---------- Forwarded message ---------- Date: 18 Mar 1999 18:10:24 -0000 From: Lar Kaufman <lark@xxxxxxxxxxxxx> To: blinux-develop-request@xxxxxxxxxx Subject: bomb shell I apologize for making remarks that may have appeared "political" in trying to examine what I perceived to be a political bias. The suggestion of political arguments made me rethink an assumption on my part, however. I'd like to suggest a different line of examination regarding the basic user interface that might once and for all preserve user access to the OS, if I could get some feedback from "those who know" about some specifics of this and that technology, since I think I may have a novel solution to the accessibility problem. Instead of asking "what other non-proprietary solution offers as flexible solution to user access as emacs" I should have asked "how can we give Linux low-level support for a rich, non-proprietary user interface?" The answer that came up when I asked that was not emacs, but XML. If Linux were capable of presenting information in the form of XML structures, it would be a simple matter to plug in, at the shell, desktop, or application level, a web browser interface; the operating system would not have to make assumptions as to the user's mode of access except to deal with the installed hardware in traditional Unix ways. Let me hasten to add that full XML need not (probably should not) be supported, but useful structural markup should be supported. For example, Linux would not send a message tagged as "bold" information, but it could send a message tagged with "emphasis". And Linux need not then concern itself with how the information it is presenting will be rendered for the user, only that it was sent to devices that are supported by default or user-chosen utilities to render it appropriatedly. The tricky part is to define what subset of XML tagging is *necessary* for OS-user interactions and what tag handling (recognition/stripping/ignoring) Linux will perform on input from the user. And, if necessary and appropriate, defining new tags to submit to the W3C to support open systems accessibility at the lowest level. Getting an English voice message output to a sound card, for example, would not be difficult, but getting a vocal response returned to Linux and recognized as a choice or command is a tougher problem that might best be handled by making the user's shell, desktop, or browser smart enough to process the signal into the needed data for Linux to use. Does this seem viable without significantly distorting the kernel? -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