Today, the Annual International Software Development Conference held in London by InfoQ and TRIFORK finished. It is, probably, too early to report how many developers came to listen the presentations, but the keynote on the conference opening gathered a few hundreds people and only 42 per cent of then came from the UK; others - came from all over the world.
I talked about the conference tracks before. Now, I think, it would be interesting to share a few of my notes,particularly, about the Keynote, and, as you have, probably, guessed already, about SOA in the conference - the subject I've spent the most of time on.
Sir Tony Hoare, who devoted his academic researches of almost 30 years 'to the pure aspects of the science of computer programming', gave the opening keynote. He described relationship between the roles of Scientist and Engineer in the programming field. At one point of his presentation, he listed qualities and characteristics of both roles against each other and commented that, in his mind, the characteristics of a Scientist were more applicable to men while the characteristics of an Engineer were more applicable to women. It appeared that typical characteristics of a SW Architect were almost 50-50 per cent mixture of both lists. At this moment I finally understood why our SW Architects could never mature. Indeed, when a baby grew up, she or he became an Engineer or a Scientist...
The second keynote 'Transforming Software Architecture with Web as Platform' was literally fired by Dion Hinchcliffe. The level of assurance and energy demonstrated by Dion was amaising. However, the arguments about, e.g., Cloud Computing (CC) did not convince many in the audience even when he declared that CC could provide reduction of IT cost to the level of magnitude. The following questions in Q&A illustrated that SW developers did not really believed in that but rather shared several business concerns about CC economy that I mentioned in my previous post. Unfortunately, the response to such questions was referring to the 'specialists' who supposed to know how to add security, transparency, flexibility, etc. Did Mr. Hinchcliffe calculate how much those 'additions' would cost? I doubt.
All together, this reminded me the case where one company decided to move its IT to India on the basis that the cost for the company of one SW developer in India was estimated 7 times less than in the UK. Those estimates, actually, counted all SW and HW licenses cost, all training cost, all business analysis and architecture cost into the cost of a UK developer. Moreover, nobody counted the cost of re-development and bug-fixing performed exclusively in the UK (it was too long and costly to explain the errors and bugs over the time zones) as well as nobody counted hours and hours spent by senior management (paid big money) on the phone when bugs were identified and mitigating development had to be set up. Those who tried to re-calculate the estimates including all mentioned factors and some others told me that the cost of the developers in both countries was almost the same...
Finally, let me tell you what impression I had from the 'SOA in Real World' track hosted by Stefan Tilkov, the only SOA dedicated session in the conference. The track question was whether SOA is really dead. However, the speakers did not try to answer this question but jumped on the related topic - SOA did not die (fortunately) but, as OASIS SOA RA draft articulated, Service Orientation (SO) switched from pure technical problems of application integration somewhere in the depth of IT onto the task of adoptiation of business changes in the technical solutions. This topic sounded in practically all presentations. SO came tightly coupled with the words 'business change'. The presentations were about 'can SO help in handling such changes and how to do this'.
When I listened to presentations in other tracks and talked to people, I had found that, in reality, SOA appeared in 60-70 per cent of discussions on different business and technical topics but nobody wanted to talk about SOA per se. I think this was the shock-effect of the announcement of SOA-death-in-IT, but, from another hand, this also was a sign of SOA commoditisation. This is the evidence of wide acceptance SO while SOA itself still has a lot of problems and it still evolves under the cover of business topics.
Leave a comment