I read this as the idea that a database can be made "as easy to
use as a spreadsheet", although I can't actually imagine what that
might really mean, since people can't find the spreadsheet they
want anymore, either.
I think that Larry's expression of the differences between the MSFT
and ORCL views of the future is more cogent, and while it is obviously
biased, it seems a bit more balanced than Bill's presentation.
The basic difference is, is it more valuable to let everyone have
a database wherever they want it, or to let everyone have access to
the data and manage it centrally. I think that it's disingenuous
for MSFT to say that we can have a zillion little databases all over
the place and worry about keeping things consistent after the fact.
That trick has never worked, and I don't see why having some new
icons on the desktop is going to make it work this time.
On the other hand, if you can convince people with small databases
to buy your less expensive software with the idea that it will scale
up by the time they need it to be bigger, and that you will copy any
feature of the higher-priced DBMS vendors by the time people need it,
then you might be able to make a case. The fact that MSFT hasn't
been able to do it so far doesn't seem to bother a lot of people,
and for all I know MSFT hasn't really tried yet. They presumably do
try to make NT reliable, and it's still not very good, and an RDBMS
kernel is at least as complex as an O/S kernel.
I am long ORCL and I sold my MSFT last summer, which wasn't that good
a move according the the "smart money", but I still think that ORCL's
strategy is more promising... |
