Received: (from majordom@localhost) by hyperreal.com (8.8.5/8.8.5) id LAA13841; Mon, 16 Jun 1997 11:18:04 -0700 (PDT) Received: from twinlark.arctic.org (twinlark.arctic.org [204.62.130.91]) by hyperreal.com (8.8.5/8.8.5) with SMTP id LAA13831 for ; Mon, 16 Jun 1997 11:18:01 -0700 (PDT) Received: (qmail 3503 invoked by uid 500); 16 Jun 1997 18:25:25 -0000 Date: Mon, 16 Jun 1997 11:25:25 -0700 (PDT) From: Dean Gaudet To: new-httpd@apache.org Subject: config idea, scoping 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 We have various modules which define essentially the same directives but with slightly different names. Such as AuthDBFile and AuthDBMFile. This has always seemed a bit messy to me. It also means it's somewhat hard to avoid name-clashes. What if we required scoping like this: File foobar Group blah This allows module writers to standardize the config directives for things like auth modules. It also makes it clear which module is being configured. But it makes the config grow in size. And another misc idea is: default When the config parser sees the keyword default it calls create_dir_config for the module, to reset the current per_dir_config for that module. This is useful in turning off features at lower levels of directory trees or whatever. But this is only a half solution for a general problem that we have: Not all config directives are reversible. Oh yeah we also need to do a systematic renaming of directives such as MaxClients, MinSpareServers, MaxSpareServers, MaxRequestsPerChild. We're calling the children three different things there. We should stick to one name, and since "server" is already overloaded so much, and so is "child" I'm thinking that "worker" would be an appropriate name (it also takes into account threading). Dean