Older blog entries for yosch (starting at number 142)

Sent from my $DEVICE

Certain mobile devices add their marketing signature by default to outgoing emails. I find this virtual positional goods statement rather annoying, even if unintentionally left as the default configuration. It feels like posturing as well as showing you're unable to configure your device to personalize your own signature and remove corporate ads.

Is the content of your email not discredited by affirming that a $DEVICE is yours, all yours? Does this trigger to compare $DEVICE against $DEVICE bring anything at all? Are you a peon in the platform wars? Especially since I already know what platform and email system you use and can react accordingly?

The best potential comebacks range from sarcastic to constructive:

• “Sent from my subterranean luxury bunker.”
• “Sent from my $DEVICE… while eating caviar… on a yacht.”
• “Sent from my re-flashed toaster.”
• “Sent from my microwave oven.”
• “Sent from YOUR $DEVICE.”
• “Sent through OUR shared interwebs.”

A look at the fonts in Android 4.0 Ice Cream Sandwich: Roboto and Droid


Plenty of news sources have picked up on today's Android Ice Cream Sandwich SDK release (AFAICT no sign of full source release yet) and among all the various nifty features in this version there's Roboto (Regular|Italic|Bold|BoldItalic) a new font developped in-house, IOW not commissioned via an external foundry. Apparently Christian Robertson is the designer. He is also credited as the author of the AndroidClock font for version 3.1. It looks like an original design (with smallcaps included).

But unfortunately there's no license in the metadata but only the very basic copyright and trademark statement...

IMHO all the smart folks in the Android team should know better than distributing something without any indication of actual licensing, you know it's going to go around the intertubes very fast: just a few hours after the release, plenty of sites offer a download of standalone files which may or may not be a version of Roboto. It's out in the wild but nobody knows exactly if and how you can actually use it, branch from it, etc...

The rather misguided "font data copyright" is still there although fonts are software, the fsType embedding restrictions are set to Preview & Print embedding which will limit various use scenarios of the font, there's nothing in the description field, no upstream URLs and no FONTLOG inside or outside the font. Could do better I guess.

Also in the new SDK I can see that the Droid font family has been expanded with support for Armenian, Ethiopic and Georgian. Great news for users of these writing systems in Android! The Droid fonts now explicitly indicate in the metadata that they are licensed under Apache 2.0 (which wasn't always the case but was thankfully fixed) but today they still have the fsType embedding bits set to Editable embedding which also limits various uses of the fonts. At some point they were potentially going to be available under the OFL as well but apparently this has been put on hold.

So here's to hoping that in upcoming versions the Android team will indicate licensing intent more clearly, fill in the useful metadata fields and fix the embedding buglets in these fonts. Android users and people who may want to use these fonts elsewhere thank you in advance.

Overview of advanced typographic features in LibreOffice

Don't miss the slides of "Towards Desktop Publishing" by László Németh from fsf.hu given at the LibreOffice Conference in Paris by András Tímár.

It's a great overview of all the new advanced typographical features available in LibreOffice.

Awesomely beautiful !
A huge well-deserved kudos to all the people involved !

New open fonts projects

Here's yet another small batch of upstream open font projects I recently heard about:
Anka/ Coder, Khottabun, Euler , Punk Nova, AnjaliOldLipi.

Also, the big and ambitious Lohit font collection was recently re-released under the OFL by Redhat to make commmunity maintainership easier.

Check out the upstream websites for all the details about the scope of these projects and how you may benefit from them (and possibly contribute).

Please don't use flags on your multilingual website

As the web becomes increasingly multilingual and more domains host content in various languages, an anti-pattern tends to surface: the use of flags to represent sections in different languages.

As more content gets written (or translated) and added to a website, the section and navigation system gets redesigned accordingly. But the underlying assumptions about hosting multilingual content and the best way to present it to a bigger and more varied audience are often not being properly thought through. I imagine that this is not just because monolinguals or ethnocentrists may be in charge but because i18n best practises usually come as a afterthought.

