>
>> 2) What happens with data send to the server? Does the server throw
>> the partly received data away?
>
> That's not libcurl's business and I would say that depends entirely on
> what you did and what server you're operating against. You should
> consult the server docs/code/company/organization to get an answer to
> that.
>
OK, I'll ask about it.
>> How it is achieved that the server does not operate on incomplete
>> data? How it is signaled to the server, that the data is not complete?
>
> That's not signalled at all. You cut off the connection, and it is
> most likely disconnected. The latter is of course a clear signal to
> the server that the client won't send anything more! ;-)
>
Yes, I thought it is handled that way...

>> Therefore I am quite interested in your opinion about possible risks
>> of aborting HTTP requests ?
>
> I would say that they're likely done better in the latest release than
> in 7.16.3, but you won't get any guarantees from me... :-)
>
OK, I don't want to have some from you :-)

Thanks for your opinion on these topics,
Bye, Stefan
Received on 2008-04-05