Re: New browser controls and an API for using them

Since I started this thread I've done some more thinking about it and
I agree that controls like cut, copy and paste would not work well as
they really only apply to editing situations. But I am still convinced
of a few points.

1)  Yes, users are used to navigating the history stack with 'back'
and 'forward' buttons and it would take some re-learning to get used
to knew buttons, but users can't learn to use controls that don't
exists and that all most browsers have ever offered.

2)  Just because 'back' and 'forward' navigation is the way it's
always been done doesn't mean it's the best way (I will also admit
that it might be the best way, but I think it's worth serious
discussion).

3)  The controls that the users are going to use most often are the
ones that are consistent and that they know where they are. This is
why the back button gets so much use, is because it's always there in
the same place and always dose what the user expects. Everything on
the web page itself is subject to change and uncertainty.

Because of this I think it's worth discussing the possibility of new
controls and how they would work, to allow users to better interact
with web page/application in the future. I think this is a topic for
specifications and not implementations (user agents) because different
implementations may access these types of controls differently, but
each user will get used to what his or her implementation dose.

I've been thinking a log about the rel and rev attribute on the <a>
and <link> tags in this way and I think the specification already
offers a start at such controls. Opera has a Navigation bar that gives
the user access to W3C defined link types. I've working on a similar
plug in for Firefox.

The shortcomings of the the defined link types is they mostly to
scientific documents, with glossary, index and first/last, next prev
controls. Web site are not usually laid out this way, and web
applications still trying to discover what they will look like but it
will be entirely new.

1)  I am thinking we need to have new link types defined that meet the
needs for the web how it is today and how it will look tomorrow.
Microformats http://microformats.org/ starting to be defined to add
these link types.

2)  Browsers need to implement controls for the user to access these
standardised link types, I think this should be part of the official
recommendation.

3)And scripting events need to be defined so that developers can have
scripts interact with users activating these new controls in the same
way a developer can capture an onclick event from the user click an
link on the page. Possibly having all links (<a> and <link>) fire an
onactive event that happens when the link goes into it's active state,
just before the new http request is made.

4)  Web developers need to be educated and encouraged to use 'rel' and 'rev'

With a system like this in place, it the user wants to go to the 'home
page', see the 'site map', view his 'shopping cart' etc... he or she
will not have to look around for the link to take them there. It will
be right in the browser controls, right where it's always been.

So far I've mainly been talking about navigation, so how dose this
relate to web apps. Using defined controls can catching the events
(like the onactive event) developers can add functionality to there
web application that the user will easily understand and that will be
consistent from one web application to the next.

Maybe it's not cut, copy, paste and undo. But I think there could be a
need for a soft-refresh control in AJAX application, so the user isn't
tempted to click the page refresh button when they want to force an
update page content. Or a setting/preferences control that gives the
user access to user defined settings. Brainstorming can come up with a
lot more ideas of controls that would be consistent across most all
applications.

Sorry for being so long winded, and again I'm willing to admit this
isn't the best road to follow, but it sound good to me. I just
strongly feel as I said above that doing thinks they way they've
always been done, just because that's what everyone is used to, is a
bad idea and that considering better ways to interact with web content
(especially as web applications using the AJAX model are starting to
become more common) is a good idea.

Kenny

Received on Tuesday, 13 December 2005 16:07:30 UTC