httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "William A. Rowe, Jr." <>
Subject RE: [inq] short, concise, and to the point
Date Tue, 10 Oct 2000 00:09:36 GMT

We are 100% on the same page here.  1.3.13 will roll on the 10th
and be released on the 13th.

But I did a search on 'case' through RFC2616... the only thing applicable
to our debate follows:

   When comparing two URIs to decide if they match or not, a client
   SHOULD use a case-sensitive octet-by-octet comparison of the entire
   URIs, with these exceptions

I've read it over and over, and the only absurd aspect is that an author
who creates the HREF /testit.html in one case and /TESTIT.html in another
is that it is not cached (if the browser followed the client rule above).
Please refer me to the right RFC and paragraph (assuming for the moment
that I'm blind).  If there is no citation, we can proceed with the argument
that <Apache's URIs> are case sensitive, and I'm still with you.

Seriously, the -other- server (in the sense of 10%+ market penetration) 
definately does not follow the behavior you advocate.  This makes porting 
webs from Apache to the other server trivial, but from the other to Apache 
is not.  And the redirects you advocate destroy the POST body (internally, 
for certain... I'm not sure about POSTS redirected via the client.)

I want both facilities.  A case-neutral option for NTFS file systems and
their administrators who don't care, and an case-sensitive option for
the rest of us :-)  URL redirect is optional, based on another security
option (that some administrators would refuse to enable, for good reasons.)
But short names are moot, and will be resolved away in all cases.

View raw message