Re: Crypto Mini Faq

From: Paul Pires (diodude_at_got.net)
Date: 03/30/04 

Date: Tue, 30 Mar 2004 20:51:00 GMT

Tom St Denis <tom@securescience.net> wrote in message news:6Xkac.568$jS5.97@news04.bloor.is.net.cable.rogers.com...
> Joe Peschel wrote:
>
> > Tom St Denis <tom@securescience.net> wrote in
> > news:qikac.2125$Kr1.50@news04.bloor.is.net.cable.rogers.com:
> >
> >> What gets me more than the occasional mistake [let's recall knuth is
> >> on his third edition after all....] is the lowering of the bar...
> >>
> >> Sure crypto can be taught well... but being taught well doesn't mean
> >> being taught stupid.
> >>
> >> E.g.
> >>
> >> p4 of that excerpt...
> >>
> >> "encrypt - Scrambling data to make it unrecognizable"
> >
> > "Scrambling data" doesn't sound so so bad.
>
> Scrambling is just such a lame word.

It's a great word. It embodies confusion and diffusion
both. If I give you 4 scrambled egss, can you give me
back just the yolks?
>
> >> Um how about
> >>
> >> "encrypt [or encipher] - concealing the meaning of a message"
> >>
> >> Not only is the latter description accurate but it is more meaningful.
> >
> > How would this description be more meaningful to the intended audience?
>
> Well because... um ITS WHAT YOU ARE DOING!!!!!!!!
>
> Message has a meaning... you are concealing it.
>
> Maybe there should be "information theory for dummies" first...

First we need the idiots guide to flame proof Pseudo-code
composition.

>
> >> What exactly is "scrabling" data?
> >
> > I dunno, either.
> >
> >> I mean you could permute the data
> >> and it's [literally] scrabled.
> >
> > It is?
>
> Um, you permute eggs to "scramble eggs" right?

Nope.
>
> ;-)
>
> >>That's not "encryption" by any means.
> >
> > But if you permute the data, it is scrambled and it is encrypted.
>
> Damn, screw this, ethics are soooooo last week.
>
> >> The "books for [yuppy impatient] dummies" are just symptomatic of a
> >> greater problem.
> >
> > What problem is that?
>
> That people want a quick fix to all problems. Hey, you wanna learn a
> science? Great, pick up some books, learn, experience, rinse and repeat.
>
> No, instead people want a reward right now. Hence the low quality book. I
> bet if Knuth spent 3 months writing TAOCP it wouldn't be as popular as it
> is now.... Hard work is supposed to be rewarded not a hockey-season worth
> of tacking notes together.
>
> Have you ever walked in a book store recently? I bought TAOCP from a
> chapters for crying out loud [um about 5 years ago]. Now? You can't even
> find a single math text in there. It's all "self-help" and "ASP.NET for
> dummies" bulldung.
>
> Point is why can't people take things with a modicum of pride and respect.
> Being a cryptographer is not "easy" by any stretch of the imagination. You
> have to be good at math and computer science, you have to be able to attack
> things from weird angles, think like the "enemy" and use the rules of
> science. [This is not unlike other fields I guess].
>
> Or put it another way. You spend [in my case] nearly 10 years teaching
> yourself computer science via literally 1000s of different "labs" [e.g. how
> todo BWT compression, how todo BN math, how ...blah] only to find some
> jackass who read some fluffy book and can spew out the latest verbiage gets
> a job instead of you....

Hmmmm. Carrying around a bit of charge are we?
How about a grounding strap for testy Tom's?

Paul
>
> Tom