The VT100 support words, history package, and so on are very useful. I've been experimenting with changes (so far I've changed a few control- character keystrokes). Now I'm trying to understand error reporting. Does gforth separate parsing of an input line from execution, or not?
I'd like gforth to show simple errors by putting the cursor after the error in the input line. My terminal program (OS X Terminal) can treat the window as a series of arbitrarily long lines of text, perfect for gforth's idea of output. But I only know how to use escape sequences to move the cursor in the current line, not how to back up past a return that has been printed. For parse errors, I think the terminal package can always assume the cursor is still on the input line. For execution errors, the terminal package must assume new lines have been printed, unless there's some reliable way to ask gforth? That's why I'm hoping the types of errors can be separated. Something like curses (where gforth copies the entire window of text and tracks what it thinks the terminal is doing) is not simple like the current design. So I'm planning to avoid such drastic changes for now, even if they allow editing of the entire window. Thanks, -- Derek --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
