Interesting problem

21 Jun 2005 - 4:46pm
8 years ago
11 replies
387 reads
russwilson
2005

I'm working on an interesting problem right now
and would welcome some suggestions from others.
(Is there an IX pattern associated with this??)

Let's say I have the following:
(component names are made up)

Component Number of Components
----------------------------------
Box 1..N
Bubble 1..N
Bone 1..N

There are 1..N Bubbles and 1..N Bones associated
with any Box (and vice versa).

The user can choose any component first, and then
any component second, and then the final component.

So, the user can choose "Bone A", after which the
list of Boxes and Bubbles should be limited to only
those associated with Bone A. If the user then chooses
"Bubble B", then the list of Boxes should be further
limited to only those Boxes assoicated with Bone A AND
Bubble B.

The order of these selections is completely left to the
user.

I'm trying to devise an elegant way to present these choices
to the user. My current mockup involves 3 pulldown lists
with each subsequent list changing based on the previous
lists choices.

Does anyone have any ideas/thoughts?

Comments

21 Jun 2005 - 4:53pm
Juan Lanus
2005

Russell:
Could you state some numbers, i,e. how many of each (boxes, bones,
bubbles) exist?
--
Juan Lanus
TECNOSOL
Argentina

On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> [Please voluntarily trim replies to include only relevant quoted material.]
>
> I'm working on an interesting problem right now
> and would welcome some suggestions from others.
> (Is there an IX pattern associated with this??)
>
> Let's say I have the following:
> (component names are made up)
>
> Component Number of Components
> ----------------------------------
> Box 1..N
> Bubble 1..N
> Bone 1..N
>
> There are 1..N Bubbles and 1..N Bones associated
> with any Box (and vice versa).
>
> The user can choose any component first, and then
> any component second, and then the final component.
>
> So, the user can choose "Bone A", after which the
> list of Boxes and Bubbles should be limited to only
> those associated with Bone A. If the user then chooses
> "Bubble B", then the list of Boxes should be further
> limited to only those Boxes assoicated with Bone A AND
> Bubble B.
>
> The order of these selections is completely left to the
> user.
>
> I'm trying to devise an elegant way to present these choices
> to the user. My current mockup involves 3 pulldown lists
> with each subsequent list changing based on the previous
> lists choices.
>
> Does anyone have any ideas/thoughts?
>
>
> _______________________________________________
> Welcome to the Interaction Design Group!
> To post to this list ....... discuss at ixdg.org
> (Un)Subscription Options ... http://discuss.ixdg.org/
> Announcements List ......... http://subscribe-announce.ixdg.org/
> Questions .................. lists at ixdg.org
> Home ....................... http://ixdg.org/
>

21 Jun 2005 - 4:58pm
russwilson
2005

Sorry --

Possible dimensions are:
Box --> in the 1000s
Bubble --> in the 100s
Bone --> in the 10s

-----Original Message-----
From: Juan Lanus [mailto:juan.lanus at gmail.com]
Sent: Tuesday, June 21, 2005 4:54 PM
To: Wilson, Russell
Cc: discuss at ixdg.org
Subject: Re: [ID Discuss] Interesting problem

Russell:
Could you state some numbers, i,e. how many of each (boxes, bones,
bubbles) exist?
--
Juan Lanus
TECNOSOL
Argentina

On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> [Please voluntarily trim replies to include only relevant quoted
> material.]
>
> I'm working on an interesting problem right now and would welcome some

> suggestions from others.
> (Is there an IX pattern associated with this??)
>
> Let's say I have the following:
> (component names are made up)
>
> Component Number of Components
> ----------------------------------
> Box 1..N
> Bubble 1..N
> Bone 1..N
>
> There are 1..N Bubbles and 1..N Bones associated with any Box (and
> vice versa).
>
> The user can choose any component first, and then any component
> second, and then the final component.
>
> So, the user can choose "Bone A", after which the list of Boxes and
> Bubbles should be limited to only those associated with Bone A. If
> the user then chooses "Bubble B", then the list of Boxes should be
> further limited to only those Boxes assoicated with Bone A AND Bubble
> B.
>
> The order of these selections is completely left to the user.
>
> I'm trying to devise an elegant way to present these choices to the
> user. My current mockup involves 3 pulldown lists with each
> subsequent list changing based on the previous lists choices.
>
> Does anyone have any ideas/thoughts?
>
>
> _______________________________________________
> Welcome to the Interaction Design Group!
> To post to this list ....... discuss at ixdg.org (Un)Subscription Options

