[
Lists Home |
Date Index |
Thread Index
]
On Jan 31, 2004, at 5:03 PM, W. E. Perry wrote:
>
>> Perhaps UBL will become the authoritative nomenclature in many
>> real-world
>> settings, but we shall have to see about that.
>
> How is that possible without vitiating the real-world expertise of
> local
> processes? The leap from homogenous enterprise networks to the
> heterogeneous internetwork topology liberates local processes from
> naming
> conventions and from fixed, agreed data structures which constrain the
> locally appropriate expression of idiosyncratic local expertise.
My personal best guess is that UBL will not become a widely used
"authoritative nomenclature". On the other hand, it may hit the 80/20
point and get viral adoption -- what do I know? Prediction is hard,
especially about the future :-)
Whether it does or not probably depends on the extent to which it
really does capture that which is homogenous across local processes,
and serves as a good enough naming convention that creates so much
value via the network effect as to outweigh its deficiencies as a
"locally appropriate expression of idiosyncratic local expertise." It
depends more on whether UBL gets the kind of buzz that HTML and RSS
got, which overcame their woeful deficiencies vis a vis technically
better hypertext and news syndication formats. I'll bet against this
just because UBL attempts to standardize the things at the very core of
every company's business process that are under tight management
control, and hence not susceptible to being played with and debugged by
geeks until they reach unstoppable critical mass, e.g. the way HTML and
RSS came to be what they became. More loose (and locally specialized)
integration via contextual inferencing and data transformation is
probably almost as efficient as UBL-based translation, and probably a
lot less threatening to Pointy Haired Bosses.
But again, we shall see what we shall see; I'll be somewhat surprised
if UBL succeeds, but not astonished.
|
