So I did create the sample app and it worked. That led me to find the
bug in a completely unrelated portion of code that was doing a close()
on an uninitialized file handle. Im not sure exactly why that caused
the behavior I saw, but it was certaintly a bad thing to do. Sorry for
the false alarm and thanks for the pointers.
Rob

On Aug 24, 2004, at 6:37 PM, Rob Mason wrote:

> So the behavior I see falls along a couple lines:
>
> 1) I see output on the screen which is the content of one of those
> buffers (write_buffer) periodically. Somehow its getting dumped to the
> screen although nothing in the source touches it.
> 2) I get the wrong HTTP response code back -- it looks like the
> response code is from another operation.
>
> Its very easy to repeat, but the app its in is >10K lines. Because of
> the multithreaded requirements its not going to be easy to put it in a
> small example I think, but if thats what's needed I can try to create
> a sample app that creates this behavior.
>
> Rob
>
> p.s. I apologize to folks for the double posts before, my mail client
> used the wrong mailbox to send.
>
> On Aug 24, 2004, at 5:44 PM, Daniel Stenberg wrote:
>
>> On Tue, 24 Aug 2004, Rob Mason wrote:
>>
>>> Thanks for the comments on my threading/programming practice, but it
>>> really didn't address my real problem though. Any ideas on that
>>> part?
>>
>> I agree that given your responses, the issues I raised all seem
>> irrelevant to the problem as you reported it. They just seemed odd.
>>
>> I have no explanation to the problems you report. Can you give us a
>> more detailed description of the "corruption between the two streams
>> in terms of the responses and behavior" ?
>>
>> Is it hard to repeat? Can you post a fully working example source
>> code that repeats the problem?
>>
>> --
>> Daniel Stenberg -- http://curl.haxx.se -- http://daniel.haxx.se
>> Dedicated custom curl help for hire: http://haxx.se/curl.html
>>
>
>
Received on 2004-08-25