Return-Path: Delivered-To: apmail-jakarta-avalon-dev-archive@apache.org Received: (qmail 75601 invoked from network); 13 Jan 2003 16:52:19 -0000 Received: from exchange.sun.com (192.18.33.10) by daedalus.apache.org with SMTP; 13 Jan 2003 16:52:19 -0000 Received: (qmail 9337 invoked by uid 97); 13 Jan 2003 16:53:37 -0000 Delivered-To: qmlist-jakarta-archive-avalon-dev@jakarta.apache.org Received: (qmail 9302 invoked by uid 97); 13 Jan 2003 16:53:36 -0000 Mailing-List: contact avalon-dev-help@jakarta.apache.org; run by ezmlm Precedence: bulk List-Unsubscribe: List-Subscribe: List-Help: List-Post: List-Id: "Avalon Developers List" Reply-To: "Avalon Developers List" Delivered-To: mailing list avalon-dev@jakarta.apache.org Received: (qmail 9275 invoked by uid 98); 13 Jan 2003 16:53:35 -0000 X-Antivirus: nagoya (v4218 created Aug 14 2002) X-Injected-Via-Gmane: http://gmane.org/ To: avalon-dev@jakarta.apache.org Path: not-for-mail From: Leo Simons Subject: thoughts on gump & Avalon & dependencies Date: Mon, 13 Jan 2003 17:53:54 +0100 Lines: 48 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@main.gmane.org User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2.1) Gecko/20021130 X-Accept-Language: en-us, en Sender: news X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N Hi peeps, Bit of a rant below, but I felt a need to explain my comment about avalon being the gump problem child. If you haven't been ignorant about gump (like I was), don't bother reading on as you'll already know this :D There's a few packages @ avalon which cause a lot of gump to fall down when they fail to build. The obvious example is avalon-framework. If it fails, almost every other gump target wrt avalon cvs fails as well, and that leads to many more projects failing, like james, cocoon, jmeter, ojb, and lots more. Another such package is logkit. You can look at the ordering of the projects at http://cvs.apache.org/builds/gump/latest/ to figure out roughly how vital a link in the dependency process a particular package is (it shouldn't surprise anyone that ant and xerces lead the way). Quite a few of the excalibur packages are high on the list as well. One that has been failing over the last few days a few times was AltRMI (because Paul's been doing refactoring. The gump failure was corrected, then more changes were introduced, and it broke again, fixed again, broken again). It was interesting to see what happened (I've been closely watching the process). Among the packages that fail to build because AltRMI doesn't build are most of excalibur, fortress, phoenix, cocoon, large parts of turbine, scarab, and more. I think that's a lot! Things are qoing rather well (Paul's been good at catching the gump errors as he refactors, and doing stuff about it), so it is a mild annoyance to all those other projects. Imagine if this goes on for weeks (as has been the case for a large part of excalibur for some time now). Not only are no nightly builds then available, it also means that as the build fails early on in the chain, it is not possible to find out why it might fail later, which can cause incompatibilities to arise between, say, turbine & scarab, which can go by unnoticed. Imagine how much worse the situation gets when people do what we suggest (ie don't worry too much about a dependency on avalon) but the builds keep failing. Large parts of jakarta wouldn't build very well anymore. That is why gump builds should always be a priority to any project that takes itself seriously. cheers, - Leo -- To unsubscribe, e-mail: For additional commands, e-mail: