>
> ---------- Forwarded message ----------
> From: Berthold Stoeger <[email protected]>
> To: Dirk Hohndel <[email protected]>
> Cc: [email protected], [email protected]
> Bcc:
> Date: Tue, 12 May 2020 07:46:00 +0200
> Subject: Re: RFC: Statistics in Subsurface
> On Dienstag, 12. Mai 2020 00:02:06 CEST Dirk Hohndel wrote:
>
> > Here's a crazy idea how this could look for example:
> >
> > The filters are built incrementally, with a drop down menu that allows
> you
> > to add criteria or constraints. Like date range, tags, people, etc
> >
>

I really like the idea of building the filters incrementally.  This style
of user interface is showing up in other tools I use, so I think it will
feel familiar.   I also believe most people filter on just one thing most
of the time, then occasionally go for more complex filters with many
criteria.  At least that's my pattern of use.


> > once the user picks one, that is added to the list and can be populated.
> > In the example the user first added 'date' and then 'tags'.
> > The tags one shows the idea of having additional options (all/any/none
> and
> > substr/starts/exact, just like we have today). Using font size and font
> > color to make this seem less cluttered.


Later on, Dirk is asking questions about visualization of the stats.
That's a tough one.  I"m not convinced we're a good judge of the average
user.  For example, the box & wiskers plots are something I deal with at
work, so they are very familiar and compact.  That said, some of my
colleagues look at those things blankly, especially when the distributions
are non gaussian.   I don't know how to satisfy all, but I think good
filtering and simple min/mean/max are already valuable.

...Hartley
_______________________________________________
subsurface mailing list
[email protected]
http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface

Reply via email to