X
Tech

Old ideas still needed

Maybe there isn't a huge market in voice recognition, or in the OCR realm. But there is a great need for it. And these are the kinds of projects where open source can make real change happen, in the lives of real people.
Written by Dana Blankenhorn, Inactive
tillie.jpg
Google has helped bring Tesseract, an OCR program originally developed by Hewlett-Packard, to the open source world under the Apache license.

A ZDNet story on the move hints that Google may already be working on a fork that could come out without some of the commercial restrictions found in the present software.

Tesseract isn't a great OCR program. It was basically orphaned when H-P got out of the OCR business back in the 1990s.

But it's part of a trend worth discussing.

That is, the crying need for open source help in the area of interfaces.

The transformation of text into something readable by the computer is just one such area. So is the area of transforming voice. (I have a dog in this hunt, since my mom (above) has been legally blind for about 25 years.) Since IBM donated some of its voice recognition code to GNOME and KDE under an Apache license, two years ago, the former has pushed a number of accessibility initiatives, while the latter has added voice to KHotKeys.

All well and good. But I keep thinking of a 20-year old movie clip, of the late James Doohan trying to use a Mac, picking up the mouse and saying "computer" to it, then tut-tutting on being told he has to use the keyboard.

He still does. We all do. And we are all weaker for it.

Maybe there isn't a huge market in voice recognition, or in the OCR realm. But there is a great need for it. And these are the kinds of projects where open source can make real change happen, in the lives of real people.

Do it for mom.  

Editorial standards