derek bambach <[log in to unmask]> writes:
> if i'm in the middle of filling out a complex form for an on-line stock
> trade, and i need to see something in the help, i don't want to loose my
> context - i want to see the help side-by -side with the form i'm filling
> out so i can understand what i'm doing. then i just close the help window
> and move on.

So I use the middle button of my mouse and open a new window
*intentionally* and not by demand of the site's author.

> it is also very useful when your user is in a secure server space
[...]
> if you make them move to another page in the original window, they
> may not be able to back into the secure page easily or at all - they
> may have to log back in and perhaps start filling out a form all
> over again. making them do that is not very good design (IMHO).

But if you need to relogin if you have left the protected server and
come back later, the design is broken anyway. So why work around a
non-functional design instead of doing it right from the start? Most
browsers I know remeber authorization information during sessions.


> and sometimes that series or process is actually important, and
> interrupting it by straying outside of the path could have negative
> consequences for the user (like making them start something again).

So warn your visitors that they'd better open a new window if they
follow such a link. Or rethink the design. I cannot think of a single
reason to impose restrictions like "don't interrupt this suite of
pages" on a visitor. I'd certainly consider these pages as unfriendly
or even hostile. The Web does not work like a classical application
and every attempt to circumvent that is bound to fail sooner or later.

> there is no design stragegy that is universally good, and there is
> no golden rule that says the end-user is the absolute arbiter of
> good design.

All of the above said, I couldn't agree more with you on your last
sentence.

        Jo