Subject: Re: How accurate is Metcalfe's law? (Was: Ximian software)
From: Ian Lance Taylor <>
Date: 06 Jan 2002 21:19:23 -0800

"Stephen J. Turnbull" <> writes:

> >>>>> "Ian" == Ian Lance Taylor <> writes:
>     Ian> I do understand that.  It's not what I meant.
> If you understand that it's the diamond-water paradox, then you just
> don't want to believe the consequences, I guess.

No, I mean that I do understand the diamond-water paradox (which is,
after all, pretty obvious), but that that was not what I was talking
about.  You misdescribed the point I was trying to make, no doubt
because I did not describe it clearly myself.

>     Ian> Since the price of software is not so tied [to developer
>     Ian> compensation],
> It is so tied, it's just that the tie is demand-pull on developer
> salaries, not cost-push on software prices.  The software that people
> want to buy commands high prices, firms move in to exploit that, they
> hire developers because they need to work around IP, developers become
> scarce, employers bid up their salaries.  If you don't believe that
> is a fact, then please explain to me why (1) the two free[1] "used
> labor" markets for professionals in Japan are securities traders and
> software engineers; (2) there exists the oft-bemoaned shortage of CS
> educators; or (3) the enormous attraction of software engineering to
> Indian knowledge workers, in India where the returns to bureaucratic
> posts are even larger than in Japan.

The price for in-house software development, and for contract
development, is clearly directly tied to developer compensation.  That
seems to me like enough to explain your observations.

As a side issue, working around IP has never been part of my
programming experience, except for the special case of public key
encryption.  Nobody who wants to write a new word processor or a new
spreadsheet or a new version of Windows has to work around any IP.
(Well, all significant programs violate software patents, but most
people don't work around them; they just ignore them and hope for the

>     Ian> However, I'm a programmer, and I don't believe that software
>     Ian> [creation amounts to "accidental" discovery of obscure
>     Ian> ideas].  Although there are a few exceptions, most good ideas
>     Ian> in software are obvious
> Fred Brooks says that's wishful thinking.  Programmers have very
> strong incentives believe that; I'm not surprised you do.  By the same
> token, I have to say that on this I trust the managers more than I do
> the engineers.

Why do programmers have a strong incentive to believe that?  I would
think that programmers would have an incentive to believe the reverse:
that they were very clever in their ideas.  I've certainly known
several programmers who thought just that.

Can you point me to where Fred Brooks said that was wishful thinking?
Since I thought I mostly agreed with him, I'd be interested in finding
the difference.

I'll add that most managers I've worked with tend to make the mistake
that programmers can do anything which can be described.  They tend
not to think that programmers spend their time scratching their heads
and looking around for a good idea.

>     Ian> The ASP model, in which the program is made available as a
>     Ian> service accessible over the Internet, is another approach.
>     Ian> It is entirely feasible to charge users $1 each time they use
>     Ian> the service
> Sure.  So you want to allow monopolists to achieve _first_ degree
> price discrimination via cookies, data exchange, and other means of
> personal characteristic tracking?  With much less possibility of
> competition because only the UI (and not the implementation) is
> available for inspection, and NDAs are much stronger than copyright?
> Where are you going, man?

You said ``Software licensing is _all about_ compensating people's
time, because it's the only[1] way to link the value of the
developer's time to the value received by users (including downstream
developers).''  I was challenging your statement: it is not the only