If you design your website with a navigation system forcing users to go to their preferred content by clicking on a flag, you're making a very infortunate statement about the classification and power relationship between languages and in the process you're very likely to alienate people visiting your website or making use of your webapp. Consider how you'd feel if you had to pick the flag of a foreign country to get to the section most relevant to you? For example if you were Belgian and had to pick the French flag to get to the content relevant to you? Cameroonian and had to pick the UK or French flag? Australian and had to pick the US flag? Swiss and had to pick the German, French or Italian flag? Or Taïwanese and had to pick the Chinese flag? See the pattern and how this is a big can of worms that makes people uncomfortable? Unlike what powers-that-be would like you to believe, flags don't map to languages and country borders don't correspond to linguistic communities. A flag represents a country and many many countries use the same language or have more than one official language (and often more lesser-known languages not officially recognized at the country level).

Instead of flags you should be using the name of the language in autonym form and the corresponding two-letter or three-letter code for the language as standardized in ISO 639.

More details in the W3 i18n best practise WG note and in the "flag or no flag language links" article by Motiva web consulting.

A Debian-branded pocket multi-tool

I've carried around a Victorinox WorkChamp in a belt pouch for ages (more recently along with a mobile device running mostly FLOSS). Now I think my next multi-tool may well be Debian-branded.

An obvious way of bringing together two famously solid, sharp, practical and malleable DIY multi-tools.

"Tools of the trade" or "tribal markings"? Both!

OS X shipping more open fonts by default with Lion release (10.7)

I find it interesting to see that while the previous version silently shipped Menlo - a derivative of Bitstream Vera - as the default monospace, the trend picks up in a clearer way as the recently released OS X Lion (10.7) now ships with two big open font families : Stix and Nanum, both major projects commissioned to be released under the OFL (along with various unamed Non-Roman fonts for Devanagari, Gujarati, Gurmukhi, Tamil, Hindi, Marathi, Nepali, and Panjabi, Bengali, Kannada, Khmer, Lao, Malayalam, Myanmar, Oriya, Sinhala, Telugu and Urdu... No word on their licensing though).

This choice contributes to show how sometimes the quality and scope of open fonts are too good to pass and it gets a vote of confidence by being explicitly described in the release notes.

Explicit support for the OFL in Gforge/FusionForge

Looks like newer versions of the Gforge and FusionForge "software forges" have updated their classification to include explicit support for the OFL. No surprise there as this classification is based on the official OSD-compliance status, but having font designers and font developers be able to pick the license explicitely when using these hosting services is always good.

Even if various projects have moved on to other hosting platforms, G|FusionForge is still being used by many upstream communities. This will facilitate collaborative font development and also make it easier to find out about new open font projects.

Future creative shifts around the (open) font design toolkit?

While many designers are eagerly waiting for the promised FontLab update, it seems other tools are looking to challenge the status quo and redefine what it means to do font design with digital tools:

According to the abstract of this upcoming AtypI talk, a spiro-based web-based collaborative font editor is in the works and is to be released under an open source license. Which license will be chosen and how well the software will work in a self-hosted scenario and outside of a "cloud" infrastructure remains to be seen.

Looks like earlier experiments in the area like Typism have not really taken off. We'll see what happens.

Another tool taking a different approach (albeit proprietary and platform-specific) is Glyphs.

The way open font design best practises shape and support these new paradigms - and obviously the underlying licensing and collaboration culture - will be interesting.

In the meantime apt-get install open-font-design-toolkit.

Mozilla's font inspector API

Yesterday's Mozilla platform meeting minutes has this:
"Jonathan Kew’s font inspector API landed. You can now write extensions that will tell you exactly what font(s) are being used to render any given DOM range including the entire document, or a single character. For downloaded fonts you can get metadata including the license."

See bug #467669: Need chrome-accessible API for getting list of font faces used by content and the corresponding one in Firebug: bug #3071 Display sample of WOFF file inside Net Panel.

Yet another huge step forward for webfonts and webdesigners! Major kudos for making this happen!