Return-Path: Delivered-To: new-httpd-archive@hyperreal.org Received: (qmail 8012 invoked by uid 6000); 8 Aug 1999 21:10:57 -0000 Received: (qmail 8006 invoked from network); 8 Aug 1999 21:10:55 -0000 Received: from ns.skylink.it (root@194.177.113.1) by taz.hyperreal.org with SMTP; 8 Aug 1999 21:10:55 -0000 Received: from kim.ispra.webweaving.org (va-157.skylink.it [194.185.55.157]) by ns.skylink.it (8.9.1/8.8.8) with ESMTP id XAA08148 for ; Sun, 8 Aug 1999 23:09:54 +0200 Received: from kim.ispra.webweaving.org (kim.ispra.webweaving.org [10.0.0.2]) by kim.ispra.webweaving.org (8.8.8/8.8.5) with ESMTP id VAA03641 for ; Sun, 8 Aug 1999 21:10:08 GMT X-Passed: MX on Ispra.WebWeaving.org Sun, 8 Aug 1999 21:10:08 GMT and masked X-No-Spam: Neither the receipients nor the senders email address(s) are to be used for Unsolicited (Commercial) Email without the explicit written consent of either party; as a per-message fee is incurred for inbound and outbound traffic to the originator. Posted-Date: Sun, 8 Aug 1999 21:10:08 GMT Date: Sun, 8 Aug 1999 23:10:08 +0200 (CEST) From: Dirk-Willem van Gulik X-Sender: dirkx@kim.ispra.webweaving.org To: new-httpd@apache.org Subject: patching the sha1/md5/base64 things into 2.0 Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: new-httpd-owner@apache.org Precedence: bulk Reply-To: new-httpd@apache.org Status: O Just to rationalize/clean whilst having a chance; whilst patching in the base64 and sha1 things into 2.0 I'd like to do replace the current base64 to64() encoding in ap_mda5.c (used only for our private 'hack' of the password when crypt() is not there) "./0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"; by the more standard/mime-ish "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; Which is used in various other places for the uudecoding of passwd strings, sha1, MD5's etc. And which is now in ap_base64.c Would anyone object, or did I miss something crucial ? Secondly; as this will break things anyway; would anyone be offended if we made our own MD5 style passwd the default, even if crypt() is on the platform (we still would accept crypt(), like in the last re- hack I did on ap_checkpass.c based on non-prefix). This would make us 'more' compatible between unix/us, windows and some of the cripled unix-es you find in europe (e.g. my SunOs machine here ;-). Please shout ?! I know these are small changes, but there are so many things I'd like to rationalize before 2.0 starts to be carved in stone and becomes a 'beta'. Dw