Re: GUI questions

[Ben Pfaff]

John Darrington <address@hidden> writes:

> On Tue, Oct 02, 2007 at 07:21:16AM -0700, Ben Pfaff wrote:
>
>      >      examples, procedures can't run without any data, so should we
>      >      gray out the Analyze menu until there's data?  
>      >
>      > I think not.  Procedures can run without data, although in most cases
>      > they will give trivial (and perhaps meaningless) answers.
>      
>      How about if there are no variables?
>
> I'm not sure if there are any procedures which can run with no
> variables (perhaps EXECUTE can?).  In most cases it won't make any
> sense to run a procedure without variables.   In these cases if, as
> discussed previously, OK and Paste are insensitive until a variable is
> selected, then it'll not be possible to run such a procedure using the
> GUI.

Clearly that works.  The bit about it that worries me is whether
it will be obvious to a user that OK and Paste are insensitive
*because* no variables have yet been selected.  In the case where
no variables at all exist, it seems like it's an even larger leap
of logic: from "Hmm, there are two empty boxes with an arrow
between them" to "I must need to define some variables so that I
can some variables appear in the leftmost box for me to move to
the rightmost box".

>From this point of view it might actually be friendlier to pop up
a message box "You need to define add some numeric variables to
your data set before you can compute descriptive statistics." in
this case, since it tells the user what's wrong and how to fix
it, instead of just giving them a couple of empty boxes and
possible mystification.

Is there a solution that has the advantages of each approach?
-- 
Ben Pfaff 
http://benpfaff.org