[Info-vax] Seasons Greetings

[Arne Vajhøj]

Main, Kerry wrote:
>> From: info-vax-bounces at rbnsn.com [mailto:info-vax-bounces at rbnsn.com] On
>> Behalf Of Arne Vajhøj
>> Companies can not afford not to do SOA. It is pretty expensive not
>> to reuse.
> 
> No. - that is one of the biggest fallacies about SOA. In theory yes, re-use
> always makes sense, but the biggest challenges with SOA is that it requires
> massive changes in a companies culture.

Not necesarrily.

> Ever wonder why DCE never made it big in the real world? Again, the answer
> is not technical, but rather company culture and politics.

Since DCE and SOA is very far from each other I do not see the point.

> Here is an article from CIO magazine that sheds additional light on SOA
> challenges:
> 
> "The Truth About SOA"
> http://www.cio.com/article/print/21975

Much of it is true.

But it is not particular about SOA.

The bottom line is that companies which does not have control over their
IT can not change IT strategy.

It is not even IT specific.

If the company does not have control over their sales people, then
they can not change sales strategy.

>> SOA is most definitely not about replacing systems. You could argue
>> that SOA is about not replacing systems.
>>
>> Distributed environments is a reality today. And it is not going
>> to go away tomorrow.
> 
> Again - another fallacy that needs to be challenged.
> 
> While I have no doubt you really believe this from a programmers point
> of view, you also need to consider that this distributed approach is 180
> degrees different from the Operations side of the house which is implementing
> massive server and DC consolidation projects. They are under massive pressure
> to reduce IT costs and distributed systems are extremely expensive and
> difficult to manage because of the inherent complexities.
> 
> Of course, distributed systems will always be around, but the huge focus
> today is on centralized strategies. Shared services, cloud computing and other
> buzz words are just today's terms for what used to be known as timesharing
> (SLA's, service catalogues, capacity planning, centralized control of resources,
> very high security and HA etc).
> 
> VM consolidation will be the next big thing once a company completes most
> Of the VMware/other server virtualization projects.

But they will never get out of the multiple servers and multiple
platforms situation.

>> SOA is not a university thing. They still do OCAML, Haskell and
>> similar - SOA is practical thing.
> 
> SOA is based on strategies that are great in theory, but are extremely difficult
> to justify in terms of ROI as well as politics and changing an entire companies
> culture.

Unless the company culture is chaos, then little changes to it should
be needed.

>> Typical SOA advocates have 10-25 years of experience.
>>
> 
> With no fully completed company wide SOA projects to base their decisions on.

True.

But then if they 50 years ago had said "you want to build a computer,
but you have not done that before - you should stick to pencil and
paper" then ...

And BTW it is the wrong approach to want to transform the entire
company at once.

Arne