On a VHS, you can pause a video playback and go forward or
backward frame-by-frame to inspect a scene that you might have
missed because it was played too quickly.
Similarly, when browsing the Internet, a URL may change very
quickly, but the user actually wanted to go back to inspect that
certain
URL. In frame-by-frame navigation mode, you can do that,
because sudden changes in the URL, such as by JavaScript redirect,
are intercepted and awaits for user confirmation with a
unobtrusive
dialog embedded at the top of the page before going to that new
URL. Some examples:
1) When a browser tries to load a URL, call it "X", and the HTTP
response header returns a redirect address, or there is an http
equivalent redirect specified in the HTML meta tag, "X" is skipped
when pressing the back button later. This is the standard browser
behavior in order to make the back button actually function.
In frame-by-frame mode, "X" is not skipped. You can go back to
"X",
and any automatic URL changes are intercepted. The user can then
inspect what the url address "X" was, or, if it was an HTML tag
that
specified the redirect, the user can take a glance at what content
that HTML page showed, if any.
2) When loading a page with URL "X" that does JavaScript redirect,
as
opposed to an http response header redirect, this can cause a
problem later when the user presses the back button, because the
browser may faithfully load "X". And when "X" is loaded, the
redirect
is automatically triggered, making it difficult for the user to go
back
twice. Usually when this happens, the user has to press down on
the
back button to get the browser history menu and skip a URL from
that.
In frame-by-frame mode, when the user goes back to page "X", the
auto-redirect is intercepted, allowing the user to press back again
to
go before "X".
3) Page "X" embeds within its page, another page "Y" from another
domain, using iframe tags. The content author of the page "Y" that
loads in the iframe does not appreciate this, so page "Y" has a
javascript redirect that causes page itself to break free from page
"X". That's fine and dandy for the author of "Y" because it gives the
author more control. But what about the user's control?
In frame-by-frame mode, the user gets to arbitrate between what
the author of "X" wants and what the author of "Y" wants. By
preventing the page from automatically changing its address
without
user control, the user can now press the back button and get to
see
"Y" in "X" as the author of "X" intended, so that the user can view
what's in "X" and maybe click something useful on that page. (e.g.
diggBar, Google Images)
Those are three examples, but there may be more unlisted cases
where giving users the control over the content authors' is
benefitial
to the user, in the form of frame-by-frame navigation. However,
this
is a tradeoff between the conveniences of having everything
automated for you versus being able to control everything. In this
case, the control over URL changes.