[Pkg-fonts-devel] Open Font License 1.1 Released

[Nicolas Spalinger]

MJ Ray wrote:
> Nicolas Spalinger wrote:
>> Many other key reviewers [namedrops] explained that the name change
>> requirement is a desirable feature for fonts and that so-called
>> "ready-to-eat" derivatives are problematic. A branch is something
>> different by definition and it should identify itself as such and not
>> masquerade itself as something else to the user.
> 
> As explained repeatedly, users should not have to configure font
> substitution for every new font 

Mmm, What?

The configuration of external font substitution systems like
fontconfig are outside the scope of the license.

See the updated FAQ entry 2.8

> and especially should not have to
> configure gentium.deb (a simple packaging of gentium, not a font branch)
> as a substitute for gentium!

It would help if you explained what you have in mind a bit more clearly.

Since when is a diff.gz understood as a derivative?

Concerning the Gentium package in Debian, the current maintainer has
been unresponsive for a long time to upstream's requests for various
updates and the availability of updates in the pkg-fonts svn (like
removing the local.conf which is only there as an example). I'll file bugs.


>> The font name
>> protection is a key feature of the OFL to guaranty artistic integrity to
>> a font designer and actually make him consider releasing his work under
>> a free license.
> 
> The fundamental problem is that OFL attempts to use copyright instead of
> other laws (trademarks, fraud, moral rights, and so on).  In common with
> many other licences doing similar things, OFL needs to be used carefully
> to avoid trampling on the four freedoms.

I disagree. Theoretically this reasoning is all good and well and I
would agree with you, but in practice such an approach is needed for
fonts. Ever wondered why we had so few free/open fonts until there was a
good way to reach out to the font designer community with something
which makes sense to them?

As for similarity with other licenses, there are name-related
requirements in various DFSG-approved licenses.

The Free Software Definition has this to say:
"However, certain kinds of rules about the manner of distributing free
software are acceptable, when they don't conflict with the central
freedoms. For example, copyleft (very simply stated) is the rule that
when redistributing the program, you cannot add restrictions to deny
other people the central freedoms. This rule does not conflict with the
central freedoms; rather it protects them."

> Worse, it's a false sense of security for font designers.  

A false sense of security? It's the best way we have found to protect
artistic integrity while allowing derivatives.

> OFL does nothing to stop an unrelated font calling itself the Reserved
> Font Name (because it cannot do anything about it).

Yes, and that's quite normal. If you start from scratch do anything you
want but if you derive from an existing font don't abuse the namespace.

> Names should be controlled
> by trademark if one feels strongly enough to hinder other fonts under
> that name. 

Please re-read the license:
It's not about hindering other fonts but *allowing derivatives* while
keeping the namespace sane.

I'll remind you of the current Debian policy which already has something
about not creating namespace chaos:
http://www.debian.org/doc/debian-policy/ch-customized-programs.html#s11.8.5

"Font packages must not provide alias names for the fonts they include
which collide with alias names already in use by fonts already packaged.

Font packages must not provide fonts with the same XLFD registry name as
another font already packaged."

Even the classic “knuthian” TeX license and its children have similar
requirements.

Please understand that this requirement makes sense.

> However, OFL's drafters seem to think font designers are
> copyright-ignorant or just plain stupid enough to think RFNs are a big
> selling point.

How many designers have you been in touch with before making these very
aggressive statements?  Not exactly the best way to engage with them and
show them the potential of a collaborative approach.

As you know the OFL has been presented at AtypI and we have received
constructive feedback from the design community.

Seriously, do you really think we'd add such a clause if it wasn't needed?

> I feel Condition 3 can be used in a way that follows
> the DFSG, but I continue to ask for its removal.

Great to hear that you think it clause 3 can be DFSG-compliant.

As for the removal, have you read Victor Gaultney's answer on this?

To quote him on OFL-discuss:
"I do understand your point, and even agree that trademark should be
enough. But in practice it's not. RFN is a critical feature of the OFL,
and does not stop designers from declaring their trademarks in addition
to protecting the name through RFN. So people can fully use both."


> I'm happy to note that I think OFL 1.1 is the first version that *can*
> be used for software that would follow the DFSG.

Glad to hear your opinion on the DFSG-compliance of OFL 1.1.

As for the status of OFL 1.0, there are many other opinions, the vast
majority of them recognizing the freeness of the license and actually
welcoming it as a solution.

Anyway, copyright holders can relicense to OFL 1.1.

This is the plan for the various SIL open font families.

> [...]
>> The ftp-masters *have decided in favor of the OFL 1.0* and so we already
>> have various quality open fonts in main (like Gentium, [...]
> 
> When?  Did ftpmaster get asked to consider the problems with OFL 1.0?
>
> This sort of unsubstantiated claim has been made repeatedly, in the
> press, at conferences and elsewhere.  It's apparently based only on
> the acceptance of a few packages.

What?

If the various fonts are in main then presumably a decision has been
made in favor of the license.

> That says little itself: far more
> obviously DFSG-busting licences have snuck into debian in the past!

Yes and at times some debian-legal contributors have been wrong or their
analysis voted against.

AFAIK the ftpmaster team still decides what goes in or not.

> Furthermore, I held back from filing the bug reports against ttf-gentium
> and others because I thought OFL 1.1 was heading the right way and I
> didn't want to cause ftpmaster unnecessary work.  I left a licensing at fsf
> ticket dangling while OFL 1.1 was created, instead of pushing for OFL
> to be removed from their lists.  

Right...

> I feel it is *very* bad behaviour to
> now punish giving OFL time like this.

Well, I'd say your goals of contacting the FSF behind our backs to get a
license-list removal isn't terribly helpful.

> 1.0 was not perfect.  Stop claiming it was.

I've never said it was perfect. (The world is not perfect).

It may not be perfect for some of your use scenarios but it was
certainly good for getting various font free-ed and for engaging
designers and encouraging them to releasing their fonts and making them
available in various distributions.

Let's build on that momentum.

> Irritated,

Sad that you seem to be missing the goals we have here.
Anyway, thanks for the time and energy you have spent on this.


I'd like to hear what other debian-legal and pkg-fonts contributors think.


-- 
Nicolas Spalinger
http://scripts.sil.org
http://alioth.debian.org/projects/pkg-fonts/


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : http://lists.alioth.debian.org/pipermail/pkg-fonts-devel/attachments/20070301/7d4e41c6/signature.pgp