Since there is a proposal to backport this directive to the 2.0 branch, would it 
make more sense to rename it to something else and avoid the confusion before it is 
backported?  It just seems a little strange to be getting an error header back in a 
successful response.  If the point is to allow a specific header to be returned for 
all responses, we should probably expand the options for the Header directive rather 
than add a new directive.

Syntax: Header set|append|add|unset|echo  header [value [env=[!]variable]] [persist]

persist 
The response header with this name is included even on errors and persists across 
internal redirects. 


Brad

Brad Nicholes
Senior Software Engineer
Novell, Inc., the leading provider of Net business solutions
http://www.novell.com 

>>> [EMAIL PROTECTED] Wednesday, June 02, 2004 4:02:25 PM >>>
* "Brad Nicholes" <[EMAIL PROTECTED]> wrote:

>    If an ErrorHeader is defined, shouldn't it only be included on
> non-200 http responses (ie. only on 3xx,4xx,5xx responses as the
> documentation states)?

No. The documentation states:

| This directive can replace, merge or remove HTTP response headers *also*
| during 3xx, 4xx and 5xx replies.

It's a kind of a misnomer anyway (named after r->err_headers_out).

nd
-- 
"Eine Eieruhr", erkl�rt ihr Hermann, "besteht aus einem Ei. Du nimmst
das Ei und kochst es. Wenn es hart ist, sind f�nf Minuten um. Dann wei�t
du, da� die Zeit vergangen ist."
                             -- Hannes H�ttner in "Das Blaue vom Himmel"

Reply via email to