From new-httpd-owner-new-httpd-archive=hyperreal.org@apache.org Tue Apr 21 15:46:55 1998 Return-Path: Delivered-To: new-httpd-archive@hyperreal.org Received: (qmail 21878 invoked by uid 6000); 21 Apr 1998 15:46:53 -0000 Received: (qmail 21871 invoked from network); 21 Apr 1998 15:46:50 -0000 Received: from gensym.com (192.156.185.2) by taz.hyperreal.org with SMTP; 21 Apr 1998 15:46:50 -0000 Received: by gensym.com (4.1/SMI-4.1) id AA21917; Tue, 21 Apr 98 11:46:45 EDT Received: from unknown(1.0.2.6) by ftp.gensym.com via smap (V1.3) id sma021914; Tue Apr 21 11:46:15 1998 Received: from siam.gensym by gensym1.gensym.com (4.1/SMI-4.1) id AA16394; Tue, 21 Apr 98 11:46:14 EDT Received: by siam.gensym (SMI-8.6/SMI-SVR4) id LAA19282; Tue, 21 Apr 1998 11:46:14 -0400 Date: Tue, 21 Apr 1998 11:46:14 -0400 From: bhyde@gensym.com (Ben Hyde) Message-Id: <199804211546.LAA19282@siam.gensym> To: new-httpd@apache.org Subject: Re: [PATCH] sections Sender: new-httpd-owner@apache.org Precedence: bulk Reply-To: new-httpd@apache.org Brian Behlendorf wrote: > I am against designing a programming language piece by piece. ... > I believe the configuration file syntax should avoid being procedural > ... A procedural configuration file would also become very difficult to > represent in a GUI. ... I think for 2.0 we should examine moving the > config language to something like XML. Sentiments I agree with. > I think the config engine should be frozen for the rest of 1.3 to allow the > GUI efforts to grow ... Yes. The only reason to depart from that plan would be, for example, discovering a requirement to make dynamicly loaded modules work out. Recalling that Ralf's example of what IfDefine is good for was toggling the loading of modules, I happily note that is already possible with -C. Ben Laurie wrote: > R of Unusual Size wrote: > > Hmm. That's not how I interpreted it. A separate namespace, > > yes - but definitions set on the command line, not by modules. > > E.g., "httpd -Dsym1". > You misunderstand, I think. What he wants is for modules to push a > symbol into the namespace, too, so instead of: Yes, I prefer single set of symbols that controlling which parts of the configuration data are considered applicable, not two and not the one IfModule is currently using. - ben hyde