Re: [libreplanet-discuss] Are proprietary software and science compatible?

[Mike Gerwitz]

On Sun, May 15, 2016 at 14:14:48 +0200, Fabio Pesari wrote:
> I would appreciate this article more if it wasn't written by people who
> can program.

I agree.

There are smaller articles/editorials that were published in Nature that
mentioned the issue (few and far between), but I'd have to dig them up,
and they weren't substantial; I read them in print, so I don't have the
links.  But even a mention gets people t

> I'll make a quick example - Stephen Hawking. He literally depends
> exclusively on software (and hardware) to interact with the world and if
> I were him, I would worry a bit more about software freedom, because
> otherwise anyone can put anything in his mouth, but I don't think I've
> ever heard him talk about free software (even if ACAT has been released
> under the Apache License 2.0, it's Windows only).

I do wonder if he has given consideration, or if he doesn't care because
he trusts that others have his interests in mind and he wouldn't modify
the software to begin with.

In any case, it is unfortunate.

> To the article (I'll get to the rest of your mail below).

>> One proposed solution to the problem of ambiguity is to devote a
>> large amount of attention to the description of a computer program,
>> perhaps expressing it mathematically or in natural language augmented
>> by mathematics.
>
> This is actually a better solution than the one they proposed because it
> allows free/libre software developers to implement a new program without
> looking at any proprietary code, which would get them into trouble for
> copyright infringement (and prevent anyone who's actually looked at it
> from writing any code, like it happened for Wine).
>
> This also only works for first-party code. If the code depends on a
> third-party program or library (the majority of cases), there's still a
> problem.

It's useful, but in practice, a formal mathematical definition rarely
manifests in a disciplined manner in procedural code (especially not for
scientists who don't have formal, disciplined experience writing
software).  Declarative languages and ones with theorem provers may be
less error prone, but it's the subtle mistakes people make in software
that are especially problematic.

(Obligatory "The Limits of Correctness" reference:
https://www.student.cs.uwaterloo.ca/~cs492/11public_html/p18-smith.pdf).

But I can agree that a formal mathematical definition, where possible,
should always supplement the software.

>> and finally that the development of program code is a subsidiary
>> activity in the scientific effort.
>
> This one is a barrier but only mentally; writing software can't be
> considered a subsidiary activity anymore, in any field.

Agreed.

> I think that rather than characterizing proprietary software as evil, we
> should call it for what it truly is, irrational.

That gave me a chuckle, but you're absolutely right. :)  Though I've
seen many programs that I really can call "evil".

-- 
Mike Gerwitz
Free Software Hacker+Activist | GNU Maintainer & Volunteer
https://mikegerwitz.com
FSF Member #5804 | GPG Key ID: 0x8EE30EAB

signature.asc
Description: PGP signature