Am Thu, 02 Nov 2006 23:48:49 +0100 hat Daniel Stenberg <daniel_at_haxx.se>
geschrieben:

> On Thu, 2 Nov 2006, Olaf wrote:
>
>> I did a patch for this BUG - and for me, it seems to run. Sorry, but I
>> had
>> to introduce a new ErrorCode - that is only used internally. I can also
>> define a flag in one of the structures to transport the needed
>> information
>> between the calls.
>>
>> I tried not to change to much. Would you please have a look on the
>> changed
>> files (diff against 7.16.0 is attached) if you see any problems. I am
>> open
>> for all comments.
>
> Thanks!
>
> I didn't really like the addition of a new external error code just to
> deal
> with an internal situation. We could solve that in many different ways,
> but
> here's one take at doing the same without the extra public error code. I
> also
> modified the check for "Connection: close" to use the already existing
> internal function for it.
>
> Any particular objections on this approach? Does it still work for you? I
> don't have any decent way to test this right now.
>
Hi,

your approach is smarter and fits better into the code - and it works also
fine
in my environment for https and http. So, I am ok with this version. For
me,
Bug#25 is closed.

Thanks.
Received on 2006-11-03