rainer.jung at kippdata
Jul 13, 2012, 12:01 PM
Post #12 of 21
On 13.07.2012 18:02, Jim Jagielski wrote:
> If these can be added somewhat quickly, I'm willing to fast-track
> them into 2.4.3.
I drafted a patch available at
- I didn't fix the indentation in include/httpd.h in order
to keep the patch readable.
Some of the new codes have a short description which is a bit longer
than the longest one used up to now.
- I didn't "fix" the old define named "HTTP_REQUEST_URI_TOO_LARGE"
which should have been "HTTP_REQUEST_URI_TOO_LONG" since it
is defined in a public header file
- I included all changes proposed by Julian
- there is a big gap of unused numbers between 208 and 226 which
I filled with "unknown" as was done before due to the limitations
in ap_index_of_response() (focus on performance there)
- I added canned error strings for the new codes
- I did not yet define new error documents. The new pages could be
Furthermore some other 4xx and 5xx codes already defined in httpd.h
also have no error page:
I guess that means defining ones for the new codes is not a must ...
- I did not check, which of the new codes actually should change
behaviour of the web server!!
> On Jul 12, 2012, at 9:07 AM, Jim Jagielski wrote:
>> On Jul 12, 2012, at 2:34 AM, Julian Reschke wrote:
>>> On 2012-07-11 19:15, Roy T. Fielding wrote:
>>>> I don't know of any issues with 308, and Julian generally knows what
>>>> he is doing with regard to HTTP. In general, we should consider
>>> Thanks :-)
>>>> the IANA registry to be authoritative unless it is a known bug,
>>> In which case we should fix the registry.
>>>> which means we should support everything in
>>> Yes. If we want to get all of these in, I can open a separate ticket and provide another patch.
>>> Best regards, Julian