[buug] (no subject)

[Rick Moen]

Quoting Johan Beisser (jb at caustic.org):

> There's a few that are true UNIX(R), as well.

And yet, nobody really cares.  As I said, the few people who did died by
around 1992.  So did, by and large, the companies.  BSDi is long gone,
for example.  Open Group still does some Single UNIX Specification (or
whatever it's called, these days) certification business to this day --
but that business has just about gone away because nobody gives a damn.


> I have to disagree. 

About "freeware" being ill-defined?  Well, sorry, it is.  

I hear the term bandied about to mean all sorts of things, and users of
the term almost invariably have no clear idea what it means, and morever
of what _they_ mean when they say it.



> "Open Source Software" has the vague definition of "Freeware" while
> not really being exclusive. 

You are flamboyantly mistaken, here.  Open source software has a formal,
highly useful, specific definition:  http://www.opensource.org/docs/osd

Honestly, did you not _know_ that?  It's OK if you didn't.  I'm just
surprised.


> "Freeware" harkens back to the "Shareware" software of the 80s and
> 90s.

"Shareware" is another ill-defined term, although rather less so.  Some
"shareware" permitted the recipient to create and send out derivative
works, some did not.  Some had source code access, some did not.  So, it
was a pretty highly vague conceps.  I believe the entirety of what it 
entailed was that you would have some right to redistribute (though
sometimes with severe and even ridiculous strings attached), and the
software held out its hand for money and/or created a legal obligation
to pay someone.



> Freeware was the "free to use, free as in beer" software that many
> people used.

That is one thing people have been known to use the term to mean.  
There are others.  And the term as you construe it is fuzzy, because it 
says nothing about the legal right of modification, independent
development, and distribution of derivative works.


> I really disagree. "Freeware" is "Free as in beer." 

Again, you are confused.  You have misread what I said.  I said that
"freeware" as the term is usually construed is one pole of an axis
on which acquisition cost is denoted.


> "Open Source" doesn't mean Free at all. It just means the source code
> is available to you to tinker and modify if you'd like.

You really ought to read what I _said_ before purporting to disagree
with it.  

I said that the term "freeware" contemplates one point on an axis of
acquisition cost.  I asserted that this particular axis isn't especially
interesting.

I pointed out a separate, orthogonal metric of legal rights to source
forms, right to modify, right to redistribute modified works, right to
use for any purpose.  That is the axis on which proprietary terms lie on
one end, and various types of open source aka free-software terms lie on
the other.


> Yes, although the history is a little more muddled than that.

You don't need to tell _me_ how muddled it is.  And I was around while
it was happening.

This is obviously a newcomer to whom I was addressing.  Were you
expecting me to hit him with the whole damned levenez.com history chart?


> BSD really is more than just the kernel at this point....

Um, BSD was always more than just the kernel.


> It's Unix-Like, and that's usually enough. 

Yes, you seem to be strenuously in agreement, which might be fun but
also a waste of time.  As I said, Tastes Differ[TM].


> Striving for POSIX compliance means there's a principle of least
> surprise when bouncing from one Unix to the next.

Rough POSIX.1 has been shown useful (except of course that such things
as the Microsoft POSIX Subsystem for Windows shows you can achieve that
and produce something utterly useless).

Let me correct that:  Rough implementation of the Core Services
_portion_ of POSIX.1 has been shown useful.  Remember, we don't all need
real-time extensions, and POSIX.1c threading isn't really great.