I'm increasingly getting to be of the opinion that the current timeout behaviour of the menu system is ridiculous. I did it this way because someone on this list had expressed concern that serial port glitches would trigger a character which then would cause the thing to sit indefinitely. Well, first of all, such an automatic system would hardly use the menu system, and second of all, I think it's optimizing for the wrong thing. Thus, I'm proposing -- and I would like comments -- of reverting the menu system to "you press a key - the timeout stops." I might eventually introduce an optional "grand timeout" for both the CLI and the menu (if no selection within time X, boot the default regardless), but it wouldn't make it into 3.11. Yes/no/maybe? -hpa
> Thus, I'm proposing -- and I would like comments -- of reverting the > menu system to "you press a key - the timeout stops." I might > eventually introduce an optional "grand timeout" for both the CLI and > the menu (if no selection within time X, boot the default regardless), > but it wouldn't make it into 3.11.If there is a "grand timeout" option I do not have a problem with reverting the changes to menu.c32. But please introduce the "grand timeout" option before you make any changes. Cheers Alex
H. Peter Anvin wrote:> I'm increasingly getting to be of the opinion that the current timeout > behaviour of the menu system is ridiculous. I did it this way because > someone on this list had expressed concern that serial port glitches > would trigger a character which then would cause the thing to sit > indefinitely. Well, first of all, such an automatic system would > hardly use the menu system, and second of all, I think it's optimizing > for the wrong thing. > > Thus, I'm proposing -- and I would like comments -- of reverting the > menu system to "you press a key - the timeout stops." I might > eventually introduce an optional "grand timeout" for both the CLI and > the menu (if no selection within time X, boot the default regardless), > but it wouldn't make it into 3.11. > > Yes/no/maybe? > > -hpa >If someone doesnt like the existing timeout, they can always hit TAB on any selection. That stops the countdown indefinitely, and escape brings you back, allowing you to choose another item. All else being equal, customers prefer menus, and whether this glitch-prevention timeout was paranoid is irrelevant; its in now, and harmless. (unless its clutter that makes the next feature harder) maybe display/feedback tweaks are in order: while counting down: hit TAB to halt countdown, or to edit selection. while in boot-edit mode, add a hint: hit ESC to return to main menu
> -----Original Message----- > From: syslinux-bounces at zytor.com > [mailto:syslinux-bounces at zytor.com] On Behalf Of Jim Cromie > Sent: Tuesday, August 30, 2005 3:10 PM > To: H. Peter Anvin > Cc: SYSLINUX at zytor.com > Subject: Re: [syslinux] Opinions about timeout > > > H. Peter Anvin wrote: > > Thus, I'm proposing -- and I would like comments -- of reverting the > > menu system to "you press a key - the timeout stops." I might > > eventually introduce an optional "grand timeout" for both > the CLI and > > the menu (if no selection within time X, boot the default > regardless), > > but it wouldn't make it into 3.11. > > > > Yes/no/maybe?I have not used the new simple menu system, but this idea seems eminently reasonable to me. It brings menu behaviour back into line with what you get on the cmdline, it sounds like.> If someone doesnt like the existing timeout, they can always > hit TAB on > any selection. > That stops the countdown indefinitely, and escape brings you back, > allowing you > to choose another item. > > All else being equal, customers prefer menus, > and whether this glitch-prevention timeout was paranoid is > irrelevant; its in now, and harmless. (unless its clutter > that makes the next > feature harder) > > maybe display/feedback tweaks are in order: > > while counting down: > hit TAB to halt countdown, or to edit selection. > > while in boot-edit mode, add a hint: > hit ESC to return to main menuUsers are fickle and resistant to reading things. It is likely that the timeout expires before they figure out how to stop it. -- Michael