[Sigia-l] Yay! Flame on! was "Who Really Turns Off JavaScript?"

[Alexander Johannesen]

Hi,

On 11/9/05, Timothy Karsjens <tim at karsjens.com> wrote:
> Yes, I was offended.

I'm interested in *what* exactly though, because it seems that the
dispute is over stated cold facts and an emotional reaction to
repetition refuted. The wholesale of statistics? I think Ziya *was*
venting shere frustration (or possibly even surprise) over even
flashing these kind of statistics in our eyes as proof of anything,
but he can chime in if he wants to ...

> In the greater scheme of things, I have to wonder whether or not I make
> my points as concisely as I should on this list.  Then, I have to wonder
> if there is anything that i can do to make my point make sense to certain
> individuals that are so, mired.

It could be that you *are* making your point as clear as you can *and*
be wrong at the same time that makes certain people react to your
statements. :) But hey, just guessing here. The certain mired people
you're reffering to *may* actually have a point. I'm not saying this
in a flamatory way, just pointing out that even if you percieve fault
it doesn't *nescessarily* make it so. Let's explore that option ;

> We are talking about new things that are out there in the industry, and my
> point about hitting 100% functionality still stands on even more solid ground
> because what we are talking about is *new* and *not* standard.

What is your definition of "new"? AJAX is a new *name* for something
I've been doing for at least 6 years in various incarnations, 3 with
XMLHttpRequest, and I've learned a lot in those years, enough to
design systems that reach roughly 90% user-guessed browser
compatibility, which really is all you can hope to achieve.

And seriously, all you need to work towards to; it's about using the
standards more than tricking subpar browsers and / or users into your
definition of what "100%" and "75%" really mean. It is contextual and
certainly biased, unless you've got concrete stats about all your
users, and comparative stats from everyone you're talking with, which
I dare say you nor I nor anyone don't really have.

Also, just as an aside, *testing* wether a browser has JS turned off
or simply don't support it or don't support your test, is hard to do
(both in terms of how and the expense of it) and can't be done on a
http level and hence most often isn't tested but simply inferred from
the http level information. It is often misleading, and *certainly*
not proof.

> I do not care if I use a decade old technology, as long as the functionality
> works 100% of the time.

Isn't that, as Ziya says, a pipe dream? How can you achieve 100%
functionality in *anything*? In fact, what is the definition of "100%
functionality"? That everything intended, any idea and tested function
taken down into code, works all the time without ever being wrong? I
think you'll find that claim to be hard pressed against reality, no
matter if they're computers, cars, beds, houses, toys ...

>  If I have a captive audience, where I can be assured that using something
> new will work 100% of the time, then I will use it.

Hmm, I'm sure you're thinking of something specific with this example,
but I don't understand what that is.

> My point in a nutshell is this; I am not comfortable delivering a product that
> does not work 100% of the time.

Then you should feel unconfertable *all* the time. :) Or are you
claiming you never have bugs? Never have a poor design? Or a bad idea?
Or you don't rely on other peoples hardware or software? Things always
just work exactly according to plan? If all this is true, then you are
beyond human and there would be hundreds of large companies just dying
to employ you.

I think a lot of "100% functionality" comes from systems that are
built robustly. But robust systems (to go back tot he computer world
again) are designed with systems failing *in mind*, meaning the design
knows about systems failing from time to time; they *know* they can't
be 100% functional all the time. Functional means there are inputs and
outputs and the completeness of functionality is that the output
matches the functional specification, and unless your functional spec
is artificially small I think that the "100%" thing is a meme we hail
for kicking our arses forwarded but hardly something we can hope to
achive in reality. IMHO, of course.

> Other people seem to be comfortable with not doing that, but I am not.

I'm comfortable knowing that I've designed a system that follows just
the standards, and yes, I'm comfortable knowing that there are people
out there that can't get 100% functionality out of my system because
of circumstances they themselves have put themselves in (being
personal or organisational reasons). I let the standards and the ideas
behind them guide me more than the 100% engineers wet pipe dream.

> I am comfortable with where *I* am in the industry.  I have never had an
> unhappy client, period.  I have always delivered 100% usable functionality,
> period.

I've never had an unhappy client tell me so either, and I've delivered
systems that certainly ain't got 100% "usable functionality" (not sure
what that means, but I'll guess that it is the part of your systems
functionality that actually is usable :). And yet we quibble like
this.

> It is all about doing what works.

First you need to find out what "works" really mean.


Alex
--
"Ultimately, all things are known because you want to believe you know."
                                                         - Frank Herbert
__ http://shelter.nu/ __________________________________________________