Re: Should we talk about "open source" hardware?

[Valentino Giudice]

> According to the Open Source Initiative's definition, that is NOT
> open source.  It is not their fault that people stretch the term.

To this, I shall add that the standard meaning of "open source" (that
supported by OSI) is recognized by a huge number of organizations, as
well as government agencies. This is on top of it being generally well
known.

The difference between Free Software and Open Source in terms of
software categories are even smaller than those outlined by:
> https://www.gnu.org/philosophy/open-source-misses-the-point.html

The article used to confuse Visual Studio and Visual Studio Code. This
was corrected.

What was NOT corrected is the confusion between Visual Studio Code and
Code - OSS:
- Visual Studio is a proprietary IDE by Microsoft.
- Code - OSS is both Free Software and Open Source, published under
the Expat license. It is a text editor which supports extensions
(becoming effectively an IDE, but unrelated with Visual Studio).
- Visual Studio Code has nothing in common with Visual Studio, other
than the name (this is the mistake that was corrected). Visual Studio
Code is provided only in binary form. It is not Free Software nor Open
Source. However, it is known that most of its codebase is the same as
that of Code - OSS (or at least that's what they claim:
https://github.com/microsoft/vscode/issues/60#issuecomment-161792005).

The error on the GNU website is claiming Visual Studio Code is Open
Source. It is not and Microsoft doesn't claim it is: instead, it
describes it as "*built* on open source".
The confusion stems from the fact that the repo for Code - OSS is
called "vscode". But they clarify that distinction in the repo.

In addition, OSI doesn't seem to have any stance on whether any
executable at all qualifies as open source, as they only use the term
to refer to source code. It's not really an explicit disagreement with
the FSF, just a narrower scope.

Debian uses the same definition of "Free Software" that OSI uses for
"open source". Despite this, Debian is more conservative than the FSF
on what counts as Free Software.

I'd argue that if any difference at alle exists between the two
software categories, it is due to slight differences of
interpretation, and not to the actual text of either definition.
But, more likely, there is no such difference at all and different
communities disagree on those licenses that sit on the very boundary
of that one category. Boundary which will always exist and always be
blurred, as is the boundary of almost any other class humans define.

Obviously the free software movement and the open source movement are
two different movements, based on different principles. And it makes
sense for the FSF to both disagree with and criticize the open source
movement.

But when it comes to differentiating software categories, almost every
time I see someone draw a distinction (obviously people like you, RMS,
are an exception to this) do so based on one of the following errors:
- Defining "open source" as software of which the source is available,
regardless of licensing (should be called "source available").
- Defining "free software" as software which is free of charge.
- Defining "free software" as software under copyleft FLOSS licenses
and "open source" as software under lax permissive FLOSS licenses
(does not require using the word "copyleft").
- Defining "free software" as "software under the GPL".

I've seen the last two misconceptions from people that appeared to
strongly support (rather, they believe they supported, but actually
misunderstood) the free software movement, often as a way to bash the
open source movement and, in some cases, claiming that licenses such
as the Expat license would never qualify as free.

I believe that when it comes to describing software or licenses (and
not movements!) "free software" and "open source" should be used as
exact synonyms in almost every context (which is compatible with
always choosing to pick the former), while recognizing that a few
licenses and programs lie in the boundary of that category, and that
they are luckily a small (and probably ever shrinking) minority.