GPL really free was Re: [Fsfe-ie] 1-page letter, faxes at the ready (IFSO, irish org)

[Niall Douglas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 28 Sep 2003 at 2:40, David Golden wrote:

> > I'm not disputing that. I just feel the GPL has bad knock on effects
> > considerably underestimated by its proponents.
> 
> I don't think the effects are underestimated by me, anyway, but  I
> would note that "bad" is in this instance quite thoroughly in the eye
> of the beholder...

This is a particularly modern viewpoint of bad being subjective. I 
disagree - there is most certainly absolute good and absolute bad.

> > Information is power, and thus it is valuable to those who wish to
> > exert power. For example if most Americans knew the true state of
> > Saddam's Iraq, they would never have let Bush invade that country.
> 
> Having value does not make something property, nor is it a valid
> argument for making it into property, particularly if it's non-scarce.
>  It certainly doesn't have a simple and unchanging value, e.g. value
> you could depreciate with a simple rule like office furniture.  Think
> about the value of timing of information - yesterday's news today is
> much less valuable to most people than today's news today.

But the same goes for say gold - one day it's worth something, next 
day something else.

I don't think you'll get any disagreement from anyone here that 
information which has neglible reproduction cost should be much freer 
than it is. But I think we can all also see why there is a desire to 
control it and make it property by those who have something to gain 
from it being so.

> It's a little more accurate IMHO to say asymmetry of information can
> translate to power - the value was not in the information in your

Asymmetry of *anything* translates to power eg; guns, wealth, food 
etc. Even asymmetry of love, faith, truth etc.

> All in all, life would be a lot easier if people stopped trying to
> treat bits like bricks...

Are you referring to the innate interconnectedness of all things in 
the universe?

> Thus far, we've bandied about the term "innovation". That has a 
> economic meaning  - "innovation" is the introduction of something new
> to a market, often with disruptive effects, think "act of innovation",
> and a more common, loose usage, largely synonymous with invention,
> think "an innovation".   As far as I can tell, thus far, we've been
> using the common-usage one, since the economic usage allows evolutions
> to be innovations. I have continued with the common usage below:

I would define "innovative" as doing something not done before in 
that particular context.

> Components of Linux are innovative. System evolution by subsystem
> revolutions.  System revolution by subsystem evolutions. By
> gerrymandering your system boundaries and lumped-approximating, you're
> probably always going to be able to declare "innovative" and
> "non-innovative" bits - "An OS written in C on a Computer? Been done.
> Bah."

I was more meaning that Linux was ten years behind the state-of-the-
art when Linus first wrote it. And it's continued to remain 
substantially behind best practice ever since with very substantial 
resistence to adding things like threading support which NT had in 
1991.

>  Could Reiser rewrite filesystems without having to rewrite everything
>  else if 
> linux weren't modular?  

We've had modular operating systems since at least the IBM/360. OTOH 
what Reiser has done in the more recent versions of ReiserFS is gone 
places which to my knowledge haven't been visited before and the next 
versions he's sketched out makes me feel quite excited :)

> Another thing is that people often regard as innovative stuff that
> isn't really but that they don't know the history for.  You mention
> EROS, well, that's an evolution of KeyKOS.  There's a also quote,
> which I'm about to fail to attribute, which paraphrased is - "you can
> make it big in Computer Science by looking at what people were doing
> years ago, waiting for people to forget about it, and releasing it as
> a big new thing":  XML<--Lisp sexps, only badly done,
> OODBMS<--Hierarchical DB (bad idea then, bad idea now...), etc.

Absolutely, but then all innovation comes from what was done before. 
I know EROS borrows heavily from prior systems but when I was reading 
its specs it was still doing things which hadn't been done before on 
an operating system to my knowledge (certainly not in its 
predecessors).

> > There is virtually nothing on Windows or Linux which
> > qualifies as "blue sky" innovative. 
> > If you want to see "blue sky"
> > innovative, go look at EROS, Plan 9, Syllable or even GNU Hurd
> > (though ten years ago). That kind of innovation demands total
> > rewrites from the ground up.
> 
> And those projects are... open source. Anyway, in the open source
> world, features and pieces can be interchanged readily - even if no
> end-user ever bothered with Hurd or EROS, they have already had
> lasting effects on the design of Linux and other OSes.

