[thelist] targeting effectively (was: navigation through form posting)

[aardvark]

> From: Matt Liotta <mliotta at iname.com>
>
> I never said I was working hard to reach 90%. I was pointing out that
> you can easily reach 90% of your target, but trying to get the last

actually, it's more of an 80/20 rule, instead of 90/10... even if all your users
are on IE6, for instance, a chunk of them will have some whack installs, be
behind firewalls, be using it on something like AOL, be surfing with stuff
turned off, etc...

> 10% can cost you more than it is worth to do. Now maybe the situation

yes, it can, but don't assume that it will... don't just spout the 90% number for
all cases, because once you've done that, you've lost the ability to be
objective on a project-by-project basis...

i'm trying to demonstrate that your assertion is, for the most part, flawed... will
there be exceptions?  yes... but i think you're limiting yourself by arguing for
them first, and relying too heavily on browser-specific features...

> that some or even most people are in here allows them to easily get
> 99% of their target, but that doesn't change the fact that it will
> still cost more development and QA time to get the last 1%. You can

it may, or it may not...

in fact, i can get to that 99.5% mark on many projects, but only the 99%
mark on others... all *without* incurring extra costs... yes, there is a cut-off,
but instead of that cut-off being determined by my own assumptions, it's
based on that final little asymptotic curve that pushes it out -- and it's usually
not nearly as steep as you think...

> move the numbers around all you want, but the bottom-line is
> unchanged. There is simply no way to easily code for every browser and
> platform out in the wild.

well, i'd argue that...

on one hand, i can make a site that will work on 100% of browsers/platforms...

on the other hand, if i want to win a design award, i might have to cut that
back to 99%...