> ... http://discuss.ixdg.org/ Announcements List .........
> http://subscribe-announce.ixdg.org/
> Questions .................. lists at ixdg.org Home
> ....................... http://ixdg.org/
>

21 Jun 2005 - 5:12pm
Juan Lanus
2005

Yup! A thousand is too much for a pulldown.
More questions: is this application to be run into a browser or in a
desktop with that fantastic communication speed between the data
sources and the UI?
--
Juan

On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> Sorry --
>
> Possible dimensions are:
> Box --> in the 1000s
> Bubble --> in the 100s
> Bone --> in the 10s
>
>
> -----Original Message-----
> From: Juan Lanus [mailto:juan.lanus at gmail.com]
> Sent: Tuesday, June 21, 2005 4:54 PM
> To: Wilson, Russell
> Cc: discuss at ixdg.org
> Subject: Re: [ID Discuss] Interesting problem
>
> Russell:
> Could you state some numbers, i,e. how many of each (boxes, bones,
> bubbles) exist?
> --
> Juan Lanus
> TECNOSOL
> Argentina
>
> On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> > [Please voluntarily trim replies to include only relevant quoted
> > material.]
> >
> > I'm working on an interesting problem right now and would welcome some
>
> > suggestions from others.
> > (Is there an IX pattern associated with this??)
> >
> > Let's say I have the following:
> > (component names are made up)
> >
> > Component Number of Components
> > ----------------------------------
> > Box 1..N
> > Bubble 1..N
> > Bone 1..N
> >
> > There are 1..N Bubbles and 1..N Bones associated with any Box (and
> > vice versa).
> >
> > The user can choose any component first, and then any component
> > second, and then the final component.
> >
> > So, the user can choose "Bone A", after which the list of Boxes and
> > Bubbles should be limited to only those associated with Bone A. If
> > the user then chooses "Bubble B", then the list of Boxes should be
> > further limited to only those Boxes assoicated with Bone A AND Bubble
> > B.
> >
> > The order of these selections is completely left to the user.
> >
> > I'm trying to devise an elegant way to present these choices to the
> > user. My current mockup involves 3 pulldown lists with each
> > subsequent list changing based on the previous lists choices.
> >
> > Does anyone have any ideas/thoughts?
> >
> >
> > _______________________________________________
> > Welcome to the Interaction Design Group!
> > To post to this list ....... discuss at ixdg.org (Un)Subscription Options
>
> > ... http://discuss.ixdg.org/ Announcements List .........
> > http://subscribe-announce.ixdg.org/
> > Questions .................. lists at ixdg.org Home
> > ....................... http://ixdg.org/
> >
>

21 Jun 2005 - 5:15pm
russwilson
2005

It will be run in a browser.

-----Original Message-----
From: Juan Lanus [mailto:juan.lanus at gmail.com]
Sent: Tuesday, June 21, 2005 5:12 PM
To: Wilson, Russell
Cc: discuss at ixdg.org
Subject: Re: [ID Discuss] Interesting problem

Yup! A thousand is too much for a pulldown.
More questions: is this application to be run into a browser or in a
desktop with that fantastic communication speed between the data sources
and the UI?
--
Juan

