Universal GUI (was Re: HIG...)

Richard Gaskin ambassador at fourthworld.com
Wed Jul 23 18:34:01 EDT 2003


Ray G. Miller wrote:

>> However, about Apple--I can no longer consider them to be the
>> ultimate example of good interface, as they were before.
> 
> Agreed, Curry, but what other consistent Guidelines are there?
> Micro$lot? ;-) 

The current mess confronting multi-platform developers is costly to us all,
with incalculable time wasted by the majority using Revolution, Java,
RealBASIC, Director, Flash, or other cross-platform development system.  I
wouldn't be surprised if the worldwide aggregate of productivity lost to
such things came to several tens or possibly hundreds of millions of dollars
annually.

Some time ago I was considering an article with petition on a "Universal
GUI", with the theme of "Put up or shut up."  There's probably a more polite
way to phrase that, but I believe the central idea is important:

As more and more development becomes cross- or multi-platform, although most
modern GUIs share a majority of UI elements (common window trimmings, menus,
button types, etc.), each OS has enough distinctions to drive everyone crazy
trying to be "HIG-compliant" across conflicting HIGs.

So maybe we developers could turn the tables: rather than enslaving
ourselves to sometimes arbitrary specifications, we take it upon ourselves
to make one recommendation for a Universal GUI.

Any OS vendor could make suggestions for deviations from common behaviors,
but they would only be incorprated into the Universal GUI spec if it's
supported by research, with methodologies and results available for review
so we can distinguish the truly research-supported recommendations.

We should never have to write two layout routines for dialog controls, in
which Win and Mac reverse the order of default buttons, for example.  We
could pick one layout, based on research or, if such supporting
documentation is unavailable, prevalence among current GUIs, making the odd
one out either substantiate their difference with research or be ignored.

One could rightly argue that a Universal GUI could lead to a "lowest common
denominator" GUI.  We'd have to be watchful of that, but at the same time
the sum of common elements is not bad, and in its simplicity there may be
great value.  We could still keep unique appearances, since each OS provides
a heathly set of hooks for rendering controls.  But there's no reason
layouts, behavior, and nomenclature couldn't be made consistent, with the
option supported by the strongest research setting the standard others would
be asked to comply with.

I realize it would be an uphill battle and likely without victory.  Worth
pursuing, or better off left as a thought experiment? ;)

--
 Richard Gaskin 
 Fourth World Media Corporation
 Software Design and Development for Mac, Windows, Linux, and the Web
 ____________________________________________________________________
 Ambassador at FourthWorld.com                http://www.FourthWorld.com
 Tel: 323-225-3717        AIM: FourthWorldInc       Fax: 323-225-0716




More information about the use-livecode mailing list