Re: Design principles and ethics (was Re: Execute without read (was [...]))

[Bas Wijnen]

On Fri, Apr 28, 2006 at 02:45:32PM +0200, Pierre THIERRY wrote:
> > > I fully support the fact that you won't implement yourself some
> > > schemes.  But as an OS architect, you must not close doors to other
> > > implementers of thoses schemes, at least for this only reason.
> > As an OS architect, no.  As a free software programmer and activist, I
> > have strong reasons to do so.
> 
> I admit I do sometimes think the same way (and do act in this way), but
> I'm not sure it's not morally objectionable, sincerely.

It isn't, it is worth a lot of praise.  The position of "I didn't do anything
wrong, I only made it posible" combined with "I didn't know I was doing
anything wrong, it was so easy" directly leads to bad things happening.  Both
parties involved are to blame IMO.  And that implies that making bad things
possible without a good reason (such as that it makes good things possible as
well) is morally objectionable.  There is no ethical rule that says you should
make everything possible that you technically can.  You seem to think that
there might be.

> > I would recommend that Alice and Bob go to a keg party and have a beer
> > over their differences.
> > 
> > I would further educate Alice about [...]
> > 
> > Furthermore, I would talk to Bob and warn him about [...]
> 
> That's a faithful commitment to social progress in the world, and I can
> only promote it, but it does not solve the problem here.

There isn't really a problem here in the first place, so why should it be
solved? ;-)

In practice, Alice can give Bob a binary, and it's well-known that
reverse-engineering a binary is much harder than simply redoing the thing.
And anyway, if Alice wants to show this thing to Bob, then it is very likely
that they will sit behind the computer together.  Alice would definitely
notice if Bob starts viewing the code.

> Your conclusions about the fact that reverse engineering is harder than
> writing from scratch, that NDA are dangerous and than educating people
> instead of limiting them are, at best, true in the general case.

Eh, what is true in the general case is also true in a specific case...  You
don't seem to mean that.

> But building principles on the general or, worse, ideal case is a very
> dogmatic position, IMHO.

Of course it is.  What's the difference between a dogma and a design
principle?

> In fact, you do with the OS what you would tell Alice not to do with her
> source code, I think. You should not prevent people to do morally
> objectionable uses of the system. You should go educate them so they
> don't want anymore.

It's not like it's impossible to do.  Alice can simply set up a server which
lets Bob start the program (on her resources, so Bob can't debug it).  Call it
a constructor. ;-)  We should still educate Bob that he should talk to Alice
and they would both be more productive when they look at each other's code.

> > It's really sad that some students have swallowed up the story of
> > "everybody against everybody" so early in childhood that they bring
> > this attitude to university.  It's something that we should work
> > against, not support.
> 
> That's not the matter here. The people I love around me somethimes have
> to protect me from myself, and I sometimes have to protect them form
> themsleves. And we are generally wery grateful that we did that to each
> other. That could be the case between Alice and Bob here.

In that case, Alice probably doesn't want to let Bob see her program at all,
she will just want to talk to him and explain what he should do.  Or if they
do, she will want to be there when Bob is testing her program, so she can
explain why things are the way they are.

This use-case simply doesn't need a technical limitation on Bob's abilities.
Marcus claims (so far non-publicly) that there are no use cases which are
worth supporting.  I have my doubts about that, but I think I agree.  I
suppose we'll find out when he brings out his whole argument. :-)

Thanks,
Bas

-- 
I encourage people to send encrypted e-mail (see http://www.gnupg.org).
If you have problems reading my e-mail, use a better reader.
Please send the central message of e-mails as plain text
   in the message body, not as HTML and definitely not as MS Word.
Please do not use the MS Word format for attachments either.
For more information, see http://129.125.47.90/e-mail.html

signature.asc
Description: Digital signature