1 Oct 2010 yosch   » (Master)

Thoughts on the public release of the Ubuntu font family

OK, Mark Shuttleworth, has now announced the first public release of the Ubuntu open font and the buzz about all this is starting to appear across the interwebs... and I feel I need to add some background to various discussion items and bring in extra perspective on important aspects:

Before I talk about the licensing issues I think it's very good to see how various recommendations from me and others in the open font community are now being followed in the structure of the source tree: detailed FONTLOG, explicit naming of font files, double-checking and getting all the metadata right in both binary fields and human-readable headers, releasing various source items alongside the final ttfs with documentation of the formats and how to use them, trademark guidelines, documentation (more coming). Thankfully what has now been released - on track for the RC and the upcoming 10.10 release - is much much more than just a unreviewed code drop over the wall like it has too often been the case for many other commissioned fonts in the past. So very well done on that front!

It's up to us in the open font community to come alongside established designers and entities involved in commissioning open font design to make them aware of potential issues, to recommend and advocate best practises from our experience and from a FLOSS perspective and overall to mutually learn and share tools, methodologies and as a result improve the overall trajectories of open font projects creation, release and maintainership.

Mark was already well-aware of many of the issues surrounding libre/open fonts from previous face-to-face and email discussions over the past development cycles of Ubuntu. Not doubt looking at the previous commissioning/release/maintainership story of Ubuntu-title and the relationships with the trademark policy and the various community branches was very helpful too. Earlier on, he was one of many reviewers who, when asked when we met at an international conference in Africa (WSIS), shared his own thoughts around collaborative font design and the OFL model as it was being formed, reviewed and refined by the wider community after SIL's extensive research. Over the past few years via more discussions with him and members of his team - and also via input from others like Dennis Jacquerye, Ben Laenen, Simos Xenitellis and Dave Crossland from the open font community - he's realised the need for bigger focus on fonts. And it's certainly good news to see the willingness to tackle that and to put more resources, time and energy into getting it done as best as possible for the benefit of the growing worldwide Ubuntu community. It's also another very significant contribution to the wider FLOSS ecosystem. I personally like reading that he's now becoming "a font nerd" in his own words as indicated in this recent Scribus-related post. Compare that to an earlier post back in 2007...

Even if some people have criticised the way the commissionning has been done, they will increasingly realise how bridging the FLOSS culture and the expectations of proprietary foundries is no easy feat at all. Getting an established for-profit foundry closer to the way FLOSS is usually produced and maintained remains an interesting challenge where few answers are easy... But it looks like this particular commissioning is another big step in the right direction. Although more recently it took quite a few emails (since the discussion at the last LGM) and some friendly pushing to get reactions to my recommendations on the Ubuntu font project specifically, I'm now reasonably happy with the current structure of the open font release and also with the plans going forward. I really appreciate the common efforts of Canonical's design team and DaltonMaag to share their design choices and methodology at the last UDS and through various blog posts, making efforts to keep the community aware of the direction the project is going, keeping people with font interest and experience engaged and the whole conversation going.

More recently Paul Sladen's great efforts to interact with us, triage and tackle the relevant bugs on LP have been a great help. He was also the one taking on the challenge to apply the needed modifications (in midstream) to the final source tree which got released by the foundry. Thanks a lot for your very worthwhile work Paul!

Also Richard Lee's fonttest webapp for testing and feedback on the design itself is pretty amazing and has great potential. It was good to see how the need for a dedicated feedback and release process pushed the need for new Launchpad features beyond the usual bug report channels and teams. I think the phased beta and the private repository helped with not getting overwhelmed by unusable feedback. Various people are looking forward to enjoying more of Launchpad's features for their font projects.

My feeling is that over the past few months, there has been quite a turnaround of awareness in many foundries towards the increasingly interesting potential of getting commissioned to do open fonts. I think that DaltonMaag is now a leader in that peer group (although big foundries like Ascender and Paratype have already released open fonts in the past but via a private process). As Mark has indicated, Bruno Maag and his team will be looking more closely at the open font design toolkit (apt-get install open-font-design-toolkit) in the future to review and look at interesting challenges for improvements and solidification based on his expertise. We have plans for working these comments into a usability and workflow review of our existing design toolkit which everyone should benefit from. Stay tuned for work from the LGM community on this...

Concerning the licensing issues themselves, I really appreciate the conscious efforts to work upstream with SIL as the author and steward of the license. Getting the friendly folks at SFLC involved again privately with me as the direct correspondent is certainly a very good thing. But our explanations and advocacy of some aspects of the existing model were not convincing enough to keep Canonical from engaging in creating their own solution. Given the timeline constraints and upload requirements in the end they decided to do their own version of the license but thankfully with a view to only being an interim solution with serious hopes of getting it merged back in the future.

Via last minutes late-night comments, recommendations and fixes to make this fffork^Wbranch as appropriate as possible for this interim phase of the release of the Ubuntu Font project, we thankfully got Canonical to respect the requirements of SIL as author and steward of the license as clearly expressed in our FAQ to differentiate the Ubuntu Font License from the original OFL:

8.4 I really like the terms of the OFL, but want to change it a little. Am I allowed to take ideas and actual wording from the OFL and put them into my own custom license for distributing my fonts?

