When Louis Braille, back in 1821, first developed a system of forming letters from rectangles of embossed dots, he probably was not thinking too much about how to display letters on tactile devices hooked up to a computer. In the last 100 years, Braille has gone through content revisions mostly relating to syntax and typography, but the basic design of the embossed alphabet has remained the same. To accommodate users for whom audio computer screen readers were not an appropriate solution, braille displays --tactile consoles that act as computer monitors and read monitor output line by line-- were developed.
Braille displays, while marvels of adaptive technology, have their drawbacks. If one goes tactile display shopping, the first obvious problem is their price. While the modern visual monitor can run into the thousands of dollars for large and exotic displays, the average monitor can be purchased for $100 or less (I recenty bought a 19" LCD monitor on eBay for less than $80). Braille displays can easily cost over $3000. The size of a 40/80 character display can more resemble the keyboard of a musical instrument than a computer keyboard and, while there are some smaller more portable displays available, the typical workstation display is a stationary device
Braille embossers (printers) are available if one wants to save and print data to have a permanent record or to have a copy of a document to read when one isn't sitting at a workstation. Braille embossers, like the tactile displays, are also big, heavy and expensive, but they do allow tactile information to be portable.
It is one thing to have portable information but quite another to manage that information. With so much information available all the time on the web, it is difficult for anyone to determine what information they want to read and how to get that information in a sensible form. When you remove the ability to see information as it is, by default, provided, the difficulties encountered jump by an order of magnitude that a sighted person can hardly understand. Technology, sometimes the devil that inflicts all this noise on us, is also sometimes the angel that helps make sense of what we need.
RoboBraille is a service provided by the Synscenter Refsnaes, the national center for blind and partially sighted children and youth in Denmark. It provides a unique e-mail-based service that translates documents from text, html, xml and .doc formats to mp3 encoded speech or contractual Braile, by e-mail. You can create a document in a supported format, submit it to one of RoboBraille's many specially purposed e-mail addresses as an attachment and, depending on the document size and RoboBraille's workload, the translated document is returned via e-mail. The tests I made around midnight eastern time, returned a one page document in about 2 minutes. There are many supported translation formats (and just as many e-mail addresses) that accept e-mail. A drawback (for me) is that the site is entirely in Danish (they do support English documents to English braille tranlation). After struggling to use my college German to figure out the site, I used a Google translation to decipher the pages. One file I submitted to be translated from text to an audio mp3 format generated an error response from RoboBraille. The e-mail RoboBraille returned to me was entirely in Danish and the best I could do was figure out that, for some reason, it just didn't work.
Translating documents using a free e-mail service is a step in the right direction and I greatly admire what the RoboBraille project has accomplished, but I've been looking for a more comprehensive solution to tactile information distribution. I found an important piece of the puzzle in Sébastien Sablé's utility, Libbraille. Libbraille's source code is freely available and after some fits and starts, I was able to configure and compile it on a Unix server.
Libbraille is not a standalone program, but a library of functions that allow other programs to make use of its translation utilities. When compiled into a program as an option, Libbraille allows that program to produce its output in braille (it supports French, German, British and American English constructs). Libbraille will directly produce output on a braille display device as well as encode its output in a variety of formats, including plain text.
I've grabbed the rss stream of Serendipity35 from the past 2 weeks, piped it through Libbraille's text output and used Duxbury's braille translation software to create an embossable version of the feed. I've taken thiis post and converted it to produce a braille single article download. While I have quite a way to go to produce fundamentally accessible Web 2.0 content, solutions are close enough to touch.