Re: [Usability] Indicating required fields

[Kalle Vahlman <kalle vahlman gmail com>]

2005/9/20, Alan Horkan <horkana maths tcd ie>:
> 
> On Tue, 20 Sep 2005, Kalle Vahlman wrote:
> 
> > Date: Tue, 20 Sep 2005 08:58:05 +0300
> > From: Kalle Vahlman <kalle vahlman gmail com>
> > Reply-To: zuh iki fi
> > To: Usability gnome conference <usability gnome org>
> > Subject: Re: [Usability] Indicating required fields
> >
> > 2005/9/20, Alan Horkan <horkana maths tcd ie>:
> > >
> > > On Mon, 19 Sep 2005, David Zulaica wrote:
> > >
> > > > Date: Mon, 19 Sep 2005 11:57:24 -0700
> > > > From: David Zulaica <david pseudo-cyb org>
> > > > To: 'Matthew Thomas' <mpt myrealbox com>,
> > > >      'Usability gnome conference' <usability gnome org>
> > >
> > > > Subject: RE: [Usability] Indicating required fields
> > >
> > > > The required fields could have a different color border.
> > >
> > > Subject [             ] (required)
> > >
> > > It is ugly but it is accessible and unambiguous.
> > > Colour alone is not likely to be enough.
> >
> > Colour can be a good start at least, if it goes away when text is
> > entered. The colour should signal "hey, here's a field you haven't
> > filled in yet!" so if you forget to put a value in and the action you
> > were going to do is not availabe, you'll spot the missing one more
> > easily as it will have different colour wrt to the filled fields.
> 
> As Matthew already mentioned the thing to do would be to focus the cursor
> on the first field which has not already been filled, complete with
> preselected text.

Well, that's one way of doing it, yes. I still would get annoyed by it
as it requires me to do something the program already knows isn't
possible. This is just making the user activate the validation of
data, instead of doing it automatically.

If the button was disabled and there is a hint that "i'm not filled
in", the user doesn't have to try to activate the process in vain just
to get the program to say "hey! lookey here, you missed this", he can
easily see that somethings missing from the fact that the action is
unavailable and there's some visual hint where the problem is.

> Without a specific example in mind it is hard to tell (and this is all
> ridiculously hypothetical but fun nonetheless)

Isn't Usability just like that always?  ;)

> but if a field is not
> important why show it at all?  Unimportant fields could be left hidden
> (behind a disclosure triangle or by some other method) or not shown at
> all.  If a field is not mandatory do not waste the users time asking them,
> design for the user not just the paymaster.

It's not black and white like that. Do we make all optional values to
be hidden by default, so that the user needs to expand every one
separately? After all, one field could be unrelated to the other
optional field, so they can't be in the same group.

This obviously should get some test implementations and field testing
to determine whether it would be better or worse than "late"
validation, it won't get decided here.

But it's fun to discuss it anyway :)

-- 
Kalle Vahlman, zuh iki fi
Powered by http://movial.fi