On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> Sorry --
>
> Possible dimensions are:
> Box --> in the 1000s
> Bubble --> in the 100s
> Bone --> in the 10s
>
>
> -----Original Message-----
> From: Juan Lanus [mailto:juan.lanus at gmail.com]
> Sent: Tuesday, June 21, 2005 4:54 PM
> To: Wilson, Russell
> Cc: discuss at ixdg.org
> Subject: Re: [ID Discuss] Interesting problem
>
> Russell:
> Could you state some numbers, i,e. how many of each (boxes, bones,
> bubbles) exist?
> --
> Juan Lanus
> TECNOSOL
> Argentina
>
> On 6/21/05, Wilson, Russell <Russell.Wilson at netqos.com> wrote:
> > [Please voluntarily trim replies to include only relevant quoted
> > material.]
> >
> > I'm working on an interesting problem right now and would welcome
> > some
>
> > suggestions from others.
> > (Is there an IX pattern associated with this??)
> >
> > Let's say I have the following:
> > (component names are made up)
> >
> > Component Number of Components
> > ----------------------------------
> > Box 1..N
> > Bubble 1..N
> > Bone 1..N
> >
> > There are 1..N Bubbles and 1..N Bones associated with any Box (and
> > vice versa).
> >
> > The user can choose any component first, and then any component
> > second, and then the final component.
> >
> > So, the user can choose "Bone A", after which the list of Boxes and
> > Bubbles should be limited to only those associated with Bone A. If
> > the user then chooses "Bubble B", then the list of Boxes should be
> > further limited to only those Boxes assoicated with Bone A AND
> > Bubble B.
> >
> > The order of these selections is completely left to the user.
> >
> > I'm trying to devise an elegant way to present these choices to the
> > user. My current mockup involves 3 pulldown lists with each
> > subsequent list changing based on the previous lists choices.
> >
> > Does anyone have any ideas/thoughts?
> >
> >
> > _______________________________________________
> > Welcome to the Interaction Design Group!
> > To post to this list ....... discuss at ixdg.org (Un)Subscription
> > Options
>
> > ... http://discuss.ixdg.org/ Announcements List .........
> > http://subscribe-announce.ixdg.org/
> > Questions .................. lists at ixdg.org Home
> > ....................... http://ixdg.org/
> >
>

21 Jun 2005 - 5:40pm
Robert Cornejo
2004

Agree with Juan that form elements like drop-downs and
multi-row lists won't work for 1000s of items. There
are issues of display, choice-in-context and memory
limitations.

One approach might be to use a simple search with
grouped results (a la Amazon) for Box, Bubble and
Bone.

Box:
result A
result B
...first 10/20...
{see all results - link to listing}

Bubble:
result 1
result 2
...first 10/20...
{see all results - link to listing}

Bone:
result X
result Y
...first 10/20...
{see all results - link to listing}

Could also use a secondary element on the search for
limiting scope, like both Ebay and Amazon, with the
default for "all categories".

If you have usage stats there's also the "most
popular" listing and "most recent" linked list.

Another approach might be paginated listings with
column sorting (if applicable to the data) and useful
sorting defaults.

= R

--- "Wilson, Russell" <Russell.Wilson at netqos.com>
wrote:

> > Possible dimensions are:
> > Box --> in the 1000s
> > Bubble --> in the 100s
> > Bone --> in the 10s

____________________________________________________
Yahoo! Sports
Rekindle the Rivalries. Sign up for Fantasy Football
http://football.fantasysports.yahoo.com

21 Jun 2005 - 7:06pm
Susan Farrell
2004

I've seen configurators tackle large numbers (tens of thousands) of
dependent choices pretty well. You choose one parameter and then the
narrowed list of further parameters appears. Sometimes you can make
the first choice based on the gating factor that matters most to you,
for example size or price.

Here's one for laptops:

http://www.ibm.com/products/finder/us/finders?pg=nbfinder&N=200004

Susan

21 Jun 2005 - 7:23pm
Donna Maurer
2003

This is generally known as faceted browsing. there are a bunch of
examples linked here:

http://iawiki.net/FacetedBrowsing

Donna

On 21 Jun 2005 at 17:06, Susan Farrell wrote:

> [Please voluntarily trim replies to include only relevant quoted
> material.]
>
> I've seen configurators tackle large numbers (tens of thousands) of
> dependent choices pretty well. You choose one parameter and then the
> narrowed list of further parameters appears. Sometimes you can make
> the first choice based on the gating factor that matters most to you,
> for example size or price.
>
> Here's one for laptops:
>
> http://www.ibm.com/products/finder/us/finders?pg=nbfinder&N=200004
>
> Susan
>
--
Donna Maurer
blog: http://maadmob.net/donna/blog/
work: http://steptwo.com.au/
AOL IM: maadmob

21 Jun 2005 - 7:52pm
Josh Seiden
2003

One of the things to keep in mind in situations like this is that
there are many many possible combinations of widgets that can be
strung together to "solve" this problem. Often, the widgets
themselves hold no cue as to which assemblage of widgets is
correct, because the widgets provide no evaluation context.

One way to approach this problem is not to get generic (bone,
bubble, box) but instead get very very specific.

For instance: What is the user trying to do? While it may be
technically true that any attibute could be selected first, it is
often the case that your user population will elevate one
attribute above the others. Try defining one attribute as the
driver, and see what happens.

