Re: What if we designed and built a "32bit" Apple II?

[mjmahon@aol.com (Michael J. Mahon)]

Roy replied:

>Michael J. Mahon wrote:
><snip>
>
>> That is why I said that if you love doing this development, go for it.
>> But if your satisfaction requires dozens of others to hop on the
>> bandwagon, then you will be disappointed.
>
>Here's is where we do disagree. "Dozens" is doable. "Hundreds" ain't

We agree.

I was talking about people working on essentially _independent_
software or hardware projects, dependent only on a common
understanding of a machine architecture.  For example, you
write a spreadsheet while I build a compiler.

If we're talking about an active collaboration on a single project,
then even dozens is pushing the envelope, in my experience.

I've always found the optimal number of people on a project
like a compiler to be two to five.  If a project is bigger than can
be done by such a small group, then it must be divided into
relatively independent smaller projects, each of which can
be designed and implemented by a small team.  (A project
team of one person is suboptimal, because everyone benefits
from having someone to discuss things with.  But sometimes
it's all you've got. ;-)

As you add people to a tightly-coupled (read: poorly structured)
project, the communication requirements go up as the square
of the number of people, and the useful work goes up much less
than linearly with the number of people.  At some number of people
around "dozens", a project is usually segmented out of desparation,
because little progress is being made.  ;-)

(Of course, that _post facto_ segmentation under duress is almost
always terrible, but that's another story.)

-michael

Check out parallel computing for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/