httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "William A. Rowe, Jr." <>
Subject Re: cvs commit: httpd-2.0/modules/mappers mod_negotiation.c
Date Tue, 07 Aug 2001 00:50:57 GMT
I agree with you, if no files exist.  Some do, therefore the server is misconfigured, 
or the files shouldn't exist in the first place.

Please see mod_negotation.c line 818 for another example where we blow up with an
error 500 when the negotiation map file is completely empty.  If we are going to
make this consistant, there are some other errors in mod_negotiation that need
some further review.

We might have somebody else even serve one of those files, but the new rules mean
nobody else will get a chance.  That's why I feel it is a 500, but I don't have
a strong enough opinion to argue the point, so the patch is committed.


----- Original Message ----- 
From: "Roy T. Fielding" <>
To: <>
Sent: Monday, August 06, 2001 6:40 PM
Subject: Re: cvs commit: httpd-2.0/modules/mappers mod_negotiation.c

> If the file with the unrecognized extension did not exist, would the
> result be a 404 or a 200?  If it would have been 200, then 500 would be
> a reasonable error response.  If it would have been 404, then 404 is the
> only reasonable response --- an admin can look at the error_log to find
> out why.
> The specific example you described in the commit message should always
> result in a 404.  This is a protocol answer.
> ....Roy
> > I believe we want a 500, here's why...
> > 
> > A 404 error is far more likely to go 'unnoticed', although it's pretty obviously
> > a configuration error (why are we searching for indexes that are misconfigured?
> > If the admin wants to deny access, Options -Includes and leave DefaultIndex out.)
> > 
> > A 500 tell the admin, as well as the user, that we are in trouble, and something
> > needs to be fixed.  If you yell loud enough, the problem gets fixed.
> > 
> > Note this is an unusual error, but people do stumble across it.  A 404 would 
> > minimize the problem too far.

View raw message