Specifying the navigation between fields

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2005-01-23 16:07 -0700, Micah Dubinko wrote:
>The following is a "think out loud" session for my HISC action item. 
>Anyone feel free to jump in with any questions, clarifications, 
>whatever--I'd find it most helpful.

Thanks for starting this thread, Micah!  Since there is no-one on the 
teleconference, I'll put my questions into this post (probably should have 
before the meeting anyway!).

>The discussion is around navigation paths through forms. For what we're 
>trying to do here, does a single navigation path (thread) woven through 
>everything make sense?

Yes, I think there should be just one path, but it might be hierarchical 
... if hierarchical means sets of fields on multiple "pages".

>Or do we need a heirarchy of navigational paths? If yes, how do we present 
>that in a concrete way?

I think by specifying the links in one direction and inferring the reverse 
direction.  BTW, having such links will help us be methodical and allow us 
to parse the links to look for holes or ambiguities.

>  Two potential use cases for heir nav. One is when there are multiple 
> navigation techniques, like control-to-control, and page-to-page.
>
>  <group id="page1" nextfocus="page2">
>    <input id="ctrl1" nextfocus="ctrl2".../>
>    <input id="ctrl2" nextfocus="ctrl3".../>
>    ...
>    <input id="ctrlN" nextfocus="ctrl1".../>
>    <!-- or maybe next is a specific control on p2 ?? -->
>  </group>
>  <group id="page2" prevfocus="page1">
>    ...
>  </group>
>
>  The two paths never intersect.

Yes, I can see that ... though two things:

(1) where does one start (do we need a first-focus= attribute on <group>)?
(2) the "last" control on a page must not have a "next" (wouldn't it 
categorically be the first?)

Is there only one way to jump forward and backward between fields, or could 
a user sitting at the last control on a page choose one "jump forward" to 
wrap to the first control on the same page and a different "jump forward" 
to jump to the first control on the next page (and same in the backward 
direction).

Probably not ... my guess is there is only one "jump forward" and one "jump 
backward" keystroke.

Perhaps, then, our specification could indicate for a <group> whether 
boundary jumps remain within the <group> or move to an adjacent <group>.

>  Another use case is when the higher-level path is a "summary" of the
>  overall navigation paths.
>
>  <group id="seller" nextfocus="buyer">
>    <input id="name_s" nextfocus="addr_s".../>
>    <input id="addr_s" nextfocvus="???".../>
>  </group>
>  <group id="buyer" nextfocus="items">
>    <input id="name_b".../>
>    <input id="addr_b".../>
>  </group>
>  <group id="items" nextfocus="endnotes">
>    <repeat nodeset="...">
>      <input id="quan".../>
>      <input id="desc".../>
>      <inupt id="price".../>
>      <input id="ext".../>
>    </repeat>
>  </group>
>  <group id="endnotes" nextfocus="buyer(?)">
>    ...
>  </group>

I'm missing something about the above ... you've abstracted the groups by 
conceptual purpose, would the field linking be inferred within a group?

>  Random thoughts:
>
>  How does navigation work in repeating sequences, like line items?

I like what you say above along the lines of <repeat> but since it appears 
to be a property of the group, it might also be an attribute of <group> ... 
unless in one group you have both repeating sets and non-repeating sets.

>  How do you have navigating off the end of one repeat row go to the start 
> of the next?

That would be implied by the ancestral <repeat> ... but does "repeat" mean 
one action at the end of a field creates a new repeating set, while a 
different action navigates to the next group?

I gather user interfaces are *not* standardized in technologies like 
X-Forms ... but are there standardized concepts of the user interface 
actions that are available from the operator of the software?  Will that 
guide us in our available options?

>Thanks for listening. :)

Thanks for speaking!  I hope other members chime in with their thoughts ... 
anyone done user interfaces for forms-like widgets?  Can anyone cite 
guidelines on available widget operator interactions?

.......................... Ken

--
World-wide on-site corporate, govt. & user group XML/XSL training.
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/o/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Breast Cancer Awareness  http://www.CraneSoftwrights.com/o/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal