[OpenFontLibrary] Bruce Perens criticized SIL Open Font License

Dave Crossland dave at lab6.com
Tue Apr 7 04:46:45 PDT 2009


2009/4/7 Nicolas Spalinger <nicolas_spalinger at sil.org>:
>>
>> Basically, his opinion is that the license does not make clear that
>> even though the OFL effects do not extend to works OFL material is
>> embedded in, the embedded material is still protected by the OFL. So
>> according to him one could strip licensing of any OFL font by
>> embedding it then extracting the result.

I flip flop on this; my gut feeling is that SIL is right, but Perens
has a point. I can't tell until a lawyer voices their opinion.

> His statement that SFLC attorneys (or others in
> the community for that matter) have not looked at the license is simply
> untrue.

Could you state which ones did? :-)

> I think you will all agree that an embedded font (an actual part of the
> document it is embedded in) is not fully bit-identical to its original
> upstream format:

Its a lossy conversion, but it isn't that lossy; there is an echo of
the idea that final OTF files are 'not lossy enough' compared to the
true SFD/VFB source files to necessitate their provision to users.

I heard a rumour that the Reading MATD portfolio site at
www.typefacedesign.org has PDFs from which fonts were extracted and
posted on some Chinese or Russian forum (communists, eh? ;-) The
rumour was that the rips were of much lower quality than the student's
real, unpublished, OTFs - incomplete metrics, glyph encodings wrong,
that kind of thing.

But they could be brought up to reasonable quality by anyone who knows
how to do type design.

And that is really the point of the OFL2PD process that Perens warns
about; that Person A published a medium quality OFL font, Person B
makes a PDF with all the glyphs in, Person C rips the glyphs out of
the PDF and publishes a shitty quality Public Domain font, and Person
D tunes up the PD font into something of high quality and publishes it
as a proprietary font. Easily enough Persons B and C are in fact
Person D.

I believe that this scenario is misguided because proprietary-minded
type designers would never do this, because their principles direct
them not to. They genuinely believe that developers have a inalienable
right to control the users of their works, and this means they respect
free software licenses' control of the uses of such works. Albeit
grudgingly, since there is an obvious political difference, haa ;-)

And so I don't think this problem, while _theoretically_ valid, is any
cause for concern.

Since there is no (culture of) automatic upgrade clause in OFL fonts,
1.2 would be annoying anyway - when you wanted to mix two OFL works
you have to check the license version precisely, so the less versions
of OFL, the better; 1.1 is fine by me.

> It's well-worth noting that Bruce's concerns didn't prevent the OSI
> board from now officially recognizing the OSD (Open Source Definition)
> compliance of the OFL v1.1 and listing it on the approved licenses page:
> http://www.opensource.org/licenses/alphabetical

Hah! I didn't catch this! Well that is good news! :-)

Although, I suspect that Liberation was GPLd because OFL wasn't OSI
approved (and Red Hat has contractual obligations to only published
software with OSI approved licenses, I think?) and since I advocate
strong copyleft fonts, I'll be sad if Red Hat starts OFLing fonts in
future.

> Now, how do other font license who claim community input and review
> actually deal with the extraction issue? That would be our next
> challenge... Though experiment: By Bruce's line of thought any random
> person could include all GPL-ed fonts inside a zip and copy them back
> out to re-release them... The current GPL font exception certainly does
> not cover that very clearly (and I'm not talking about the other
> problems). All the source requirements (as hard to define as they are)
> are now gone!

I've come to the conclusion that until we get lawyers to suggest how
to deal with this, we are wasting our time :-)


More information about the OpenFontLibrary mailing list