If they weren't open source, we wouldn't know about them :)

> > And that kind of sweeping change is something free software cannot
> > do because you'll never get enough of a team together to agree on
> > something totally unproven. Only investment capital can pay people
> > to do the work required to prove a concept.
> >
> You say that, yet 75% of your examples are GPL and quite actively
> developed? Whatever.  That's a baseless assertion, as far as I'm
> concerned. 

Not at all. I never claimed that they weren't GPL. I *did* say the 
GPL works against blue-sky innovation.

Let's look at the history: EROS was developed by academia, Plan 9 by 
AT&T labs, Syllable by one guy plodding away alone for nearly three 
years until he'd proved enough of his concept to get people to help 
him.

Now look at GNU Hurd. They had the idea back in 1991 and it's STILL 
not finished today. It's not the fault of its developers - its 
because when an idea is unproven, it's hard to attract volunteers who 
don't want to also add to the recipe.

> Quite apart from the fact that that assumes people need
> paying to do the work (history has many counterexamples, and I know I
> do blue-sky stuff just for laughs), significant Investment capital can
> be, and has been in the past, given to open-source people, companies,
> universities, and so on.  Your generalisations about human motivation,
> which you've done a few times, are iffy, and probably reflect most on
> your own motivations.

You've just proven my point - some external entity needs to invest 
some capital to employ workers.

Now consider an innovative project. One guy makes it look exciting 
but if he's GPLed it business wouldn't bother investing. Thus either 
he plods away alone for a few more years, or it withers and dies.

At least with BSD or LGPL projects business can buy it in, develop it 
and use it as a base for some product.

What the GPL is *extremely* good at is making superior clones. Look 
at KDE. It looks like Windows, works like Windows but is also 
slightly better than Windows. But it's still cloning Windows - 
nothing radical there.

You can carry on throughout most free software - it either clones 
other Unix software or it clones Windows software. A small proportion 
is original eg; the scripting languages. But there's not much in 
there which is radically innovative.

> > Agreed. I prefer a clause like the LGPL has which mandates
> > enhancements to some code are returned to the community.
> 
> Well, actually, it only mandates return if you _distribute_.  If you
> enhance in-house without ever distributing, you don't need to return.

Sorry, true. This is good actually as some enhancements do harm :)

> > On this we would disagree. I want a law mandating that source comes
> > with all code.
> 
> Nah. In the absence of copyright and patent, binary-only code would
> soon be outcompeted.

Doesn't matter. It's a case of consumer rights to me, what's just.

> > Well, they'll do anything they [MS] can to win.
> 
> Yep, though I predict they'll eventually burn out, and assuming I'm
> not hit by a bus or whatever, most likely within my lifetime.  All
> empires fall, and MS's could crumble alarmingly quickly.   I do find
> Linus' recent quote in the new york times amusing:

Certainly even the average user is getting sick and tired of crummy 
reliability - viruses, worms etc.

> [ From NYTimes, 2003-09-38 ]
> The thing is, at least to me personally, Microsoft just isn't relevant
> to  what I do. [....] I just can't see myself in the position of the
> nemesis,  since I just don't care enough. To be a nemesis, you have to
> actively try  to destroy something, don't you? Really, I'm not out to
> destroy Microsoft. That will just be a completely unintentional side
> effect. 

I wouldn't put it past MS to have Linus assassinated. If I were him, 
I'd be careful.

Cheers,
Niall





-----BEGIN PGP SIGNATURE-----
Version: idw's PGP-Frontend 4.9.6.1 / 9-2003 + PGP 8.0.2

iQA/AwUBP3ZNucEcvDLFGKbPEQJ4KwCfeISMsB/YbA7GP4xx8D7Ao+ipGwMAn0+9
MID84lzR4k12Ewk3CWKfXAuI
=Yos3
-----END PGP SIGNATURE-----