Return-Path: Delivered-To: new-httpd-archive@hyperreal.org Received: (qmail 8408 invoked by uid 6000); 19 Jan 1999 11:24:51 -0000 Received: (qmail 8401 invoked from network); 19 Jan 1999 11:24:49 -0000 Received: from mrelay.jrc.it (139.191.1.65) by taz.hyperreal.org with SMTP; 19 Jan 1999 11:24:49 -0000 Received: from jrc.it (elpc51.jrc.it [139.191.71.51]) by mrelay.jrc.it (LMC5692) with ESMTP id MAA00870 for ; Tue, 19 Jan 1999 12:26:17 +0100 (MET) Message-ID: <36A46AA6.133D42A@jrc.it> Date: Tue, 19 Jan 1999 12:21:10 +0100 From: Dirk-Willem van Gulik Organization: ISIS/STA - Joint Research Center of the European Commission X-Mailer: Mozilla 4.5 [en]C-CCK-MCD {PEN/1.00} (Win95; I) X-Accept-Language: en,zh,de,fr,sv MIME-Version: 1.0 To: new-httpd@apache.org Subject: Re: Hosting case-sensitive httpd on case-insensitive filesystem References: <199901182154.NAA07286@scv2.apple.com> Content-Type: multipart/mixed; boundary="------------04C694DF142A641E0479A864" Sender: new-httpd-owner@apache.org Precedence: bulk Reply-To: new-httpd@apache.org This is a multi-part message in MIME format. --------------04C694DF142A641E0479A864 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Wilfredo Sanchez wrote: > > What sorts of problems should I expect to run into if I run a > case-sensitive (Unix) httpd on a case-insensitive (HFS+) filesystem? > I be looking at the code in a minute, but I'm wondering if people > who've dealt with it might offer some insight. > > I've been talking to the kernel guys about detecting per-directory > what filesystem is underneath and they say no problem, but I doubt the > answer will be very portable, though it would possibly be portable to > other BSD's or easy to add to other BSD's. When I have something that > works, I'm bring it up again so we can figure out what the right thing > to do is. > > In the meantime, we compile with the assumption of having the docroot > on UFS, but people want to be able to move the docroot to an HFS+ volume > and that makes me nervous. Should it? Id' say that you could 'avoid' making it OS specific; (and in fact it would be usefull say on a BSD machine with a dos/fat16/fat32 partition. You'd be looking at a 'comapre case(in)sensitive flag' on a per directory basis (which would also cover the