http://ayende.comAyende @ RahienCopyright (C) Ayende Rahien 2004 - 2021 (c) 202660This is somewhat similar to something I've been banging on about for a while. why is it that some software needs an endless amount of testing, and the only way to cope with this expanding burden of testing is to automate, automate, automate. But some systems don't need it. I think the essence is "..if the problem is neat the code can be neat as there are no special cases. And neat code is good because it is easy to test and easy to review, and that means that the implementation quality can be very high;" it also means that number once the code works at all then system integration tests are sufficient. The rest is here: [nomorehacks.wordpress.com/.../building-stable-a...](http://nomorehacks.wordpress.com/2008/07/17/building-stable-applications/) http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment13http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment13Fri, 09 Jan 2009 11:26:36 GMTSteve, If you take the Prism guidance, and apply it to a non UI system, you get a lot of that behavior. I just finished recording an hour talk with Luke about this topic, and I would publish it in a few days. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment12http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment12Wed, 24 Dec 2008 06:30:57 GMTYou should deprecate the other post - this one says it much better. It is sad that composite architectures are so little-known, and so terribly tooled. Besides the obvious dependency inversion stuff, is there some toolset that anyone recommends for achieving composite architectures? Challenges I have encountered are security authorization (& authentication sometimes), UI merging (how do pieces of the UI from different components integrate) and module communication (how do modules communicate in a low-friction manner). http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment11http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment11Tue, 23 Dec 2008 19:07:03 GMTI think one thing that may be missing from this discussion is the power of great leadership. I have seen projects lead by various levels of leaders and at the end of the day I believe 100% that a successful project needs a great leader. And by successful I don't mean it shipped because I have been on projects that "shipped" but I didn't consider anywhere near successful in terms of creating a maintainable, trust able application that fulfilled the customers real needs. In this example it seems obvious that this open closed architecture was Oren's idea and without him the project would have gone in a more normal direction. I suspect that this more normal direction which would have resulted in a project that was much less successful. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment10http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment10Tue, 23 Dec 2008 16:26:37 GMTCan you give me a starting point? From quickly running this search, one didn't stand out: [www.google.com/search?q=JFHCI+site:ayende.com](http://www.google.com/search?q=JFHCI+site:ayende.com)http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment9http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment9Tue, 23 Dec 2008 16:00:13 GMTLuke, See my posts about JFHCI for more details. Otherwise, I would like to get a list of concrete questions that I'll be able to work from. Can you compile one? http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment8http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment8Tue, 23 Dec 2008 15:56:44 GMTCould you write a bit more to combat that very common critique? I mean, yes, you were able to get it to work and you're very smart and also _experienced_. A bit more discussion would, I think, benefit many of your subscribers. "It worked for me" is not so useful. I know you've written a bit about this stuff before, but at least I would appreciate something more concentrated. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment7http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment7Tue, 23 Dec 2008 15:54:27 GMTOh my, not at all. I used a similar approach in another project, where the team that I had was of integrators, not developers. It was very successful (given my starting point), and it served to further validate this approach. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment6http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment6Tue, 23 Dec 2008 15:52:28 GMTHahahaha, I was too succinct for Oren. This must be a record, no? You must have seen this critique coming: "The only reason your technique worked was because you were working with smart people. That doesn't work for most folks." My running hypothesis on this is that most people are being _held back_ by technology and management. I think most people could write better code (and do better architecture) if it required less of their energy to do so (some of us just have more energy to put into computers than other). I think it's possible to lower the energy requirements and it seems that you do, too. However, I would like to hear you talk about this a bit. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment5http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment5Tue, 23 Dec 2008 15:43:39 GMTLuke, I am not sure that I understand what you mean by the question. I had a good team in that project, talented, passionate and smarter than me in actually applying the knowledge in ways that I wouldn't have thought of. They didn't have as much breadth of knowledge, perhaps, and they didn't have experience with unit testing, though. http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment4http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment4Tue, 23 Dec 2008 15:32:53 GMTAyende, about tests, your last paragraph says all! I can't agree more! http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment3http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment3Tue, 23 Dec 2008 14:56:08 GMTOren, I'd like to give you an [American?] high five for that post. Someone rubbed me the wrong way about the unit testing mantra from the beginning and you're providing me support -- not just inane ramblings (not _all_ discussions of unit testing are inane), about the benefits without no solid evidence. Now let me ask you something: to the extent that you can comment on this, what characteristics did your approach require that your team of devs had? I could pull out the word "Mort" and ask you to relate, but just use your own words -- they're usually plenty succinct, and I can often understand the first time around. :-) http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment2http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment2Tue, 23 Dec 2008 13:18:41 GMTHi Ayendy, I’m fully agree with your conclusion that “unit tests are not silver bullet”, in IT in generally we don’t have “silver bullets” (or in mostly cases we don’t have) we just do some practices which can minimize our “pain”, it doesn’t necessary mean that will help us to deliver, increase quality etc.. Because there are a lot of other factors which should be considered, these practices just can increase our probability to deliver (and in mostly cases by practice, it does). http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment1http://ayende.com/3763/composite-architecture-the-open-close-principle-as-applied-to-system-architecture#comment1Tue, 23 Dec 2008 10:22:56 GMT