Re: [Fsfe-uk] Liberated software

[Chris Croughton]

On Sun, Oct 10, 2004 at 01:45:03PM +0100, Philip Hands wrote:

> Chris Croughton wrote:
> >On Sat, Oct 09, 2004 at 09:40:01AM +0100, Simon Waters wrote:
> >
> >>I think the mistake is to worry too much about the misunderstanding
> >>of the term "free", and concentrate on sending a clear message that
> >>software that doesn't cost anything, but to which you have no right
> >>to inspect, modify, or redistribute, isn't free software.
> >
> >But it is, that's the problem.  As someone else pointed out, the word
> >'free' is almost as overloaded as the word 'love', and any and all of
> >the meanings are as accurate as the others.
> 
> All of these things are ambiguous, and it doesn't help much to adopt the 
> Humpty Dumpty "words mean what I want them to mean" approach.

Exactly.  The problem is not that they mean what /I/ want them to mean
(or as Simon wants them to mean), but that they already mean a load of
things which other people already use and have used for decades.
Picking out one of those meanings and saying "this is what the word
shall now mean and nothing else" is doomed to failure.

> The fundamental problem is that only a tiny percentage of software users 
> know anything significant about licence conditions, be they proprietary, 
> or Free.
> 
> Of those, few know anything about the plethora of Free software licenses 
> that we have, or what they might imply.
> 
> Of those, only a fraction are aware that things like YaST (under it's 
> old license) or qmail are not Free Software and/or Open Source.

Having never used either YAST or qmail, I've never had occasion to look
at their licences.

> Of those, many are under the impression that Open Source is supposed to 
> be different from Free Software, for example thinking that Free Software 
> means that the software must be under the GPL, whereas Open Source 
> includes BSD, say.

GPL => FS and OSS
FS  => OSS

The inverse is not necesarily true (but a lot of people reason that "if
A implies B then B implies A", and "if A implies B then not A implies
not B", they are very common logical faults).

> Of the people that are aware that the intent is that FS == OSS, there 
> are presumably some who are unaware that despite the fact that the 
> intent was to mean the same thing, the reality is that FSF, Debian and 
> OSI have managed to differ on the interpretations about things like the 
>  APSL 1.x.

Was the intent ever that FS == OSS?  FS implies open source, yes, but
all the articles I've seen from the FSF and RMS have been very clear
that they are not the same (and that open source does not imply free).

> So we're left with some tiny proportion of the population that have a 
> clear idea of what the stuff is that the terms "Free Software" and "Open 
> Source" are supposed to describe.

And they probably don't all agree with each other (i.e. they each
individually have a clear idea but the ideas aren't idetical).

> There are many more that are adamant that they know that Open Source 
> means that you get the source and no more, for example, and would tell 
> you that qmail is Open Source, but not Free Software, because it's not 
> under the GPL (or some such nonsense).

Yup, I've come across that.

> It seems that the proposed solution to this lack of understanding is to 
> come up with yet another name for the thing that almost nobody is 
> currently capable of recognising.
> 
> Anyone not already in the know is bound to think (as they do about OSS & 
> FS) that if you have more than one name for something it must indicate 
> that there is some important difference between the named things, and 
> that that difference is the defining feature of whatever it is you're 
> talking about.

Unfortunately, labels are essential to thinking at all (try coding
without variable and function names!).  But labels also get identified
with the thing they represent, which leads to errors.

> Forget the name argument, and just explain to people that:
> 
>   The software we're interested in (by whatever name) comes with
>   permission (or freedom) to modify the software, and redistribute the
>   result, without first asking the copyright holder.
> 
>   If it doesn't have at least that, then it's not Free Software (as the
>   FSF uses the term) no matter how cheap it is, and it isn't Open
>   Source (as the OSI uses that term) even if you get to see the source.
> 
>   ... and it isn't Liberated Software, even if you stole it ;-)

But that's back to the Humpty Dumpty problem, that it's saying "as some
person or organisation uses the term", and immediately opens up the
rebuttal "But it /is/ open source and free software as Microsoft and the
man on the Clapham Omnibus use the terms".  What gives us the right to
define the terms one way and denies it to MS and the "common people" to
define them another?

> Once we have a few more people contemplating that licensing distinction, 
> we can start worrying about the best name for the thing to which we're 
> referring, but since we'll probably all be dead long before that 
> happens, I'm not overly worried about that phase of the project.

If it doesn't have a 'handle', no one will make the distinction unless
it affects them directly.  I distinguish between them because I am
directly concerned (I've had to rewrite several pieces of software
"clean-room" because the originals didn't have a licence which permitted
their use in the situation I needed), but the man in the street doesn't
care, he just wants something which does the job he wants it to do
adequately.

> P.S. I happen to prefer Free Software as the name, simply because the 
> misunderstanding of that term seems to be easier to deal with than the 
> misunderstanding of Open Source, but I'm not going to waste my breath 
> trying to persuade IBM's or RedHat's marketing departments that they 
> should drop the Open Source term.  A regular mention that they are 
> synonyms would be an improvement though.

I wonder whether the FSF would agree...

Chris C