Re: [Usability]sft+ctl+w v. ctl+q

[Calum Benson <calum benson sun com>]

On Wed, 2003-02-05 at 00:53, bordoley msu edu wrote:

> 
> quit and ctrl+q should be deprecated.
> 
> Close all windows is really vague, and non-standard, and some of us 
> (including me) consider its presence in nautilus a bug (but i'm not making to 
> big a fuss about it).

Coincidentally, there's a debate about this on the Mozilla accessibility
list just now, and a 187-comment bug report in bugzilla:

http://bugzilla.mozilla.org/show_bug?id=39057

Sample email from the thread attached :)


Cheeri,
Calum.

-- 
CALUM BENSON, Usability Engineer       Sun Microsystems Ireland
mailto:calum benson sun com            GNOME Desktop Group
http://ie.sun.com                      +353 1 819 9771

Any opinions are personal and not necessarily those of Sun Microsystems

--- Begin Message ---

• From: Vaclav Dvorak <dvorakv idas cz>
• To: Aaron Leventhal <aaronl netscape com>
• Subject: Re: Ctrl-Q (quit the whole Mozilla) too dangerous; bug 52821
• Date: Wed, 05 Feb 2003 01:58:45 +0100

Hello Aaron,

Aaron Leventhal wrote:

I'll revisit this whole area with the UI team soon. I agree it's too easy to hit Ctrl+Q and quit the app. Persoanlly, I think a yes/no confimation would be useful.

Aaron, you're a god. I love you! :-)

that people don't read confirmation dialogs is true.

Sure, I'm not disputing that, in general. But when a general guideline becomes a dogma, it's not good.

dialog is good. The mac won't get one, because we're trying to behave

I don't care about the Mac. I've never seen one except in a shop window. ;-) But as far as I can tell from Mac users' comments, it seems the current behaviour is OK for them.

I got to this whole issue late, and it's too painful read through all the bug comments because of excessive whining. Feel free to tell me any points I've missed.

I know what you mean, with 157 comments in bug 39057 (confirmation alert) and 187 comments in bug 52821 (too easy to hit ctrl+q). :-)

I'll try to summarize the important points. Sorry, it's still long, even though it's a summary. ;-)

In bug 52821, several solutions were proposed:

1) removing the Ctrl+Q shortcut entirely except for Mac (you gave a positive review to that one yourself, if I remember well) - was rejected, some people want a shortcut for quit; 2) removing Quit entirely from the menu (and, presumably, the shortcut with it) - this is actually a subject of another open bug/rfe (65121), see further below for comments; 3) changing Ctrl+Q into Ctrl+Shift+Q, except for the Mac - that's the current consensus among the bug commenters. There's a patch for it ready to be applied, but it got negative review by Kathleen Brade (comment 166).

In bug 39057, there's an interesting attachment: http://bugzilla.mozilla.org/attachment.cgi?id=94617&action=view It's a screenshot/collage of Opera's preferences window and exit dialog. Seems like the way to go.

In bug 65121 (Remove File|Exit and replace with File|Close), the proposal is essentially to remove Exit/Quit from the menu entirely (arguments: counter-intuitive (user doesn't expect browser, mail etc. to close), and there's no such thing in MSIE => ie-parity) and move Close to the last position of the File menu. Again, the Mac is an exception. There's one interesting variant: comment 80 (and then 91) suggests to do what the bug summary says, but still leave Quit available on Shift-click of File|Close. This may sound a bit weird, it's actually not so completely unusual: we have shift-reload, Windows has shift-restart and shift-delete (real delete instead of move to trash), plus it coincides nicely with the proposal to change Ctrl+Q into Ctrl+Shift+Q (which could still be done, then). It would, however, call for some way to advertise the Shift-click, which could be in a statusbar message (bug 47434) or in a tooltip (bug 122095).

The above was, mostly, a factual overview (or it should have been). What follows are my personal opinions.

In case we will really add a confirmation dialog for Quit, we must make sure we don't go from one extreme to the other. There are bugs/rfes for warning on quit/close-window when downloads are in progress (bug 28385), when forms are filled but not sent (bug 48333), and when multiple tabs in a window are open (bug 108973). We shouldn't present more than one dialog on Quit.

We should also think about what happens when the OS instructs Mozilla to shut down (e.g. when you are shutting down Windows). Normally, applications just close without a dialog, unless they have unsaved data. (I think IE just closes, always.) I suppose we should follow suit - don't present a dialog unless there are unsent forms, unsent messages, unsaved pages in composer or downloads in progress. Actually, when download resuming is properly implemented, not even the downloads should present a dialog in this (OS-initiated) case.

As for the problem of public terminals that Charles mentioned, I think that's a separate (but valid) issue. There should be a possibility to lock down any profile changes with a password. I don't know if there is a bug filed for that. Probably yes - there seems to be a bug for everything. ;-)

The ideal course of action now, in my opinion, is:

1) give review to the patch changing Ctrl+Q into Ctrl+Shift+Q and get it into the tree; this could be done fast, if there was will, so a big part of the dataloss issue would instantly go away and users would cheer; 2) reopen bug 39057, make an explaining comment, and add something like "start reading from comment 158" in the Status Whiteboard; :-) 3) drive the discussion in bugs 39057 and 65121 to determine whether both of them need to be fixed, or which one of them and how; 4) make it clearer what Quit actually does, either in the dialog, or in the statusbar (bug 47434).

Or, instead of that, just decide the issue on your "friday keyboard meetings" and enforce the decision. I'd love to show up in person (citing http://www.mozilla.org/projects/ui/accessibility/mozkeyintro.html#meetings :-) ), but it's a bit off the beaten track for me unless it's somewhere in central Europe. ;-)))

Sorry for being so lengthy again. I think there's a name for this disorder... ;-)

Thanks,

Vaclav Dvorak  <vdvo vdvo net>

--- End Message ---