To start, here is the basic list of allowed tags in comments and articles:

<a>,</a>

<b>,</b>

<strong>,</strong>

<i>,</i>

<em>,</em>

<p>,</p>

<blockquote>,</blockquote>

<ul>,</ul>

<ol>,</ol>

<li>,</li>

<dl>,</dl>

<dd>,</dd>

<dt>,</dt>

<pre>,</pre> (which get converted to a textarea)

<code>,</code>

<cite>,</cite>

<br>

<img>

<hr>

Can you give us a brief overview of what you've done?

To answer your question, I'm using a big black box with a variable

speed motor attached to an alternating beater arm. Each tag is spun around

within this black box by this beater arm inside a metal colander with small

holes of varying sizes. Those tags not marked as safe have holes of

appropriate size that can be found in this colander. Those marked as safe

do not have such holes. As they're spun around, centripetal force pulls

them out through the hole sized appropriately for the given tag. What's

left is then poured out into a baking tin sprayed with 3 in 1 oil, baked at

12,000 degrees kelvin for 3.4 nanoseconds and promptly stored in the database to

cool. Serve chilled, neat, warm, or hellfire. Serves an unlimited number

and never spoils.

Wouldn't a teflon coated ceramic work better and resist corrosion?

I experimented with teflon and ceramic, both together and separately. I

abandoned them both for separate reasons.

I found that the sharp ends of the tags would scratch nearly all the teflon

coating off the colander after only one run. This resulted in excessive

long term memory loss in those the final product was served to, hence the

temporary increase of same subject questions on thelist in a relatively

short period of time.

In my experiments I found that the initial torque requirements necessary to

bring the colander up to speed was causing undue stress on the motor due to

its weight. The advantage of inertia at that point would have warranted

keeping it if it hadn't been so expensive, not to mention there wasn't

enough room in the current design of the black box, to install a set of

counter-action, retro-motion brakes. without them the colander would come

up to speed, spin out the mash in milliseconds and continue to spin for

hours. This resulted in unnecessary spilling when attempting to feed the

next string to process. We were losing data bits from unpredictable spots

in strings until we figured that one out. Some of the secrets of the

universe will never be recovered because of that mistake.

What's the leverage ratio of the arm?

It's actually a variable ratio leverage arm achieved by a parallel coupling

between the motor drive arm and the fixed arm on the beater. Depending on

the thickness of the string, the arm can adjust its pitch to match.

You have no idea how much rethinking we had to do just to test this thing.

We had to throw all of our testing methodology out the window.

Are you using needle bearings or cartridge?

It's actually using something called a polymorphal-floatfit. maybe i'll

write an article about that actually. It's pretty complex.

I would think there would be some concern with the possible under-torque in high-content filtering

Fortunately, this device is nothing more than an object that can be

instantiated for any number of requests. we even recently fitted it with an

attachment that would take care of its own cleanup and storage. All you

have to do is install it and it does everything else automatically.

Unfortunately we forgot an off switch and it can be pretty protective. So,

now that it's installed we're basically stuck with it.

Thanks for taking the time to speak with us.

Sure. Can I go now?