We strongly recommend against creating your very own unique open licensing model. Using a modified or derivative license will likely cut you off - along with the font(s) under that license - from the community of designers using the OFL, potentially expose you and your users to legal liabilities, and possibly put your work and rights at risk. The OFL went though a community and legal review process that took years of effort, and that review is only applicable to an unmodified OFL. The text of the OFL has been written by SIL (with review and consultation from the community) and is copyright (c) 2005-2010 SIL International. You may re-use the ideas and wording (in part, not in whole) in another non-proprietary license provided that you call your license by another unambiguous name, that you do not use the preamble, that you do not mention SIL and that you clearly present your license as different from the OFL so as not to cause confusion by being too similar to the original. If you feel the OFL does not meet your needs for an open license, please contact us.

Yes, licensing proliferation is a pain for everyone and as you can expect SIL sees serious issues with the way the Ubuntu Font License is worded. Some of the concerns of Canonical do not have anything to do with the strength of the OFL from a legal standpoint, but rather reflect their opinions on how an open font model should work. But this approach can be an interim solution provided it is sufficiently differentiated from upstream and great care is given not to introduce any confusion and hostility against existing users of the OFL. And I think Mark and Canonical have shown that they want to do that which is good.

What does this mean for the many existing users of the OFL, both non-profit and for-profit? Canonical had specific needs that they felt were not sufficiently clearly addressed in the current OFL. Do the many users of the OFL - both non-profit and for-profit - need to be worried? Not at all. Should every well-meaning and ambitious user of the OFL start making their own? Please no! Canonical's concerns are very specific to the structure of the Ubuntu project, and are related to their strong opinions on about how they anticipate their own project will evolve.

The status of the OFL is that it still remains the only font-specific license having undergone proper community-wide review and having been validated by the FSF as satisfying the 4 freedoms and Free Software definition, by the Debian ftpmasters as satisfying the DFSG and by OSI as satisfying the OSD. The OFL is still the font license which key organisations in our community have joined to recommend to upstreams through the Go for OFL campaign. And in the webfonts scene the OFL is still the licence under which the vast majority of quality open fonts are released and available via the Google webfonts Directory and API or kernest for example. Mozilla is also discussing how to use more OFL-ed fonts across their websites. With all the recent buzz about LibreOffice this week, it's also worth noting - if you didn't know already - that the source tree contains various open fonts under the OFL: Gentium Basic, Gentium Book Basic. And obviously Magyar Linux Libertine/Magyar Linux Biolinum is on track for inclusion with all the amazing work done on high-end smart typographic features by László Németh (thanks to Graphite). Similarly OOo4kids - the version for children and young students - bundles Andika and Ecolier to offer increased legibility. These open fonts are especially useful in a context where literacy learning is the goal (this is a request from teachers using this resource). Another example is how the very talented and dynamic Fedora design team has recently reviewed available offerings and decided to use two fonts under OFL for their branding and visual identity: Comfortaa and Cantarell. More recently various OFL fonts have been designed and released by individuals and companies with no such concerns: Google's commissioning of Ascender for the Chrom*OS fonts being one example among many. No cause for concern or need to make an incompatible derivative in any of these projects.

The latest SFLC review, kindly supported by Canonical, has not turned up any problems and we see it as just another potential opportunity to clarify some items and think through some long-term issues together with others. We remain open to recommendations from lawyers like James Vasile at the SFLC. BTW SFLC had already given support to the license model back when the community review was going on. The favourable listing of the OFL (both version 1.0 and 1.1) on the FSF licensing list which resulted from SFLC approval remains. As indicated in our FAQ we will certainly consider ongoing improvements to the well-established OFL model, and continue to have productive discussions with the SFLC and others. At this point we have no definite changes in mind or a timeline for any future version of the OFL, and will be sure to involve the wider community in such developments. Any revision process for potential future versions will be undertaken with great care - as has been the case in the past - and previously existing licenses would remain in effect. Any potential changes would need to support, not diminish, the basic principles as set out in the currently used licensing model. Any changes will be made in the spirit of the licensing model and in the interest of all users. No retroactive changes are possible. Going forward SIL - as the license author and steward of the OFL for the community - will look at the community process and will not be in any way pressurized into catering to the needs of one entity over those of the whole community of stakeholders. We should certainly be optimistic but we will also be realistic and protect the existing values and commons of the open font community.

No designer or contributor outside of the Ubuntu font project itself should use this interim license, because it has not been community-validated and it is not recognized as valid by the FSF, Debian or the OSI. And with Ubuntu in the name itself it's very clear that it's organisation and project-specific.

I do wish the best of success to the Ubuntu font project, now in its public phase. A quality open font with an ambitious scope and expansion program will surely be a great enabler for the various language communities using Ubuntu worldwide. The concept and spirit of Ubuntu is certainly linked to enabling humanity to enjoy our big variety of writing systems! I'm probably not the only one who thinks that quality open fonts with wide Unicode coverage available in Ubuntu is in tune with the Ubuntu philosophy particularly We believe that every computer user should be able to use their software in the language of their choice. I intend to continue contributing to this project as appropriate.

Finally Canonical and other stakeholders in our diverse open font community and ecosystem have kindly indicated they would like to practically support and fund ongoing work on sustaining and growing the open font community and its work on libre/open font licensing as they benefit from it. If you also find this is something worth thinking about I recommend you take a look at http://funding.open-fonts.org.

Latest blog entries     Older blog entries

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!