[Openfontlibrary] new release of the Ubuntu titling font

[Dave Crossland]

On 12/01/2008, Karl Berry <karl at freefriends.org> wrote:
>     whether they're really Free since according to the license you can't
>     redistribute or modify since you don't have any source code...
>
> That seems like an unwarranted conclusion to me.  Anyone can release
> anything under the GPL (or anything else), if they own the rights to it
> -- that's nothing about the GPL, it's a basic fact of copyright law.
> Ipso facto, it is legal to modify/redistribute/rerelease it starting
> from what was released, since the GPL gives you those rights; given that
> no better form of the source is available, it becomes the source.

If you write 10,000 lines of C, compile it, and give me the binary
"under the GPL," I reason that I cannot redistribute the binary
because I cannot fulfill all the obligations of the GPL to my
recipients when I try to do so. Section 7 of GPLv2 makes this clear.

> Anyway, in general, no argument that a font is more than the basic glyph
> shapes, and the "preferred form" of the source is highly desirable.
> However, I still don't think releasing font outlines is comparable to
> releasing a compiled binary.

For C programs, a compiled object code file that is encoded as binary
data (a.out) has less information than the source code file that is
encoded as text data (foo.c).

For fonts, a "generated" object font file that is encoded as binary
data (font.otf) has less information than the source code file that is
encoded as text data (SFD+XGF for FontForge and XGridFit)

(NB 1: Since the GPL requires the "preferred form" of source for the
work, binary FontLab VFB source files are acceptable.)

(NB 2: Nor does the GPL require that the work depend only on free
software, so FontLab VFB source files are acceptable.)

So they are entirely comparable to me, and your distinction seems
contradictory to me.

Therefore, a GPL font without sources may not be redistributable.

I am not a lawyer, so I really want to speak to a lawyer about this,
and will be doing so very soon :-)

> The font information that is always
> present in a ttf or pfb or whatever is a lot more
> accessible/retrievable/usable than anything you can do with object code.

Font data like point coordinates is, by nature, a machine-readable
kind of data and the binary formats would be okay to reverse but are
published formats anyway (afaik)

This means that a weak-copyleft like the OFL is acceptable and
strategically desirable; that is, highly recommended when the
alternative is "redistribution but non-modification" or worse "non
commercial" bullshit.

However, that is not always the case. For example, the Ubuntu Titling
font is under a medium-copyleft and it makes no sense to me to license
it under a weaker ones. The LGPL sits with an essay explaining why it
is important to not to use it when this is unstrategic for the
software freedom movement,
http://www.gnu.org/licenses/why-not-lgpl.html and I paraphrase:

"We should not listen to these temptations [for weaker copyleft than
the GPL] because we can achieve much more if we stand together. We
free software developers should support one another. By releasing
[fonts] that are limited to [full source code redistribution] only, we
can help each other's [fonts] outdo the proprietary alternatives. The
whole free software movement will have more popularity, because free
software as a whole will stack up better against the competition."

(Also:

I never fail to find it surprising how transparent compiled C programs
are to my friends
with reverse engineer skillsets. They can modify them in a hex editor
substantially - ie, without access to the "preferred form" of the
source - and add new features.

I learned the basics of defeating copy restriction mechanisms when I
was like 15 from print outs of Fravia's pages (no internet access at
home, hah!) A large reason why I and Fravia didn't carry on, and why
reverse engineering has stayed obscure, is that there is so much
"preferred form" for free programs lying around since the late 1990s.
Similarly, the big advances in tools - IDA and UndoDB especially - are
yet to have free replacements, since they are less necessary when you
have source. But if there had never been a GNU project, its likely
that those skills would be far more widespread and the tools a lot
sharper.

However, they would never be as sharp as George's "japanese steel"
'forge, for the reasons we agree on above :-)

> I have never seen an interpretation of any free software license that
> says that independent "scaffolding" such as build scripts have to be
> released.  That's never been the intent, and I don't think such an
> interpretation would hold up.

"scripts used to control compilation" sounds like 'independent'
scaffolding such as build scripts to me :-)

Here is the full relevant text from the GPL:

"The source code for a work means the preferred form of the
work for making modifications to it. For an executable work,
complete source code means all the source code for all
modules it contains, plus any associated interface definition files,
_plus_the_scripts_used_to_control_*compilation*_
_and_installation_of_the_executable_."
-- http://www.gnu.org/licenses/old-licenses/gpl-2.0.html (1991)


-- 
Regards,
Dave