[Leica] OT: software, engineering, and microsoft

[Pete Su <psu@kvdpsu.org>]

I've been offline for a bit, but had to throw in a belated off topic 
reply to this:

> The reason for MS's success is that a software engineer can easily be
> assigned a different task. There is no huge expenditure of time and 
> investment in
> machine tools and the dozens of other industrial expenses required in 
> making
> dramatic shifts in production of new products. At MS Gates calls a 
> meeting of the
> managers and explains the project and the time requirements. A 
> software outfit
> can turn on a dime, not hundreds of millions of dollars.

Whoever wrote this knows little about software, engineering or software 
engineering.

Software by its very nature requires a great understanding of both high 
level design and arcane low level details. The notion that you can move 
software developers around like interchangable parts was discredited as 
long as 30 years ago by the classic book by Fred Brooks called "The 
Mythical Man Month". I suggest you go read it.

It is also a fact that one of the reasons Microsoft (and IBM before it) 
manages to maintain a dominant lock on desktop software is precisely 
because *changing* software infrastructure is horrendously expensive, 
and homogeneous systems are easier and cheaper to maintain and deploy. 
Whatever else you say about Microsoft, the fact that they were able to 
literally turn on a dime and decide to dominate the internet browser 
was one of the most impressive feats of ruthless management that I've 
witnessed. As someone who has worked in software for a decade, I can 
tell you that it is decidedly not easy to do something like that.

Pete

- --
To unsubscribe, see http://mejac.palo-alto.ca.us/leica-users/unsub.html