Re: Release plans

[Alan Mackenzie <acm@muc.de>]

Hi, Stephen,

On Tue, Aug 26, 2008 at 01:54:15PM +0900, Stephen J. Turnbull wrote:
> Alan Mackenzie writes:

>  > Yes, we want software to be free, but no, we don't want people to
>  > use this freedom in certain ways, ways which would inhibit the
>  > progress of free software.

> I'm not sure I agree with this formulation, but it's not what I'm
> talking about in this thread.

I think it is.  I think it's the abstract principle behind RMS's decision
not to put a binary module loader in Emacs.

If you don't like my formulation, how about reformulating it your own
way?

> I'm talking about a specific decision which is based not on observing
> people misusing freedom, not on the likelihood of people misusing
> freedom in foreseeable ways, but rather on the conceptual possibility
> that in some as-yet unforeseeable way somehow someday somebody might
> possibly misuse freedom to a disastrous end.

You might be right there.  Such a "conceptual possibility" existed wrt
patents in GPL2.  Microsoft and Novell exploited it to give special
"patent protection" (against unspecified MS patents) to direct Novell
customers only, in gross violation of the ideals of the GPL.  Rather
clumsily, they pulled the stunt as GPL3 was being developed.  I think
it's right to be very wary about exposing possible vulnerabilities to
proprietary competitors.  They will be actively seeking weak points to
exploit.

[ .... ]

>  > Eric Ludlam mentioned a product called Xrefactory a couple of days
>  > ago.  It seems to be a refactoring tool based upon (X)Emacs.  Yes,
>  > this is legitimate within the terms of the GPL, but isn't the sort
>  > of thing we really want to encourage; it's not free, neither in the
>  > speech nor beer sense.

> The reductio is obvious, isn't it?  Since any free software program can
> be abused as part of a non-free software product, we should stop all
> distribution of free software to the heathen.  That way we can be sure
> of providing the minimum encouragement to abuse freedom.

No, not at all.  The absurdum is what I pointed out yesterday, not what
you've just written: we restrict freedom to protect it against being used
to destroy itself.

[ .... ]

> Effectively discouraging non-free software is out of our control.
> "Mr. Quixote, meet Mr. Windmill...."

It is not.  There are few non-free extensions to Emacs, at least that I
have heard of.  With a binary module loader, there might well be more.

>  > I gladly accept the freedom guaranteed by professional soldiers.
>  > Just as those soldiers protect those "who don't give a damn", I feel
>  > we should protect the (software) freedom of those who, for whatever
>  > reason, wouldn't protect their own.

> At the cost of the freedom of those who would *like* to use unfree
> software: they have done the calculation on the costs of lock-in, and
> like the answers they got.

Yes, this is down in the irony and contradictions department of free
software.  But as you've noted, the lock-in is largely psychological:
there's nothing in the GPL to prevent anybody extending Emacs pretty much
however they want, and that includes adding a binary module loader -
providing the loaded modules wouldn't get too intimate with Emacs itself.

> I find your paternalism distasteful.

Yes, I expected that and I've no problem with it.  Likewise, I find
aspects of your personal philosophy distasteful too.  I look on it as a
source of fruitful debate, far better than the near slanging match we
were having just a few days ago.

[ .... ]

> Some people are willing to accept the constraints of unfree software
> constrained for commercial advantage.  Some people hate unfree software
> so much that they impose constraints on the use of their software by
> others, and claim that the net result is somehow an increase in freedom
> when in fact there is a clear decrease in options available to users.
> I don't really see an ethical difference here.

I think you're right.  That is down in the irony/contradiction/absurdity
region again, where the ethics can't help but being murky.

A while ago, I was arguing that Emacs being available on w32 was a good
thing - not from any highly principled position, just for purely
pragmatic reasons.

-- 
Alan Mackenzie (Nuremberg, Germany).