on 3/1/01 10:56 PM, Michael W. Wellman at [EMAIL PROTECTED] wrote:
> One of the problems that comes up here is that the problem is mostly an
> engineering problem. And the pre-docs are mostly interested in science, not
> practice.
>
>>>snip>>>
>
> In a naïve test environment, every input into a system doubles the test
> matrix. Which quickly brings a test department up against the "practically
> infinite".
>
This is what I mean. Current methods and practices are getting unwieldy and
are being overrun by today's programs. What we need is new, radical, ways of
evaluating software. Ways based on new concepts. New ways to evaluate
software and in its environment. A new visualization of how software works.
This is science. Engineers take this new understanding and visualization and
make something practical of it.
>> Now if Microsoft was truly as innovative as they claim, and still desire to
>> be the Pre-Breakup ATT of the software world. They need to create the
>> software equivalent of Bell Labs and put some very smart minds into advance
>> basic software research.
>
> Oddly enough, the best software ideas still seem to come from small groups
> and/or individuals. Most often working in isolation.
>
> Large teams tend to moderate towards the average...unless they've a truly
> forceful leader who is compelled towards "the right thing".
>
That¹s why I mentioned Bell Labs. During the 40's, 50's and 60's, ATT was
smart enough to hire very smart and bright people and turn them loose. True,
most of what Bell Labs did was basic, mundane engineering work. But deep
inside, there small groups of individuals and teams developing the basic
concepts and theories that we rely on today. Two that come to mind are
Shannon's' information theorems and the transistor. Without them, none of
this we're using today would be possible.
That's what I was thinking about when I suggested that Microsoft create a
software equivalent to Bell Labs. A place where very bright and talented
people can work individually or in small groups to solve the fundamental
problems of software creation and use.
--
To unsubscribe: <mailto:[EMAIL PROTECTED]>
To search the archives:
<http://www.mail-archive.com/entourage-talk%40lists.boingo.com/>
