From omg-list-errors@emerald.omg.org Thu Nov 28 09:16:11 2002 Received: from escher.cs.uml.edu (escher.cs.uml.edu [129.63.16.205]) by saturn.cs.uml.edu (8.11.6/8.11.6) with ESMTP id gASEGB2263967 for ; Thu, 28 Nov 2002 09:16:11 -0500 (EST) Received: from emerald.omg.org (emerald.omg.org [192.67.184.65]) by escher.cs.uml.edu (4.7.0.120) with ESMTP id for ; Thu, 28 Nov 2002 09:02:55 -0500 (EST) Received: from hobbit.omg.org (hobbit.omg.org [192.67.184.3]) by emerald.omg.org (8.11.0/8.9.2) with ESMTP id gASE0X317661; Thu, 28 Nov 2002 09:00:33 -0500 (EST) Received: from rwcrmhc51.attbi.com [204.127.198.38] by hobbit.omg.org asmtp(1.4f) id 4757; Thu, 28 Nov 2002 09:05:02 -0500 (EST) Received: from spectre (h0020781023b7.ne.client2.attbi.com[24.218.154.177]) by rwcrmhc51.attbi.com (rwcrmhc51) with SMTP id <2002112814010705100agm4ee>; Thu, 28 Nov 2002 14:01:07 +0000 From: "Peter J. Fontana" To: "Daniel Duffy" , "BERRISFORD, Graham" , , , Subject: RE: Issues with UML2 support for MDA Date: Thu, 28 Nov 2002 09:00:27 -0500 Message-ID: X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) In-Reply-To: <2925F4D053DFBE44BC02176424D719BC023FFB@dsserver2.datasim.nl> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Importance: Normal MIME-Version: 1.0 (Generated by Clearswift ES version 4.7.0.120) Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Status: X-Keywords: X-UID: 211 Status: RO Hi Daniel - > Speaking as someone with 15 years OO experience (and 10 in other > areas!) I have now come to the end of the road as far as OO is > concerned as a medium to capture high-level requirements and/or > architectures. Others in the past have realised that cohesion, > coupling, information hiding and decomposition (Jackson, Yourdon, > Aristotle, Julius Caeser (divide and conquer) just to name a few) > are essential. These are lacking in OO work. Very unfortunate, > because we alienate the group of non-OO people. You remarks are troubling to me, because there are techniques (like Domain Modeling) that can break up a large and complex overall problem into "logical components". With domain modeling you increase cohesion between the classes within a domain, substantially reduce coupling between domains, exercise information hiding (at the domain boundary), and of course domain are the unit of decomposition. > Objects are close to design and implementation. I use them on a > daily basis during detailed design. They are not suitable in the > early phases of software development. For the typical embedded system we work with, our clients use of UML for about 50 - 75% of the overall system (the remainder is legacy code and 3rd party components). ALL of this UML work is in analysis modeling (PIM development). These analysis models are executable, and they automatically translate their deployed implementation code from them, but they are still analysis models - independent of any specific implementation. > If anyone would like to > tell me what I am missing or do not see, please feel free to tell > how for exampl to do high-level analysis with objects. Apologies to all for the evangelical tone, but if I've stuck a chord with you please grab a white paper we have on our web site: "Durable Embedded System Architecture via Domain Modeling" from www.pathfindersol.com/papers.html. _________________________________________________ Pathfinder Solutions www.pathfindersol.com We Make UML Work. 888-662-PATH Peter J. Fontana voice: +01 508-384-1392 peterf@pathfindersol.com fax: +01 508-384-7906 _________________________________________________