|
TGIIF -- Not the raging snowstorm that was more or less predicted.
But I am under the weather, feeling the shift.
Jakob Nielsen¨s Nov 14th article
When Bad Design Elements
Become the
Standard is a provocative one for those who wish to work for
better web design. The point he makes is however valid to a large extent,
in that when most major sites adopt a particular design convention, pretty
much no matter how bad, it quickly becomes the expected norm for a large
portion of the user base. Although he does not explicitly mention frames,
I believe this too is a candidate for a middle, or strong convention example.
Frameset use spread very quickly in the face of good navigational reasons
not to use frames, particularly at a time when most used v3 browsers that
could not back through visited frames.
Jakob then goes on to give a number of interesting examples, some design
element uses he deplores, others which he finds potentially useful if only
sites would stop misusing these design elements in inappropriate contexts.
He admits that his xx% figures are simply guesses at this point and calls
for more research.
One convention he takes up is the vertical, colored navigation column --
the "yellow fever" design -- which on my site goes against the
established convention by being pale blue and on the right. Does this departure
from "most sites" really confuse users as much as he suggests it would? I
think not. (Of course, anyone used to X windows and Squeak could
legitimately complain that their scrollbars are on the left, so
having my naviagtion on the right is a pain for them.) Anyway, this
is a only part of a larger convention to style pages in columns similar to
the printed magazine page -- sometimes seemingly only to provide more advertising
space. While there may be good reason to at least suggest (CSS,
possibly table) a limited width to text, I find many of these text columns
as implemented rather painful to read unless I toggle off a number of browser
features to "normalize" the the text, especially in terms of font selection
and size... all that
itty-bitty-teeny-weeny-text
Another design element taken up is the "tab" metaphor, which is very much
used these days in GUI design, especially in configuration dialogs. Rarely
is this very helpful on webpages, especially since it relies so heavily on
graphical elements, and often uses imagemap to work out where you click.
Not mentioned was the "folder-tree" rendering of a site structure or category
index, usually in a frame context. Most often, I find this really slows down
navigation, and become useless without images turned on, just like most button
arrangements (a column of identical IMAGE boxes).
A link convention that I recently introduced on the Wiki
has as perhaps you've noted also ended up here. I don't mess with the color
or underlining conventions -- they default, so you get whatever your browser
is set to do -- but I have added a CSS element to make the link anchor
italic unless otherwise specified at the browser/user-CSS end. I found
that I preferred this myself, especially given the EmbeddedCapital
style of Wiki-links.
Reviewing my own webauthoring, I find that I these days
follow a relatively small set of guidelines:
-
safe colors (the 216 palette) for background
-
themed sections -- color and sometimes narrow graphics along left or top
edge
-
almost nonexistent use of FONT and FONT COLOR -- CSS styling where "appropriate"
-
a small globe graphic to indicate inlined URLs that point to external sites
(no forcing of new windows)
-
few inlined graphics, and thumbnails to larger ones
-
no framesets, banners or imagemaps
-
avoiding graphical icon or text buttons for navigation
-
largely "default" setting of link rendering
-
some form of "breadcrumb trail", that at least extends back to the
section parent page
-
heading links at the top of "longer" pages
Hmm, maybe a longer list than I thought :)
Forcing new windows for a followed link, incidentally, is
not a good idea. While I have sometimes been confused by this sort of thing,
I never thought much about it. In fact, at one point I reasoned that it might
even be a good convention for off-site links. When Isabel started doing large
scale searches for material on the Web, I found myself periodically called
to assist her when she got hopelessly lost. The problem was invariably that
some link had opened a new window, hiding the previous one, and creating
a dead-end for the browser back button. (Less often, but equally problematic,
were links that opened frame-pages in separate windows -- such pages frequently
lack any navigational links at all, and give no information about in what
context they reside.)
An interesting assertion I have seen lately from several
sources, is that more and more people are surfing the Web with Java and
Javascript support routinely turned off -- some give informal estimates as
high as 30% to 50% -- and increasingly image rendering turned off as well.
They are in a sense deliberately devolving the Web experience to Lynx level,
avoiding sites that have Java-only or graphic-only navigation unless absolutely
forced to. There is thus ample reason to now and then check how gracefully
your webpages degrade when viewed at this essentially text-only level.
I have an as yet unverified impression myself that text-only surfing is a
noticeable phenomenum. I use invisible hit counters on some pages, which
are technically in the form of inlined image references. This serves two
purposes in this context: they increment only on real refresh calls for images
on a page and so understate rather than overstate page hits, and comparisons
with the server hit logs should be able to indicate a trend towards text-only
surfing. I think they do, but I really don't have enough data for a long
enough timeline.
And this was a really good suggestion by Jerry Pournelle illustrating
the nature of this business, I thought, quoted from his site, apropos the
MS FOF issue:
At COMDEX I suggested to Steve Balmer that Microsoft go to the judge
and say "We give up. We'll give away our Monopoly. We will give Windows 98
to Netscape. They have to take it all, including legacy and tech support,
but we'll give them the source code. We keep Windows NT and Windows 2000
of course." Then when in about 180 days Windows 98 is no longer selling,
and it still has to be supported, Microsoft can say "See, we gave them our
best selling product, and they still couldn't go anywhere." Balmer laughed
so hard I thought he would split a gut.
The courts aren't going to be able to handle this, and by the time
they can, Windows 98 will be irrelevant.
Whatever else one may think about MS and what has been going on since the
days of DOS and Win 1-2-3, the above is likely the bottom line assessment.
To which I can add this paraphrase:
Never attribute to malicious intent what can be adequately explained
by imperfect market forces.
|
Hi,
welcome to this week's journal. The update-link
(above) points to where I last added some text, which should simplify your
keeping up to speed. Of course, this assumes I remember to move it, and you
may still have to scroll back a bit and see if I've updated more than once
since you last visited, but that is easily done.
.
Open yours now!
