On Sep 18, 2010, at 5:21 PM, Doug McDavid wrote:
And the reason this is worrisome is that it introduces massive cost and lost opportunities into enterprises.
Odd you should say that. (Or maybe you're saying there should be distributed/central repositories & it's a pity there are not?)
One of the major business benefits in this case (http://www.tdan.com/view-articles/6123
) was flexibility... systems were modified (not cloned... a VERY dangerous activity) quickly to embrace a new product line.
IF done well—I know, BIG if—I would be happy to argue for reduced cost & increased opportunities.
The current state is that most organizations don't have a clue what they have for applications, much less how changing something in Silo A impacts Silo Z... so they just make the change in A & eventually pay for fixing the mess in Z (probably without ever knowing the connection).
I have recently conducted an informal survey... and learned in no uncertain terms that "inventory/impact analysis" is NOT something to discuss. People DO NOT CARE.
How organizations cope with constant change in a complex world WITHOUT having a firm grip on their applications inventory & how to trace impacts through & across systems is something I do not & will not comprehend. I do understand it's a play on that John Houseman/Smith Barney add... "we do it the old fashioned way....[sic] BY HAND."
Language SHOULD be part of this equation (and historically has not been)... when (not if) "new" language is introduced to the organization, it can be difficult to grok how if the new term is in fact new or just new jargon for old concepts. Eg. many know "social security number" often abbreviates to SSN. But it takes a bit more language understanding to grok that SIN is the Canadian version of SSN. Ignoring/cutting off data feeds because something isn't seen as relevant is a common risk.
Another facet of the "immediate discovery" at the desktop challenge... when someone finds a "new" term, the entire process (from desktop to distributed repository & back) must QUICKLY inform the desktop user that Term A & Term Z in the context of the user's need are in fact synonyms. Sort of like MSOffice spellchecker on super steroids.