Henrik wrote:
>
> Sorry if I get a bit aggravated. I really appreciate Your effort
> Arno, but
> external problems is not the way to go this time.
No problem, I just wanted to point onto that possibility.
---
Arno Garrels [TeamICS]
http://www.overbyte.be/eng/overbyte/teamics.html
>
> //Henr
Henrik wrote:
> So how to proceed? Can we come up with some other test scenarios?
> The answer must have something to do with _how_ ICS load the dll...
I'm not so sure, as I got it you called WSocketForceLoadWinsock
as well as WSocketCancelForceLoadWinsock from the new main form,
which results in
Well ok, strange sense of humor anyway since its pretty obvious that the
virus solution is not the rescue this time. I don't blame him though. I
myself speculate wildly and tend to grasp the "all mighty virus solution"
when it is something that I can't explain or duplicate at my end when my
custome
Wilfried Mestdagh wrote:
> Ok, so it is working, but still remains the question wy the
> dynamically loading of winsock should not work as it should on some
> machines ?
>
> Arno, you mention a virus, but should it then work with the
> ForceLoadWinsock ?
I know it doesn't make much sense, there
<<
So how to proceed? Can we come up with some other test scenarios?
The answer must have something to do with _how_ ICS load the dll... or any
other ideas?
// Henrik >>
Henrik, any default units that gets automatically added in your project like
madexcept or fastmm4 ?
--
To unsubscribe or
Hello Henrik,
What you could do is set in IDE some breakpoints at the places the dll
is loaded / unloaded. To see if this mechanism is working fine on the
offending machines. It is a counter that increment / decrement to decide
if winsock should be load or unload.
That's the best idea I have so f
Hi Wilfried!
I think Arno's virus response was meant to be a joke...? Anyhow it is not
the answer this time.
So how to proceed? Can we come up with some other test scenarios?
The answer must have something to do with _how_ ICS load the dll... or any
other ideas?
// Henrik
-Ursprungligt med
Ok, so it is working, but still remains the question wy the dynamically
loading of winsock should not work as it should on some machines ?
Arno, you mention a virus, but should it then work with the
ForceLoadWinsock ? I recall virusses that did some things with the TCP
stack, but it was more the s
Wilfried Mestdagh wrote:
> Could it be the loading of Winsock dll ?
That's obviously true, since it's the only effect I can imagine
that the secondary instance could have. Is Winsock still the origin
DLL, or has it been replaced by a Virus ;-) ?
---
Arno Garrels [TeamICS]
http://www.overbyte.be
Hi!
YES! Now we are getting somewhere.
When I manually load/unload winsock as Wilfried suggested, the error does
not show. So it has something to do with the loading of the dll.
I'm sending You the files Wilfried :-)
// Henrik
-Ursprungligt meddelande-
Från: [EMAIL PROTECTED] [mailto:[E
Hi Wilfried and thanks for a fast reply!
I'm sending You the new demo.
At my end it does not make any change removing the application wide
exception handler. This is just in the ICS demo not in any of my normal
applications.
I also want to point out that:
- It's the standard ICS MailSnd demo we a
Forgot to mention that the XP box below has a Dual-Core
Intel CPU, that should by very close to Henrik's Dell box.
Arno Garrels wrote:
> Wilfried Mestdagh wrote:
>> Hello Henrik,
>>
>> this is frustrating :(
>
> Welcome to the club.
>
>> I can do a test on several
>> computers if you want. Ple
Hello Henrik,
> * My workaround so far is to create another smtp component
> that exists all the time the application is running. I
> now discovered that this is also true with the Wsocket
> component. For instance put it on the main form and
> everything works on all computers.
Could it
Hello Arno,
Just received the compiled demo from Henrik. Tested on same boxes as you
did, nothing reproducable :( Tryed with the compiled exe from Henrik as
with recompiling from in IDE.
Something not good: In FormCreate of the ICS form there is an
appllication wide exception handler assigned whi
Wilfried Mestdagh wrote:
> Hello Henrik,
>
> this is frustrating :(
Welcome to the club.
> I can do a test on several
> computers if you want. Please include in your mail the steps nececary
> to reproduce the problem.
I wanted to help as well, but on 3 boxes here NT4 SP6a, W2K SP4,
XP SP2 (auto
Hi again Wilfried!
Nice to hear from You!
I've now tried to have 'stop on exceptions' both enabled and disabled in the
IDE but with the same result: the trace "disappears" at the earlier
described line.
I'm thankful for all the help I can get so if it is possible for You to
investigate this I'm v
Arno and others,
More facts and thoughts that perhaps can shed some light on
the error or point someone with more knowledge on ICS than
me in the right direction:
* You don't have to actually _send_ an email to get
the error. It also happens if You just connect.
To try this:
- create form
Hi Bjørnar!
Thanks for Your input!
As long as the OnRequestDone only is called with RqType = smtpQuit when the
component is finished working, it should not be any problem releasing the
parent form after that has happened?!
Best Regards
Henrik
-Ursprungligt meddelande-
Från: [EMAIL PROTEC
Hello Henrik,
this is frustrating :(
When the trace dissapears it could also be an AV (unles you have 'stop
on exceptions' enabled in the IDE).
Can you send me the offending project in private mail, including source
and the compiled exe file ? I can do a test on several computers if you
want. Ple
19 matches
Mail list logo
