is no help from MS to dectect such
> problems. We already wasted 2 months tring different approaches.
>
> PageHeap that every Microsoft Employee so proudly talks about is the
> peice of crap.
>
> Can you imagine allocating 1 page after every allocation. Maybe it
> will work for test apps that Microsofties write for their own testing
> but it is a piece of crap for a big application ...
>
>
> I am so sick of this problem. MS keeps talking about .NET LongHorn,
> bullshit etc but never tries to research ways to help us fixing the
> existing problems.
>
> I wish all developers and companies out there buycott the ,NET and all
> other crap Microsofties keep pushing at us.
>
> Many of you may sense frustration in my mail and I really am
> frustrated .