Tracking change and innovation in the enterprise software development community
Terracotta:
Scale your
Apps w/ease
Offload the DB:
Learn More!
TargetProcess:
Plan, Track
Succeed with
Agile PM tool.
Free Download!
Testing Tools to
Support Agile
Software Delivery
Whitepaper
Tools to get
Visual Studio
2008 Projects
Under Control
Posted by Stefan Tilkov on Jun 07, 2008 01:48 AM
In this keynote, recorded at QCon London 2008, ThoughtWorks' Chief Scientist and bestselling author Martin Fowler and Global Head of Architecture Jim Webber share their views of the typical corporate ESB, which in their opinion has grown too fat for its own good. Martin and Jim suggest the Web's architecture as a possible and more light-weight alternative, in line with their preference for agile approaches.Create REST-enabled Web Applications
Create a photo album application with Project Zero and REST design principles
Introducing IBM WebSphere sMash, Part 1: Build RESTful services for your Web application
I really enjoyed this one. In QCon london 2007 I really liked the presentation Dan North and Martin Fowler did by then. I consider this one as a continutation!
Doesn't it lead to an ad hoc, spaghetti net of interdependencies between systems? It may work while it's small, but will it still work while it grows? It seems like 'I got it done, it works, but how to manage it in the long term is not my problem'. I don't think big, complex ESBs are the solution, but I'm not sure ad hoc services are, either.
Shhhh... don't let reality interrupt the arm waving ! Seriously though, yes web architectures can scale - but that implies behind the firewall there will have to be web style infrastructure - stuff like DNS servers and so on. The current view of using web based tech in enterprise forgets that the web actually has some pretty powerful "middleware" out there behind the scenes that most of the time Just Works, hence we forget about it. Replicating that in the enterprise would not be trivial.
Absolutely brilliant stuff! I listened to Jim's Guerilla SOA talk earlier this year, and was hoping to find a good online presentation to send around whenever these kinds of issues come up, and now I've got one. Excellent!
Excellent way of presenting the SOA reality, Gureilla SOA seems promising but a practical example of implementing such a Guerilla SOA based system will be very very helpful.
Every one(seem to) understand that pushing the complexity on to web is a good idea but how to do that is the Question? (Answer REST??)
how can I use REST for building Business Processes? I have no idea how to ;) in fact I believe REST was not meant for that. (its ROA not SOA and R may not always S, inturn S can be using many R's, hope you got what I am pointing out!)
Every one say REST!= CRUD they say internet is an example for that,is it justified? what are we doing on internet other than CR ??
Hope I am not taking off this conversation, I felt Gureilla SOA will be a combination of SOA and REST.
Again I am kind off biased with this article about REST style and SOAP Style.
Thanks all, for such a nice community to share the thoughts.
http://schneider.blogspot.com/2008/06/worst-soa-presentation-ive-ever-seen.html
It would be good if Jim and Martin could get their facts right, and not just about historical constructions. Here is the history of the bridge they make fun of: http://en.wikipedia.org/wiki/Pont_Saint-B%C3%A9nezet. The reminder of the presentation is just as atrocious.
Hi Siva, I gave another talk at QCon last year called "A Couple of Ways to Skin an Internet-Scale Cat" which details how to do boring enterprise-y things like workflow using Web technologies. I don't know if that video will ever be released, but you can grab the slide deck from my Web site, http://jim.webber.name/presentations.html Jim
Thank you for your input JJ. I feel it's a badge of honour to receive your validation. Jim
Hi Jim
Yes we have your presentation from QCon San Francisco on the scheduled as will be released the first week of August.
Diana
Thanks for the reply Jim and Diana, looking forward for that presentation.
Having worked in SOA environment for a while, I will have to agree with this presentation completely. Nice work guys! It is very funny too. I am sure that this presentation does not impress the enterprise architects as it threatens thier monopoly. The fact is that it does not have to. I used to be in that camp but not any more. And I am glad that I made the switch to the reality.
hi, maybe Martin and Jim are wrong, We daily need to integrate different kinds of systems, different
technologies...i don't think squid can do that (by the way...i have installed squid ;) ), today bussinesses
are more than a simple WS that responds with data (in any format you like....but data) today bussinesses
need integration stuff that i don't think you can do with a few WS-REST.
