Re: Overstrike mode for key driven ups?

From: William F. Dowling (
Date: Mon Oct 22 2001 - 14:48:19 BST

Russell Browne writes:
 > [...]
 > I have a hard time imagining how <tab> would move focus in a reasonable
 > way.  Every displayed value, array index, expression, and breakpoint
 > is a potential editable field.  Would you have to tab through everything
 > in the display window to get to a breakpoint you wanted to edit?

You wouldn't have to tab within the display window.  As Flemming
suggested arrow keys can be used within one logical area (display
area, target menu, source menu, etc.) And you still have the mouse.

But I agree there is a real problem with using tab to move focus
generally: you shouldn't be allowed to tab into any area you can't tab
out of, i.e. any area where tab is a legal input character.  So I
think tabbing works to move among button groups (menus), but not the
display/source/output areas. So, amending earlier proposal,
<ctrl><tab> toggles one-key-menu-state; in one-key-menu-state, <tab>
moves among menus, arrow keys move within selected menu, single key
(s/n/etc./<space> for repeat) activates selected menu item.

 > I can imagine a typing region that would take single keystrokes and
 > do the most common things.  Since it only needs to display a single
 > character it need not be very big. 

This would be fine too.

 > Perhaps a new region between the <<
 > and Help  buttons, and a little bit bigger that the << button.  I would
 > have the region redraw itself in some fashion when active to give
 > visual feedback.  Perhaps gray it out when inactive and gray out
 > the typing line when the new region is active. I would have a <tab>
 > toggle focus between the two regions, 

This means you can't type <tab> as data into the typing line.

 > and would move focus also with
 > a mouse click.  I hesitate to have it grab focus when the mouse
 > enters the region because the mouse could wander out unintentionally,
 > especially if it is a small target.

William F. Dowling
ISI/Thomson Scientific (
215-386-0100 x-1156

This archive was generated by hypermail 2.1.4 : Wed Feb 13 2002 - 21:51:34 GMT