> It would not be sufficient, therefore, to tell the braille/speech complex to > ignore cursor motion if it's to a specific line, and to always speak changes > on that line. It depends on why the cursor was moved to that line, and on why > the content of that line was changed. Exactly. The "why" of the cursor movement to a line implies a meaning that has to be recognized by the application, and if significant, passed to the appropriate level for processing in the form of structural tagging. It's easy to see that the kernal only needs to see very limited data; most information would be processed at the shell level or higher--largely by choices made by the user. Obviously, a shell could incorporate web browser functionality, especially if it need not deal with graphics rendering. Nevertheless, such rendering could be added to the shell by running a popular browser or by using add-ons just as Navigator, Mosaic, and Explorer do. -lar "The sum of all we drive at is that every man may enjoy the same rights that are granted to others." -- John Locke, 1689, A Letter Concerning Toleration