I've just recently decided to get started using Twitter, which I don't completely understand yet, but most likely because most of the users who are following me are either looking to get married in order to leave the Ukraine, or begging for credit card information in order to get their favourite crippled, blind, grandmother a new hip. Not being overly interested in crippled blind people, or getting married to a stranger, I don't really have much to do on Twitter.
I just found out about the keyword searches on it and have realised that people write on twitter using hashes and different keywords to help organise subjects, so of course did a quick search on #km to see if I could find some people to follow. One of the first things I found was the following link. (Don't worry, it's not the blind, crippled grandmother) I did a quick scan of the first sentence and thought it looked interesting, then read the paragraph starting "BTW, forget about 'enterprise architecture' or SOA....... ". What frustrates me here is trying to discredit SOA and Enterprise architecture by pretending they meant something else to begin with. SOA(Service Oriented Architecture) and Enterprise Architecture are most definitely terms used in IT, and should only be used FOR IT. They are terms used when one star trek loving, comic con attending individual speaks to another from behind their extremely large calculator and pocket protector. (No offense meant I am one of those individuals by the way) SOA and Enterprise Architecture are used to describe how software, any software, is built, they are terms that instantly allow those who speak Klingon to imagine classes, objects, layers, tiers, and databases without having to read pages of documents describing it all. It is jargon and never during it's history was it ever directly related to KM or KM software. Using the term SOA is no different from using the term mashup and whether you are using software designed with Enterprise level architecture in mind, or using software that has been created as a mashup does not tell you how good or bad the software is. It does not tell you that the software will or will not work in the business area it is meant to. It will tell you however, how scaleable the software is, how flexible it is, how quick and easy it would be to change the software as the world and the demands change around the software. To simply forget about patterns that work is a mistake. Software in the KM area needs to be able to change, needs to be able to bend to demand, and needs to be scaleable in order to deal with the world and the problems thrown at it. Using a mashup is absolutely fine, so long as the mashup itself has been built correctly, has been thought through correctly, and guess what, usually that means building the mashup on Enterprise Architecture.
Comments (1)
Hi Ken, Some very good points here. Of course SOA has become a very big business issue too and on the agenda of many business oriented conferences and many business focussed research projects (e.g. the European Nessi project http://www.nessi-europe.com/).
There is good reason for this and you say it all in the blog, which I quote with a very small modification. "Software needs to be able to change, needs to be able to bend to demand, and needs to be scaleable in order to deal with the world and the problems thrown at it."
There is one more huge business imperative that is driving this that is also applicable to KM. This is run-time composition of software service level components even where at design time those components are unknown. And in the business focussed SOA research this is because businesses want software to be composed across organisation boundaries (and by implications platforms/languages etc.) because in a more dynamic world organisations will be less rigid in the manner in which they contract and deliver services and products and will thus need software that can interact in this way.
An example from a NESSI research call is
"SOA4All is a NESSI Strategic Project with semantic technology at its heart. SOA4All's stated objective is to realise a world where billions of parties are exposing and consuming services via advanced web technology: while service orientation is widely acknowledged for its potential to revolutionize IT by abstracting from the underlying hardware and software layers, its success depends on resolving challenges in the areas of scalability, dynamicity, heterogeneity, distributivity and openness not handled by today's SOA technology. These are the challenges SOA4All will address." (http://www.nessi-europe.com/WorkingGroups/HorizontalWorkingGroups/SemanticTechnologies/tabid/241/ItemId/412/Default.aspx)
Posted by Peter Stanbridge | December 30, 2009 7:30 PM
Posted on December 30, 2009 19:30