David Poehlman wrote:
expectation is easily changed, bad design is not.
Hmm. What makes you think that?
Designs, good and bad, have shaped current user expectations, along with
publisher habits and user expectations inherited from previous media
and computer information systems.
To go back and change a design used throughout the web would be
monumentally difficult and expensive. But it is a linear, technical job.
Changing a design for one website is roughly the same as changing a
design for the next.
By contrast, changing user expectations across the web is a non-linear
task. It gets a lot easier the further along the transition you go. But
at the start, it's very hard indeed. One website changing its design
does not change expectations for users across the web. Dedicated users
changing their expectations does not change the expectations of floating
visitors. Changes in human expectations would take a long time to filter
through to search engines, which are designed around some of the same
expectations and so might send less traffic to sites that had made the
switch. In cruder terms, making a change that goes against user
expectations has the potential to cost a lot of business.
That certainly doesn't mean changing designs is never worth doing, but
don't underestimate the difficulty of changing people used to, and other
systems based around, current practice.
I agree about small biew ports so there is no harm in putting a link at the
highest point of regard that points to an index. I like the index removed
from pages altogether.
We're veering off the initial topic of where best to put the navigation,
but I think that's an interesting preference. I think keeping
navigational clutter to a minimum is good. However, I also think that
it's non-viable preposition both commercially and in terms of general
usability to reduce navigation to a single link to an index. Here's just
three reasons:
1) Iif you've got a small amount of navigation (let's say Blog, Photos,
About Me), then all you do by removing that to another page is replace
three links of navigation with one link, and force users to click twice
to get to the same content. The more minimal your navigation, the more
the returns diminish of moving it to another page.
2) If you have some navigation in the page, you can give users a hint of
what other exciting stuff is on your site. If you don't give them such a
hint, a lot of them wouldn't bother to look in the index.
3) On complex sites, you can use well-designed navigation to tell users
where they are in relation to the rest of the site:
http://www.useit.com/alertbox/breadcrumbs.html
Assuming you don't drown content with navigation, it imposes no
cognitive load on sighted users because they've learnt to simply ignore
it until they need it:
http://www.useit.com/alertbox/20000109.html
I'm a bit sceptical that removing navigation would improve people's web
experience. But if you have knowledge of CSS selectors and the display
and speak properties, you can set up user styles for some of your
favorite sites and try it out for yourself. All browsers have some sort
of facility for doing this, although the user interface is non-obvious
and will usually require you to learn some new skills. In other words,
it's guaranteed to make the user think. ;) Safari's and Internet
Explorer's implementations makes this a little difficult, often forcing
you to rely on so-called "CSS signatures" that not all sites provide,
but OmniWeb allows site-by-site user stylesheets and the Stylish add-on
for Firefox allows customization by URL. Firefox's GreaseMonkey add-on
can even alter sites on the fly, which can be more effective than
styling tweaks in some cases.
Here's some relevant resources to get you started, if you're interested:
Basic CSS tutorial:
http://www.w3.org/Style/Examples/011/firstcss
CSS selectors:
http://css.maxdesign.com.au/selectutorial/
Display property:
http://www.w3.org/TR/CSS21/visuren.html#propdef-display
Speak property:
http://www.w3.org/TR/CSS21/aural.html#propdef-speak
CSS signatures:
http://archivist.incutio.com/viewlist/css-discuss/13291
Loads of examples of user styles:
http://www.squarefree.com/userstyles/
If the next versions of HTML and XHTML includes elements for indicating
navigation areas, it would make it much easier to simply remove or hide
them. But I suspect that would be a lot less useful than being able to:
* Jump to the next navigation area with a key-combination.
* Read all navigation areas in a page with key-combination.
* Jump to the next content area with a key-combination.
* Read all content areas in a page with a key-combination.
Like I said before, an approximation of this jumping is already possible
with JAWS and Window-Eyes.
Don't make me think is a cop out.
From what, exactly?
It is also circular. It's hard to not make you think and the harder
I try, the more you think.
Any human effort may have the opposite of the intended effect. That
doesn't mean that all human aims are circular.
If it's possible to make users think harder (and you say it is), then
it's necessarily also possible to allow them to think less than that. If
it's possible to allow them to think less, then the goal of not making
them think has been achieved. So your criticism of the idea doesn't make
sense to me.
Perhaps a good example of how (I think) you're underestimating both the
power of users' habit and their intolerance of being forced to think is
the widespread failure of the warning dialogues about incorrect
certificates that web browsers pop up to protect users from phishing sites.
The Windows environment regularly pops up confirmation dialogs to ask
users if they really want to do whatever they just told the system to do
("Do you really want to delete myphoto.jpg?" and so on). This has
trained a lot of users to just click OK. So when they see a warning
dialog about incorrect certificates, they just click OK without even
reading the message.
Of course, reading the message would require the user to think in order
to extract the essential idea from a load of technobabble about
certificates.
I don't think the unwillingness to learn this stuff makes users stupid:
I think it reflects the fact that they are mostly time-poor,
non-technical folk trying to get on with a task, and their stupid
computer is getting in the way. Making security systems successful will
involve making them intelligent enough to take more of the cognitive
load off overworked, stressed-out human users.
Regards
--
Benjamin Hawkes-Lewis
