In article <fqn5rb$tfq$1@xenon.Stanford.EDU>,
amorgan@xenon.Stanford.EDU (Alan Morgan) wrote:

> In article <see-34B202.10423806032008@lust.ihug.co.nz>,
> Barb Knox <see@sig.below> wrote:
> >In article <47CF0BB3.15CDB9E6@bytecraft.com>,
> > Walter Banks <walter@bytecraft.com> wrote:
> >
> >> Barb Knox wrote:
> >>
> >> > In article <yOqdnWCX2uUPd1PanZ2dnUVZ8surnZ2d@bt.com>,
> >> > Richard Heathfield <rjh@see.sig.invalid> wrote:
> >> > [SNIP]
> >> >
> >> > > When scientists make claims that are not verifiable, their credibility
> >> > > becomes important. When they say stupid things (e.g. [...] "random
> >> > > data cannot be compressed"), I am likely to reassess their credibility.
> >> >
> >> > In what sense do you believe that random data can be compressed? (In
> >> > the usual sense of the word, statistically random data N bits long
> >> > contains N bits of information, and that can not be represented in less
> >> > than N bits.)
> >>
> >> Take a page from chaos theory even in your random data there will
> >> be patterns
> >
> >If there are patterns then it is not statistically random.
>
> Well, no. There will always be patterns if you have enough data, but
> that doesn't mean that the patterns will be long enough to take advantage
> of.

Good point.

> This distinction, of course, is not one that the random data fans
> understand. They see the occasional 1234567 in a 20GB file as a chance
> for compression. The rest of us see it as a chance to waste precious
> bytes describing the location, length, and form of the patten, resulting
> in, if they are very, very lucky, no gain at all.
>
> Alan

--
---------------------------
| BBB b \ Barbara at LivingHistory stop co stop uk
| B B aa rrr b |
| BBB a a r bbb | Quidquid latine dictum sit,
| B B a a r b b | altum viditur.
| BBB aa a r bbb |
-----------------------------