Or perhaps there is a certain type of result or result set the
user is looking for. Can you solve the problem by giving
result-set finding tools, rather than attribute specification
tools?

I understand that you may have made the problem generic in order
to discuss it publicly. But it is usually the case that generic
problems are much harder to solve than specific instances of
them--especially in software, where context is everything.

JS

> -----Original Message-----

> I'm working on an interesting problem right now
> and would welcome some suggestions from others.
> (Is there an IX pattern associated with this??)
>
> Let's say I have the following:
> (component names are made up)
>
> Component Number of Components
> ----------------------------------
> Box 1..N
> Bubble 1..N
> Bone 1..N
>

22 Jun 2005 - 12:27pm
Anirudha Joshi
2003

Joshua Seiden <One way to approach this problem is not to get generic
(bone, bubble, box) but instead get very very specific.>

I double Joshua here, but with a variation.

Another way of looking at this problem is that if you want users to
select bone, bubble, box in any sequence, then the visual narrowing of
the list may not help.

For example, the oft-quoted iHotelier flash interface narrows the room
options available if I select a date, but keeps them visible, none the
less. This keeps the user informed that he can choose another set of the
dates and will get his preferred room.

To put it in context, a business user may have a fixed schedule, but may
be a bit flexible about the room type. A budget-leisure traveller may be
a bit more flexible on his schedule, if he can save a bit. A honeymoon
couple has fixed dates and high experience demands - they may just opt
for another hotel, or even another destination. Different users have
different perspectives, and the sequence may change accordingly.

So where do I double Joshua? The problem needs to be solved 'in
context'. All the widgets that work for iHotelier (a 2 month calendar, a
room list with a view, options for number of extra beds etc.) will not
work for buying laptops or even train tickets.

That is why designers need to be paid, and creative ones need to be paid
more.

Anirudha

22 Jun 2005 - 9:10am
russwilson
2005

Thanks to everyone for their feedback.

Joshua and Anirudha - great points... I was arriving
at the same conclusion myself. (solve the problem
in context, not generically... and consider selecting
a "focus")

And the pattern I was trying to remember was Faceted Browsing -
thanks to Donna!

- Russ

21 Apr 2006 - 12:43pm
erez
2005

One way I used was initially to present only the first pulldown lists.
After the user made his first choice I made the second pulldown lists
visible (using DHTML). The same applies to the last box. This allow users to
act within a logical action flow.

Erez Kikin-Gil | Experience design

+44 (0) 791.402.1270

MSN : Erez at tiltool.com

IM/AOL: erezkikingil

Http://www.tiltool.com

-----Original Message-----
From:
discuss-interactiondesigners.com-bounces at lists.interactiondesigners.com
[mailto:discuss-interactiondesigners.com-bounces at lists.interactiondesigners.
com] On Behalf Of Wilson, Russell
Sent: Tuesday, June 21, 2005 10:47 PM
To: discuss at ixdg.org
Subject: [ID Discuss] Interesting problem

[Please voluntarily trim replies to include only relevant quoted material.]

I'm working on an interesting problem right now
and would welcome some suggestions from others.
(Is there an IX pattern associated with this??)

Let's say I have the following:
(component names are made up)

Component Number of Components
----------------------------------
Box 1..N
Bubble 1..N
Bone 1..N

There are 1..N Bubbles and 1..N Bones associated
with any Box (and vice versa).

The user can choose any component first, and then
any component second, and then the final component.

So, the user can choose "Bone A", after which the
list of Boxes and Bubbles should be limited to only
those associated with Bone A. If the user then chooses
"Bubble B", then the list of Boxes should be further
limited to only those Boxes assoicated with Bone A AND
Bubble B.

The order of these selections is completely left to the
user.

I'm trying to devise an elegant way to present these choices
to the user. My current mockup involves 3 pulldown lists
with each subsequent list changing based on the previous
lists choices.

Does anyone have any ideas/thoughts?

_______________________________________________
Welcome to the Interaction Design Group!
To post to this list ....... discuss at ixdg.org
(Un)Subscription Options ... http://discuss.ixdg.org/
Announcements List ......... http://subscribe-announce.ixdg.org/
Questions .................. lists at ixdg.org
Home ....................... http://ixdg.org/

--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.323 / Virus Database: 267.7.9/23 - Release Date: 6/20/2005

Syndicate content Get the feed