Daniel Stenberg wrote:
> I suspect you mean "headers" and not "head". HEAD is a method in HTTP
> while response-headers is the headers that comes in the server response.
Yes I know what HEAD method is. It also occurred to me that the collection
of things called "header" should have a name on its own (I'm not a native
English speaker, though), and not just "the headers". So "the head" came
quite naturally (after "headers"), especially considering, that the method
for retrieving that collection is also called "HEAD".

> But when I try the sepheaders.c example right here and now, it works
> just as expected... ?
You know why? Because sepheaders.c is DIFFERENT. Besides, it is incorrect:
it returns nmemb, instead of expected nmemb*size.

You haven't really tried looking at my example, don't you? Or you looked at
my example and thought: "Another guy who can't understand our crazy,
nonorthogonal incorrectly documented output handling interface, so let us be
condescending and let him figure it out by himself!"

So I did that. Where in the f***ing documentation it is said, that if I
don't specify header writing function with CURLOPT_HEADERFUNCTION, the
output will be done using data writing function specified with
CURLOPT_WRITEFUNCTION? Since it wasn't said so, I dutifully supplied cURL
with a head output stream and a dummy (discarding) data writing function,
expecting that the default head output function will be used. But no, there
must be some magic in the world! The dummy data writing function was used
instead.

Phew, since I have finished reacting to your condescending answer, I could
now ask: where should I send patches to the documentation?

Sincerely,
Nick
Received on 2008-04-03