According to Translate.Google.Com

The following paragraph is written in English, according to Google Translate.


Добро пожаловать на страницу проекта Templet - облачного сервиса параллельных вычислений Самарского государственного аэрокосмического университета!
Templet – это шаблон, лекало, модель для построения параллельных программ. Работа сервиса Templet основана на следующих принципах: 


 Here is a screenshot of the page I was looking at.

Here is the screenshot of what Google Translate displays, when requesting the translation:  URL  http://templet.ssau.ru/.

And here is how Bing translates that page.

Not that I like Microsoft, but if their stuff translates pages that Google won't.

Um, Really?!?!?!

ISO 639-3 lists 7,776 languages.

Ethnologue (http://www.ethnologue.com/enterprise-faq/how-many-languages-world-are-unwritten) claims that of their currently listed 7,105 living languages, 3,570 have a developed writing system, and 698 are unwritten, with no knowledge about the writing system, if any, of the remaining 2,839 languages.

SIL lists 451 languages as having died in the last millennium, and 116 languages as having died more than a millennium ago. Inasmuch as primary consideration is given to languages that have been reduced to writing, it probably is a gross underestimate of the number of languages that are either dead, or extinct.

The most recent UBS Translation Report I've found, is from 2007.  As of 31 December 2007,  the Bible (Protestant Canon) had been completely translated into 438 languages. The Bible had been either completely, or partially translated into 2,454 languages.

Wycliffe Global Alliance, which is arguably the largest organization translating the Bible, claims that as of 1 October 2014, the Bible (Protestant Canon) had been completely translated into 531 languages.  The Bible has been completely, or partially translated into 2,883 languages. Furthermore, there is active translation work in 2,195 languages.

ISO 15924 lists 199 writing systems, of which 35 are not included in the Unicode Standard.

The breakup of the Iron Curtain validated Max Weinrich's maxim: A language is a dialect with an army and a navy.

By defining one IETF Language Tag, as one language, one might be able to avoid the dialect/language difference, and the language/writing system difference. By way of example:

• TUR-Latn-TR;
• TUR-Arab-TR;
• TUR-Cyrl-TR;

Those tags are for the same language (Turkish), used in the same country (Turkey), using the Latin Writing System, The Arabic Writing System, and the Cyrillic Writing System, respectively.

And that is how Windows 8.1 claims to support almost 7,000 languages.

A11Y Tools

Earlier today, I got to thinking about devices for those with A11Y requirements.

Speech2Text4Keyboard Input

• Voice input;
• Listen to voice;
• Convert to text;
• Send text to system as if a keyboard;

ASL2Text4Keyboard Input

• Video input;
• Watches video for ASL;
• Convert to text;
• Send text to system as if a keyboard;

Text2ASL4Video Display

• Screen reader reads console input;
• That input is broken into words;
• "Known" words are displayed as a person talking, using ASL;
• "Unknown" words are displayed as a person spelling words out in ASL;
• "Query Mode" can be enabled, to add more "known" words;

Eyes2Keyboard Input

• Video Camera watches;
• Eye Movements are converted to Dasher images;
• Dasher sends text to system, as if a keyboard;

Touchscreen Reader/Presser

• Video Camera watches;
• OCR software watches video input;
• OCR software outputs text;
• Screen reader reads text;

• Voice input;
• Listen to voice;
• Convert to instructions;
• Camera guides the prong in moving to the selected item;
• Prong on device extends out to press the selected item;

Additions to the Hardware

• 10 Key Keypad;
• Joystick;
• Firing button;
• Trackball;
• Camera;

These would be used if the vocabulary could not differentiate between a keyboard value, and something to be written.

Or maybe just to be used as supplemental aids;

• Bluetooth;

Make it a Bluetooth enabled keyboard.

Existing Software

• Sphinx3: voice recognition;
• Dasher: predictive text input;
• Eyegazer: eye-tracking;
• Orca: screen reading;

Needed Software

• ASL2Text: This examines photographs for ASL signs, and outputs the corresponding word;
• Text2ASL: This displays a picture of a person making the appropriate ASL gesture;
• Prong Guidance Software: This is the hard one. I suspect that there is code from a robotics project that can be converted/modified/re-used, that will do this. Worst case scenario is to write the dumb pigeon scenario;

Platform

This is the type of one-off job, that Arduino's were designed for.