Re: Opinions on Embedded Help? + an HTML Help rant

Subject: Re: Opinions on Embedded Help? + an HTML Help rant
From: Stephen REYNOLDS <sreynolds -at- INFORMATIVEMEDIA -dot- COM>
Date: Wed, 10 Mar 1999 14:12:46 +0000

David Castro wrote:

I'm including embedded help in the next release of the software I'm
writing, and wanted to see what others on the list thought of this "new"

idea. Etc.
My only real experience of embedded help comes from looking at the
HelpExtender examples on a WexTech demo CD. On one hand I consider it an
exciting opportunity?a relatively easy way towards the Electronic
Performance Support (that was the terminology back in 1991; now I think
we?re supposed to say user-centred design) principles that have always
sounded so sensible. On the other hand, I am worried that I am seeing
another case of the tail wagging the dog. HTML-based help is opening up
new opportunities, but at the same time it is opening up a can of worms.
And it opens up the question of who will provide us with usability
guidance now that HTML, CSS, JavaScript, Shockwave, streaming video, and
Java give us new ways to present information, provide navigational
devices, integrate interaction, and provide feedback.

Wextech?s embedded help examples demonstrate dynamic procedures. For
example, you open a dialog box that has the usual fields, check boxes,
and radio buttons. Alongside is a help pane that lists perhaps six
steps. If checkbox X and radio button Z are already selected when you
open the dialog box, the procedure steps for selecting X and Z are
unavailable (greyed out). So can ignore them and just do the remaining
four active steps. Great, you say. So user friendly.

However, I think this is a naïve response. When playing with the Wextech
demos, I became aware that I really was not reading the inactive steps.
If a step was in black text, I did something about it. The moment a step
turned from black to grey (I had clicked or typed somewhere), I ignored
it. I didn?t look at it to see why it was no longer black. Surely, this
has a negative side.

Most of us?if not all of us?have happily bought into the idea of task
orientation. We accept that users should not have to learn the inner
functioning of a piece of software before using it before accomplishing
tasks. However, whether we like it or not, users do ?learn? the
software. Through experience, they construct mental models that they use
in planning how to accomplish a task, actually doing the work, or in
troubleshooting any problems. I am worried that the type of dynamic
procedures I have seen will hinder the construction of mental models by
the hiding information that helps build it. From my personal experience,
I feel that the use of greyed text encourages you to ignore potentially
useful information; you are discouraged from deciding whether the
information is important for you. Using dynamic procedures might provide
some instant improvement in a novice user?s performance on one simple
task. But in the longer term and for complex work demanding a higher
level of knowledge ??

Am I making a fuss about nothing? I wonder how much usability testing
Wextech has done on the types of embedded help they are promoting
through their examples. Perhaps Wextech would argue that they are only
demonstrating technical possibilities. But I would argue that, to
decision-makers with no knowledge of usability, these examples could be
seen to validate the approach taken (Wextech are HAT manufacturers so
this dynamic stuff must be problem-free).

With good old WinHelp, our design decisions were relatively limited, and
many people have taken the pragmatic way to develop easy-to-use help is
to produce stuff that looks much like help produced by everyone else. In
practice , this has often meant that we have copied the ideas we have
seen in Microsoft?s own help files. I have periodically seen mention of
Microsoft?s mythical usability lab (I?ve never seen or read of any
reference to any results published by this lab?strange!). So naturally,
it is safe to assume that any approach adopted by MS has been tested on
a significant number of real people. (As opposed to more academic
usability research, which often seems to content to test only unoccupied
students.) And although Microsoft are not exactly a model of
consistency, their WinHelp modules have provided a certain level of
guidance to those of us in no position to do usability testing of
alternative approaches to structuring and presenting information. By and
large, the HAT manufacturers have done little that challenged
Microsoft?s model of adequately usable help systems. For years, they
seemed more interested in giving us little graphics utilities than
usability features such as natural language search facilities. (Note:
this is not actually a dig at Wextech; they are more innovative than
most of their competitors.)

Anyway with WinHelp, most of us have happily lived with (to take an easy
example) green-underlined hypertext. We had other options, but, in most
cases, we came to the decision that we had an adequate standard. If
users were unhappy about the colour, I never heard anything about it. I
heard no uproars because we couldn?t change the colour of used links or
links being clicked. The lack or ?rollovers? did not see an issue.

Along comes HTML Help and everything changes. Where do I go if I want to
take advantage of HTMLs new possibilities while still being? (This is a
rhetorical question, not really a cry for help). Well one thing for
sure, I don?t see much usability guidance when I look at some of the CHM
files at my disposal. Thinking of hypertext links again, the help for
Windows 98 has links of different colours. Some links change colour when
they are used; others stay the same colour. And what about the help for
v1.2 of the HTML Help Workshop!! Other list members have already voiced
their disapproval at the use of hypertext links that get longer when a
mouse pointer moves over them. This causes a disturbing ?repagination?
of the help text, especially in narrow windows. Rollover colours might
be useful, but I don?t see them as necessary. Most of us coped with the
Web before Microsoft decided to give us its proprietary :hover tag. I
wonder how many people will bother to coordinate their rollover text
effects with their rollover graphics effects to get some kind of
consistent message across? I notice that many most web sites make no
attempt to provide an ?already visited? version of hotspot graphics.

The help files with the HTML Help Workshop also introduce us to some new
types of jump:

- The ?GO? words

- Text links preceded by little triangles. These expand or close
optional sections of text. Maybe Microsoft brought this in to compensate
for problems with 1.2?s secondary windows. Of course, this method of
laying text raises issues concerning indexing and full-text search.

I can?t help feeling these new types of jump are more demonstrations of
DHTML than a considered and tested addition to our help interface

Most of the people who contribute to this list or, like me, hang around
silently consider ourselves help ?professionals?. Many of us have been
in this game some time and --rightly or wrongly think we are
sufficiently experienced and user-centred to make intelligent design
decisions. On the other hand, a significant of help development seems to
be done by people with little experience of or interest in the field. I
am concerned about how all of us make the move from the fairly
constrained world of WinHelp to the freedom of HTML-based help (with the
bad influence of its unruly cousin, the Web). For some of us the
potential confusion will provide good money-earning opportunities (even
more help conferences!!!1). But what sorts of help systems will finding
their way on to the desktop? (On the current evidence, with no secondary
windows but lots of flashing bits.) And this against a background of
more and more companies supplying online documentation only.

Considering the messages I?ve seen on this list and WINHLP-L over the
last year or so, many of us are still at the stage of trying to make
HTML Help or Web Help work at all. I don?t even want to think about
JavaHelp. Some usability issues have been raised (potential confusion
with a synchronised ToC, for example), but most questions have concerned
the nuts and bolts. I wonder when topic design will take over as our
main preoccupation?

Please excuse my rant. I do not currently have colleagues interested in
discussing these types of issues. I need a way to get some ideas off my
chest. I hope that some of this stuff is of interest to someone.

Regards, Steve

Stephen Reynolds
sreynolds -at- informativemedia -dot- com

From ??? -at- ??? Sun Jan 00 00:00:00 0000=

Previous by Author: CD-ROM Search Engines
Next by Author: Re: Replacing SHED graphics with new image (while preserving the hots pots)
Previous by Thread: Re: Explanations for getting fired
Next by Thread: Interview & testing

What this post helpful? Share it with friends and colleagues:

Sponsored Ads