From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261733AbTLIVsg (ORCPT ); Tue, 9 Dec 2003 16:48:36 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262098AbTLIVsg (ORCPT ); Tue, 9 Dec 2003 16:48:36 -0500 Received: from ipcop.bitmover.com ([192.132.92.15]:62876 "EHLO work.bitmover.com") by vger.kernel.org with ESMTP id S261733AbTLIVsc (ORCPT ); Tue, 9 Dec 2003 16:48:32 -0500 Date: Tue, 9 Dec 2003 13:48:15 -0800 From: Larry McVoy To: Zwane Mwaikambo Cc: Larry McVoy , cliff white , hannal@us.ibm.com, lse-tech@lists.sourceforge.net, linux-kernel@vger.kernel.org, Amy Graf Subject: Re: [Lse-tech] Re: Minutes from OSDL talk at LSE call today Message-ID: <20031209214815.GA32633@work.bitmover.com> Mail-Followup-To: Larry McVoy , Zwane Mwaikambo , Larry McVoy , cliff white , hannal@us.ibm.com, lse-tech@lists.sourceforge.net, linux-kernel@vger.kernel.org, Amy Graf References: <189470000.1070500829@w-hlinder> <20031204033535.GA2370@work.bitmover.com> <20031204134517.0c7a4ec4.cliffw@osdl.org> <20031204234454.GA15799@work.bitmover.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Dec 09, 2003 at 04:26:45PM -0500, Zwane Mwaikambo wrote: > On Thu, 4 Dec 2003, Larry McVoy wrote: > > > > The first is triggers. The Mozilla tinderbox is driven by triggers from > > > CVS commits. I believe that triggers are resevered for the commercial > > > version of BK. > > > > That's not true. Trigger support is identical in both versions. > > Perhaps the FAQ may need updating then; > > http://www.bitkeeper.com/Documentation.FAQS.Event.html Indeed. It's worth pointing out that triggers in open source trees are quite a bit more dangerous than in controlled environment. Carl-Daniel's boss made quite a fuss over the fact that triggers are just programs that are run and can be used to cause all sorts of problems if people were malicious. I've toyed with the idea of disabling triggers in openlogging trees because of this. I'm neutral on the topic, it's not like triggers are some huge money maker that we need to reserve for the commercial version. If the general feeling is that triggers are useful and people will take responsibility for policing their own repos then we'll leave them in. On the other hand, if someone putting a nasty trigger into your tree somehow becomes the fault of BitMover because we provided the infrastructure then out they go in the next release. -- --- Larry McVoy lm at bitmover.com http://www.bitmover.com/lm