My experience integrating different Enterprise Applications is that ESB can do a lot for you, maybe you
can play with incremental services published inside the ESB.
Martin and Jim are like those generals outside reality, please switch to the reality.
Pd: How you can integrate a CICS with a J2EE application? come on...using WS-REST.
R R, thanks for your comment. Would you be nice enough to edit your profile and put in your real name? Although we haven’t published any restriction from using aliases, the intent for InfoQ discussions is to maintain a level of professionalism, in which using real names is an important factor. Thanks!
with a couple of my thoughts: On the use of Squid as middleware: http://www.manageability.org/blog/stuff/cache-tier-architecture On the use of lean methodology for SOA: http://www.manageability.org/blog/stuff/why-soa-2-dot-0 Carlos
This presentation clearly shows Martin is out of touch with the reality and technical trend and does not understand what SOA and ESB really mean and can do to business. The way they have been promoting can only work in a well defined/smaller project, which is not the case in larger business today, which has chaos of mixture of applications and data sources. And also very often, business refuses to accept "small things that work". They want us to deliver entire thing that work and bring the competitive advantages.
The presentation by the Agile gurus Jim Webber and Martin Fowler "Does My Bus Look Big in This?" was very disappointing in my opinion. I watched the presentation hoping to learn some new approaches to SOA. But the presentation says middleware is bad, big SOA projects are bad and Agile is still great so keep buying the books. Unfortunately some SOA projects really are big but SOA experts already know to decompose and iterate them as much as possible. I do agree with their points on iteration and continuous integration, which is fairly well accepted in the SOA community. Their good proxy diagram that replaces the bad middleware diagram this is just semantics. Isn't a proxy middleware too if it does anything to help the developer with integration problems? Also, middleware does not have to be expensive that only "stupid" CIOs buy. There is an open source SOA logo on the screen right behind the presenters. Middleware is becoming a commodity. The problem most companies face is not how to develop new application in an agile fashion. Most IT shops are buying ERP, CRM, SCM packages and maintaining legacy systems. SOA and middleware can help with integration problems and it does not have to be expensive or big bang. This just looks like an attempt to protect their Agile franchise from the threat of SOA. http://blogs.ittoolbox.com/eai/business/archives/gorilla-soa-25448
Great presentation! It's interesting to notice that SOA implementation actually differs quite a lot from he original intentions. I bet big vendors played quite a role in that. To some customers, SOA was a chance to get rid of some nasty vendor lock in, but this turned into an opportunity for the same vendor to feed in some more tools, more or less useful. However the main struggle goin'on is about size. Agile principles aim for small but continuous increment, but SOA is too often sold as a "big plan" that will require "big operations" to be delivered in a "big bang" fashion or require a "big team" or adetailed "top-down" specification. It's somewhat hard to define "big" or "small" in a huge organization, but the way the whole stuff is "sold" really makes a difference. However, nobody wants to turn the ESB into a mess repository, ...but still looks like a good tool to deal with legacy systems.
Agreed.
Related article - mine. How to use ESBs for SOA
Michael Mah analyzes the development process in 5 companies: 2 Agile (one of them BMC) and 3 classic. He presents the factors which contributed to the success of BMC's Agile adoption.
In this interview filmed at RubyFringe 2008, Tom Preston-Werner talks about how both Powerset and GitHub use Ruby and Erlang, as well as tools like Fuzed, god, and more.
David Laribee discusses the purpose of ALT.NET, its mission and future.
Ruby on Rails has become a popular Ruby framework for creating web applications in recent years. An aspect of creating a web application is the need to repeatedly create the same base functionality.
Steven Haines talks about tackling web application performance tuning by proposing a method called wait-based tuning.
Shaw and Fowler talk about the need for a new relationship between the business department and the IT department. Studies have shown that projects mostly fail due to miscommunication between the two.
In this article, Jim Webber, Savas Parastatidis and Ian Robinson show how to drive an application's flow through the use of hypermedia in a RESTful application.
Eccentric artist turned overnight anti-celebrity, Giles Bowkett captures the heart and soul of RubyFringe as he demonstrates his revolutionary Archaeopteryx MIDI drum pattern generator.
20 comments
Reply