Return-Path: Delivered-To: apmail-cocoon-dev-archive@www.apache.org Received: (qmail 55739 invoked from network); 29 Aug 2003 11:41:50 -0000 Received: from daedalus.apache.org (HELO apache.org) (208.185.179.12) by minotaur-2.apache.org with SMTP; 29 Aug 2003 11:41:50 -0000 Received: (qmail 712 invoked by uid 500); 29 Aug 2003 11:41:38 -0000 Delivered-To: apmail-cocoon-dev-archive@cocoon.apache.org Received: (qmail 635 invoked by uid 500); 29 Aug 2003 11:41:37 -0000 Mailing-List: contact dev-help@cocoon.apache.org; run by ezmlm Precedence: bulk list-help: list-unsubscribe: list-post: Reply-To: dev@cocoon.apache.org Delivered-To: mailing list dev@cocoon.apache.org Received: (qmail 542 invoked from network); 29 Aug 2003 11:41:36 -0000 Received: from unknown (HELO stud4.tuwien.ac.at) (193.170.75.21) by daedalus.apache.org with SMTP; 29 Aug 2003 11:41:36 -0000 Received: from student.tuwien.ac.at (chello212186106081.14.tuwien.teleweb.at [212.186.106.81]) by stud4.tuwien.ac.at (8.9.3 (PHNE_26304)/8.9.3) with ESMTP id NAA15003 for ; Fri, 29 Aug 2003 13:25:37 +0200 (METDST) Message-ID: <3F4F3831.2060209@student.tuwien.ac.at> Date: Fri, 29 Aug 2003 13:25:37 +0200 From: Andreas Hochsteger User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; de-AT; rv:1.4) Gecko/20030612 X-Accept-Language: de-at, de, en MIME-Version: 1.0 To: dev@cocoon.apache.org Subject: Re: [RT] Cocoon Distro support References: <3385.202.187.40.8.1062137538.squirrel@www.hedhman.org> In-Reply-To: <3385.202.187.40.8.1062137538.squirrel@www.hedhman.org> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N X-Spam-Rating: minotaur-2.apache.org 1.6.2 0/1000/N You've read my mind, didn't you? ;-) Seriously, I think you've brought many things down to the point. Specifically the security issues you've mentioned were given too less attention before. Good work, I like it! -- Andreas Niclas Hedhman wrote: > Hi all, > > the recent comments about production build and "binary builds" have > caused me to think about Cocoon Distros and what must core Cocoon do > to allow for Cocoon Distros from 3rd parties. > > Since this overlap quite a degree with the average Cocoon user's > need of "upgrade running systems", I think it is desirable to start > a discussion. > > Currently I see these hurdles; > * Block Modularization. > * cocoon.xconf Management > * Sitemap Management. > > > Block Modularization. > This is an on-going effort, but I have a few things I would like to > clarify. > Blocks are NOT ONLY CODE! IMHO, Blocks are the traditional JAR based > resources (classes, bundles, etc), but more importantly(!) > documentation of the block, meta-data (see Avalon discussions), > examples, source code (optional) and so on. And all of that should > be wrapped in a single "container"... > That container should be "sealed" and I shouldn't need to know > anything about the internal content, and if Cocoon needs to expand > it, it can do so anywhere with my knowledge about it. > Internally there is a default configuration file, but by putting an > external config file at the same place, it will override the > defaults, preferably merged. > Security is another concern. Do I trust any blocks? NO, therefor the > security policy is declared per block externally, but the default in > Cocoon should be pretty restrictive (like no write or network > permissions). > > cocoon.xconf > Once the Block Modularization is in place, it will have addressed > these aspects, and very little I hope will remain in this file. > So if all Block related stuff is removed, and we are down to Core > funcationality, I think a Distro maker can handle this file fairly > easily. > > Sitemap Management > I would like to see a minimalistic sitemap, basically only > containing the "AutoMount" of all directories' sub-sitemaps. > The question is whether the component declarations should be in the > main sitemap or not. I think maybe not. > The main argument here is "Start simple, expand on demand". The > elaborate main sitemap today can still be around as a sub-sitemap in > the "elaborate/" directory... > Block Modularization probably need to address sitemap declarations > as well, with a formal way of defaults being added automatically and > other sitemap declaration should perhaps be adjacent to the Block > and not to the sitemap, after all keeping the Block stuff at the > Block makes more sense to me. > > > Bottom line; Cocoon is about to broken into smaller, more managable > pieces. This will easier allow 3rd parties to package Cocoon into > nice binary distros. > > > Comments? > > Niclas >