Related link: http://biz.yahoo.com/prnews/031013/nym082a_1.html
I’m happy for Graham. This gives him a new playground to play in (webM). My guess is that it will take some time and effort to actually combine the technology so that its integrated for real, assuming that’s what they really intend to do. We just finished stage 2 of that process with the eXcelon acquisition that we did about a year ago. The initial integration can be almost instantaneous, because its all standards based, right? However, it takes some time to integrate the incumbent technology with the new stuff in terms of making the coherent experience for the user that doesn’t leave you feeling like you’re using two separate products.
Actually, webM isn’t really “the first” to have a highly distributed standards based integration fabric. The ESB has been providing a “100% standards based” integration fabric for a couple of years now. I’m happy to see other vendors recognizing the need and sharing a vision for a highly distributed integration fabric. A highly distributed integration fabric containing a robust management infrastructure is at the core of the emerging Enterprise Service Bus technology trend. Sonic would like to see Graham and webMethods “get on the bus”.
I run into Graham quite a bit at speaking venues. I had spoken with him about adopting the ESB terminology several months ago. He wasn’t too keen on it largely due to at the time that 1)The definition of ESB requires messaging as part of the core requirement. 2) None of the major vendors had gotten “on the bus” so to speak. Since then IBM has started adopting the term ESB. Perhaps he should revisit issue #1.
Dave
anonymous
>I'm happy for Graham. This gives him a new playground to play in.
What are you talking about? Who is Graham? Are we supposed to know? I just reached this site and yours was the first link I clicked on. Do you think you are talking to yourself and nobody else? A bit parochial are we? Please give an introduction so readers can know what you are talking about. Thanks.
Huh?
>I'm happy for Graham. This gives him a new playground to play in.
What are you talking about? Who is Graham? Are we supposed to know? I just reached this site and yours was the first link I clicked on. Do you think you are talking to yourself and nobody else? A bit parochial are we? Please give an introduction so readers can know what you are talking about. Thanks.
Read the Linked Article
All is revealed there. ;)
Eh?
I thought you were talking about Paul Graham of LISP fame. Who's this Graham Glass hosehead?
Stump sold out!
I read the article, and I still don't know why anyone would expect me to know who's referred to by the title of this blog entry... About the only first name you could do that with would be "Bill".
Stump sold out!
The title of the blog is a link that points to an article that explains who Graham is in a fair amount of detail. I suppose I could have been more precise and said "Graham Glass". I actually made an assumption that most people who would come to the O'Reilly site would know who he is. He is a fairly well known public figure.
If this were an actual "article", then I probably would have taken the time to make sure everything was explained a little better. But that's the nature of blog postings. They can be whatever you want them to be, and people can choose to read them or not. Sorry if I caused you to get agitated over my lack of clarity.
Dave
Dear anonymous
Shut up.
Yours,
Anonymous.
Sonic pitch
Nice pitch for Sonic. As long as one considers EAI or a "highly distributed standards based integration fabric" to solely be a message bus, Dave's argument holds water. Once one strays from that limited interpretation, Dave's argument stumbles. I find this limited interpretation to be a source of constant problems in the integration business. Too many vendors and deployments fail in this market because of it. It's nice that Sonic still has "the fastest JMS provider in the west" to keep it's cash flow running.
Sonic pitch
I'll agree with you on that "fastest JMS provider in the west", and I'll reinforce your assertion that we have cash flowing in as a result of that.
I'll even agree that a highly distributed integration fabric requires more than just messaging. The ESB is way more than messaging. Messaging is definitely a key component. The ESB also means Web services. Its transformation and routing. Its adaptors to packaged and legacy applications. Its integration with appservers. Its orchestration. Its XML caching, aggregation, and querying. Its a new model for integration that incorporates all these concepts in a highly distributed fabric based on SOA, and a specialized container that's built for housing integration components. Its about a new type of IDE built for the "Integration Architect" (you). Here's a great article that talks about the ESB concept in more detail.
Related question...
What is "the ESB"?
Related question...
The term "ESB" stands for "Enterprise Service Bus". ESB is a technology trend that is being viewed as the next generation for EAI. The ESB provides an architecture for integrating applications and services into a fabric that spans the extended enterprise in a large scale fashion. The ESB makes Web services, XML, and other integration technologies immediately useful with mature technology that exists today.
The ESB combines enterprise messaging, Web services, XML transformation, and intelligent routing of data based on content. The ESB provides a loosely-coupled highly distributed approach to integration. An ESB can utilize J2EE components such as the Java Message Service (JMS), J2EE Connector Architecture (JCA), and also integrates nicely with applications built with .NET, COM, C#, and legacy C/C++. These standards are put together in a meaningful way that defines an open-ended pluggable architecture that supports both industry-standard integration components as well as proprietary elements.
The ESB is based on today's established standards, and has real implementations that have been shipping for at least a year. Yet it is also well suited for the evolving standards of tomorrow, including all of the WS-* specifications.
The ESB provides integration broker capabilities, yet does it in such a way that is much more distributed, in a buy-what-you-need and deploy-what-you-need fashion. Beyond the base capabilities mentioned so far, an ESB can provide more advanced features like XML persistence, caching, and aggregation, an IDE for creating and debugging XSLT transformations, and advanced orcestration services. For more details, I would suggest you look at this article that was recently published in Business Integration Journal - http://www.bijonline.com/PDF/Sep03Van%20Huizen.pdf
Hhmmm...Perhaps ESB should be the topic of my next book.
Dave
Sonic pitch
If the ESB is more than just messaging, and represents your integration approach, then you \really\ can't criticize Graham Glass in the sense that you have.
On the other hand, it is comforting to know that, given this, Sonic doesn't hold the same stance as certain other enterprise software companies, which is "integration = message bus, and nothing more" (which is an extremely limited vision). Along with webMethods, IBM, etc., this provides us (integration architects) with several viable solutions to choose from.
Sonic pitch
If you're referring to my "criticism" being that I commented about Graham not adopting the term ESB - then that's now a moot point. webM has recently adopted that term.
Regarding Sonic's stance on integration - You're absolutely correct. You are raising an excellent point about a common misconception regarding the use of the term "Bus" as it pertains to integration. In addition to redefining what integration means, we are redefining what people think of when they hear that word to mean not just a "Message Bus", but an "Enterprise Service Bus". Messaging is an important part of an ESB, but only one of the enabling technologies.
Dave