O'Reilly WebBoard

Gregory Alan Bolcer (gbolcer@gambetta.ICS.uci.edu)
Wed, 16 Sep 1998 06:14:25 -0700


I recently did a review for O'Reilly's
WebBoard 3.0 [1] for the Atlantic Monthly [2].
The design of the perfect webboard is in
some ways the same as the search for
the holy grail. The only way to overcome
individual complaints and annoynaces is
to allow individual customization used
with shared data. Unfortunately, all
my criticisms were discounted because I
was told I have a fundamental bias against
their atrocious implementation of frames and
their insistence on using client side cookies,
one of which stores your password unencrypted.
The very first thing I did was write a javascript
program to go through and change everyone's
password.

Not feeling particularly biased, I'd like to
know what is the general consensus on frames
and client-side cookies on FoRK?

Are most frame-based Web pages designed poorly or is
there some fundamental flaw with non-overlapping,
tiled space allocation? After all, the CS
community as a whole stopped supporting tiled
window manager almost 2 decades ago because of
it's poor design.

Also, what is it about East coast publications [3]
that feel the need to restrict access to material
unless they can explicitly track your access
to it? Are most client-side use of cookies
abused? Are they necessary? What's wrong
with keeping server-side cookies for login
services?

Greg

[1] http://webboard.oreilly.com/ sells for $699. Someone's
got to make a better mousetrap.
[2] http://www.theatlantic.com/
[3] http://www.nytimes.com/