How extensive is this "leak". I am using Indy in commercial applications.
We recently crashed at a customer site. I cannot tell if it was from our
appliaction or others.
I belive I compiled that release with 10.1.6 though I have moved on to
10.2.3. I am using the TCPIP client only.
Make no doubt that I love the Indy suite but I need something I can deploy
with my product. I don't have a problem paying for the components and did
not go for Indy, simply because it was free.
I am unclear as to how recompiling would correct a memory leak or why an
intentional leak would be placed in the code in the first place. It it
intended to be a time bomb?
Thanks,
Larry
"Steve Maughan" <GoogleMe@www.com> wrote in message
news:47310445@newsgroups.borland.com...
> Remy,
>
> Many thanks!!
>
> Steve
>
> "Remy Lebeau (TeamB)" <no.spam@no.spam.com> wrote in message
> news:4730f407$1@newsgroups.borland.com...
>>
>> "Steve Maughan" <GoogleMe@www.com> wrote in message
>> news:4730cfea$1@newsgroups.borland.com...
>>
>>> I've been told that this is intentional (not sure why it necessary) but
>>> there is a version of Indy that doesn't have this leak.
>>> Is this true?
>>
>> No. However, you can recompile Indy with IDFREEONFINAL defined
>> (IdCompilerDefines.inc would be a good place to put it) in order to
>> remove the leaks. If you upgrade to a newer Indy 10 snapshot (what ships
>> with D2007 is *not* the latest), then it calls
>> SysRegisterExpectedMemoryLeak() when IDFREEONFINAL is not defined, so the
>> leaks do not appear in FastMM's leak report when enabled.
>>
>>
>> Gambit
