[plug] 30 theses

Leon Brooks leon at brooks.smileys.net
Wed Sep 6 12:23:08 WST 2000


[lunch break time: please read this and add theses. Each thesis should be short
(prefereably 2 lines or less) and to the point, but (and you know who you are)
not violent or needlessly offensive. I plan to collate all useful suggestions
and get them posted somewhere well-read on the Internet]

Preamble
~~~~~~~~

One might think that during the past 483 years, some civil progress would have
been made. However, it seems that in certain areas, specifically computing,
"there is nothing new under the sun". Today we are faced with a monopolistic
situation not unlike the one faced by Martin Luther when he nailed his 95 theses
to the door of the Wittemburg cathedral.

"The only thing one learns from history is that nobody learns anything from
history", claimed Hegel, and these new theses are yet another effort to prove
him wrong. It is our fond hope that you will take these ideas to heart, and
change the way you think about computers, perhaps even change the decisions you
make about buying and specifying computers.

While the modern equivalent of this action would be to "nail" our 95 theses to
Microsoft's home page, we feel that this would not even be as well received as
Martin's action was in his day, so we are publishing these theses here.

1. The opportunity for greatest freedom of choice is key to the greatest common
good.

2. Freedom of choice naturally includes the freedom to avoid choice.

3. The safest and simplest way to guarantee freedom of choice is to leave as
many choices open as possible in each step of the software food chain, from
developers through distributors to end users.

4. The only certain method of pushing control out towards each end user is to
make complete and useable source code readily available to everyone.

5. Long feedback chains mean late and often erroneous changes to software.

6. The shortest possible feedback chain, zero hops, results when the end user
can also be the developer, and vice versa.

7. Tailored gear can be made to fit better than sweatshop production-line gear.

8. Practice produces a more well-rounded product than theory.

9. Most people work best when they're working for themselves.

10. Most people take better care of something that they own. We need to hear the
lesson in "Don't worry, it's a rental."

11. Many eyeballs (and minds) make light work of debugging.

12. The only certain way to provide ownership, short feedback chains, practical
coding, self-powered and empowered workers and a broad skill base is to make
development tools readily available to the end user.

13. An example can be worth a thousand words of theory.

14. Monocultures are unhealthy, and in the long term not viable (in a natural
setting, they will die out: think about the Irish potato famine or the effects
of Texan broadacre monocropping on Russian farmland).

15. Polycultures are healthy and beneficial, even for nearby or connected
monocultures. The most robust situations in nature are extremely complex.

16. The only certain way to provide a broad spectrum of useful examples is to
establish a framework within which source code access is a right and source code
provision is an honour.

17. Entities with a larger number of centres of control are more difficult to
destroy than more "concentrated" structures.

18. Entities with a larger number of centres of control are more difficult to
subvert than more "homogeneous" structures.

19. Any significant success in a contested field will result in attempts to
subvert or destroy the successful entity from limited numbers of hostile
competitors.

20. Distribution (even abandonment) of control makes destruction or subversion
of an entity by hostile competitors difficult more quickly than it makes
assistance or empowerment of the entity by sympathisers difficult.

21. Hostile competitors must also have a right to life, however "the liberty of
the corporation must be thus far limited: it must not make a nuisance of itself
to other entities."

22. There are many "right ways" of acheiving an end; where resources permit,
many should be tried.

23. Reliability is generally more important for a user than snazzy features.

24. Reliability is generally more important for a user than the product being on
supermarket shelves (or FTP servers) for Christmas.

25. Wide document availability is generally more important that the power to
double-underscore, colour or "fontasize" about your text.

26. Keeping your passwords, credit card numbers and disk structure safe is
generally more important than being able to fly a little man around on somebody
else's screen on the fiery rocket blast from his backside, accompanied by a
jolly tune.

27. Having a public encryption system that works is better than having a private
one with your name on it and enormous security holes.

28. "Write once, run anywhere" is more safely and politely achieved with Python,
TCL or PERL than with a black-box ActiveX binary or VBS.

29. Reliability, interoperability and security are generally better achived by
bowing to a standards committee than by bullying one.

30. Paying a local computer guru $500 to set your machine up "just so" feels a
lot better than paying a remote and faceless software corporation $500 just for
a licence to use it.

31. [add more theses here: we want 95 of them in the end]

-- 
"I think that's how Chicago got started. A bunch of people in New York
said, 'Gee, I'm enjoying the crime and the poverty, but it just isn't
cold enough. Let's go west.'" -- Richard Jeni



More information about the plug mailing list