Older blog entries for yosch (starting at number 12)

I’m convinced that a lack of fonts is a major barrier to an increase in the amount of content in certain languages on the web. Web fonts would go a long way toward fulfilling the “world wide” promise of the World Wide Web.

I definitely agree with Patrick Hall's Blogamundo blog entry as seen on Planet i18n.

Cyberspace is a great chance for many communities using lesser-know languages. Quite a way to go until about 6972 languages become be first-class citizens... But it's good to hear more recognition of the key role libre/open fonts can play. Thankfully many in the open font community are busy creating and providing quality fonts while the various browsers make great progress supporting font linking. IMHO the fulfillment of the promise of a richer web is well underway.

Mobile platforms which use webkit-derived browsers already have support for @font-face CSS font linking: iPhone with its Safari Mobile edition and Android via its own browser component as you can see on Kai Hendry's photostream where he successfully gets a stylesheet calling Andika Basic to work.

Hopefully Mozilla's Fennec (in alpha) will soon offer similar support for the Maemo and OpenMoko platforms.

Besides Ed Trager's amazing font playground (a new version is in the work), I've enjoyed playing with Richard Ishida's Unicode pickers like the IPA one, creating word clouds with Jonathan Feinberg's wordle or admiring the colourful shapes of Ricard Marxer Piñón's Caligraft.

I find it very cool to see the functionality of these webapps being improved by the availability of open fonts like Gentium, Doulos SIL and many others in the background. Open fonts for an open web or something :-)

While embarking on the multi-year journey to draft the OFL and coordinate the expert and community review of the license, we analysed the needs and the existing models, how they worked and how people reacted to them. With feedback from the community we made a list of features which are crucial for good font licenses and tried our best to condense them all into an ideal model for both users and designers to build upon.

These key features are:

• use, study, modification, redistribution (the 4 core freedoms)
• bundling
• embedding and its interaction with possible strong or weak copyleft requirements
• derivative outlines and artwork status
• derivative fonts status
• artistic integrity
• anti-name collision
• name and brand protection
• reputation protection for authors
• preventing stand-alone reselling within huge collections
• descriptive changes of modifications
• clarity and readability for designers
• awareness of the software nature of fonts
• the multiplicity of font source formats, some open and human-readable and some opaque/binary
• good integration with the font design toolkit
• legal solidity through wide expert and community review
• metadata integration
• cultural appropriateness to both the type and FLOSS communities
• stable trustworthy working model with a non-profit as the steward of the license
• being reusable and not project and .org-specific
• allowing linking in a web context (more recently)

There are of course differing views along the licensing spectrum but if you take into account the specific needs of collaborative font design then your criteria may well be in tune with the elements above.

With these criteria in mind, and taking into account the need for reducing licensing proliferation, where do existing font licensing approaches fit in? The following list has some of the existing licenses used for fonts out there along with some quick comments about problems the specific approach may have.

Don't get me wrong: for other uses many of these licenses are brilliant and do a fantastic job and I don't want to ignore the efforts by the corresponding authors or maintainers but with hindsight it seems there are probably better ways to release a font under a free software license.

• Public domain: no rights reserved not even attribution, unclear under various jurisdictions which makes it problematic for a global license, fairly often found to contain elements from restricted fonts where copyright has been stripped
• Utopia license: project-specific and organisation-specific and so non-reusable
• AFPL: Alladin Free Public License: deprecated and rejected as non-free by FSF and Debian
• Various Creative Commons combinations: designed to be used for content and not software
• Baekmuk License: project-specific and organisation-specific and so non-reusable
• Hershey font license: project-specific and organisation-specific and so non-reusable
• Liberation Font License: project-specific and organisation-specific and so non-reusable
• GPLv2 without exception: causing problems with embedding and satisfying source requirements
• GPLv3 without exception: causing problems with embedding and satisfying source requirements
• LGPL v2: confusing in terms of satisfying redistribution requirements
• LGPL v3: confusing in terms of satisfying redistribution requirements
• Bitstream Vera agreement: project-specific and organisation-specific and so non-reusable
• Lucida Legal Notice: project-specific and organisation-specific and so non-reusable
• MgOpen agreement: project-specific and organisation-specific and so non-reusable, a variation of the Vera license
• Arphic Public License: project-specific and organisation-specific and so non-reusable, very closely modelled on the GPL, some clauses are odd in the context of font design
• Design Science License: meant for data and not software: not endorsed by the FSF
• Mincho License: project-specific and organisation-specific and so non-reusable

If you know about font designers wishing to release their creations under a community-validated font-specific license, then I'd recommend you point them to the OFL FAQ and use the Go for OFL campaign materials to advocate a common license which many in the FLOSS community believe caters better to these needs.

At the last EuroBSDcon which happened in Strasbourg, FR, there was a talk about using TeX in an *BSD environment (initially focusing on OpenBSD) by Ed Barrett. (Slides and audio are available).

A rather nice introduction to the wonderful world of typesetting with TeX with mentions of the great work going on porting TexLive to OpenBSD including Jonathan Kew's XeTeX and TeXworks along with demos of using various native OpenType fonts from the Open Font Library or from SIL with all the corresponding font freedom and Unicode goodness :-)

IMHO such cross-OS efforts around the Free Software typesetting stack are incredibly useful for everyone.

So another big shout out from me over here to the various maintainers for their dedicated efforts in porting/testing/packaging the whole stack like Karl Berry of TeXlive, Norbert Preining and Frank Küster of Debian and obviously Edd Barret from OpenBSD. And also Yellowshift who know about the profound meaning and benefits of giving back! Thank you all.

George Williams - typographer and font toolkit developer extraordinaire - has improved the great integration between the fontforge font editor and the OFLB Open Font Library (or OFLib). Which means that directly from your font editor (cvs version at this stage) you can search and retrieve existing open fonts and upload back your changes or your own creations. And with all the needed tags, font metadata, FONTLOG and specimen image included. Collaborative open font design to a whole new level. Brilliant :-)

Fontforge's changelog is certainly worth watching!

moyogo, IMHO they should really fix the licensing before it can be accepted by any serious distro...

Technically if the metadata stays as it is now (the standard Ascender EULA), use is restricted and designers can't branch or send patches back...

The fonts are in the base.git repo under data/fonts. There are README and Notice files mentioning the Apache v2 license but this is not reflected in the font metadata. The font metadata actually contains contradictory information!

The Ahem test font does not have any license information attached to it. No way of knowing who the upstream is. A FONTLOG should really be added.

And the Droid fonts themselves still refer to an unknown external EULA in the License Description field: "License Description: This font software is the valuable property of Ascender Corporation and/or its suppliers and its use by you is covered under the terms of a license agreement. This font software is licensed to you by Ascender Corporation for your personal or business use on up to five personal computers. You may not use this font software on more than five personal computers unless you have obtained a license from Ascender to do so. Except as specifically permitted by the license, you may not copy this font software.If you have any questions, please review the license agreement you received with this font software, and/or contact Ascender Corporation. Contact Information:Ascender CorporationWeb http://www.ascendercorp.com/"

A standard link to the Ascender restricted EULA is provided in the License Info URL field: http://ascendercorp.com/eula10.html

Not the kind of choice that will build a collaborative font designer community around this to make the Android stack more multilingual...

OTOH If this is fixed I really doubt many font designers have heard about or read through the Apache v2 license. IMHO it's in the interest of Google and the Android community to pick a community-recognized font-specific license like the OFL to provide a better collaboration mechanism to extend the font and to take care of embedding/naming problems.

I'll get in touch with the Android folks about this...

I find it both heart-warming and mind-expanding to learn about a new writing system... But this one, well... we're reaching a whole new level!

A little while ago an experienced linguist working with sign languages around the world sent me information about an open font project designed specifically to enable sign language writing systems: ISWA 2008 (International SignWriting Alphabet).

The writing systems implementation (in which this font is a key element) uses a combination of iconic symbols to represent face, hands and body movements with meaning: it's called SignWriting.

And you have to recognize that the authors have their licensing strategy well thought out for maximum distribution and collaboration: GPL 3 for the software, OFL 1.1 for the fonts and CC BY-SA (Attribution-ShareAlike) 3.0 for the content. IHMO a great combination indeed!

Pretty amazing to see such a collaborative Free Software approach helping enable yet another writing system to be created and widely used. Sounds like a great step forward to bring increased literacy and means of expression to the deaf communities :-)

Now to learn some signs and try to write them down...

Today I flashed my XO (gotten through the G1G1 program from OLPC) to release 8.2 and I'm very pleased to see all the progress made with the whole stack. There are huge improvements to Sugar: nicer navigability and usability, higher responsiveness, a wider set of improved activities, tremendously improved neighbourhood view/network management, and of course the localizers/translators have done great work to reduce the language barriers.

In short there are a lot of very nice touches here and there to make the whole experience even better for the many children out there. Big kudos to all involved :-)

Besides Dejavu with its own peculiar license, various quality open fonts released under OFL are currently available by default. You can see choose them from the Write activity (Abiword-based):

• Abyssinica SIL
• Mothanna
• Thabit

And using the Terminal activity you can easily install extra open fonts packaged by the great Fedora Font SIG from the repositories via yum:

• Ecolier Court
• Ecolier Lignes Court
• Gentium
• Gentium Basic
• Andika Basic
• Padauk
• Libertine
• Jomolhari
• Inconsolata
• GFS Artemisia
• GFS Baskerville
• GFS Bodoni
• GFS Bodoni Classic
• GFS Complutum
• GFS Didot
• GFS Didot Classic
• GFS Gazis
• GFS Neohellenic
• GFS Olga
• GFS Solomos
• GFS Theokritos
• Edrip
• Doulos SIL
• Charis SIL
• Brett
• Silkscreen
• RoadStencil

We'll see how some of these open fonts can be included by default in the deployment target where it makes sense.

High-quality redistributable and modifiable fonts certainly have a key role to play in the education goals of such a project. This is where a community-recognized license like the OFL taking care of issues like embedding, naming collisions and artistic integrity - allowing them to be used without problems as well as further maintained by the open font community - is crucial.

SugarLabs and OLPC have huge challenges to tackle in terms of i18n, so I'm glad to see the progress made in enabling wider language support by picking open fonts allowing designers to extend the fonts to meet local needs.

It's also a joy to see how beautifully the fonts render on the tiny but high-res XO screen and to imagine the many kids learning to read and write in this environment. Seeing the beautiful shapes of your mother-tongue script has to be a good motivator to learn...

moyogo, I'm really looking forward to more info about your work on fonts for African languages... So what about the reactions to your talk and the FLOSS culture/methodology aspects? Follow-up discussions?

How's the atmosphere of the conference? Beside some pictures of the conf it's good to know some talk slides will be up on the site :-)