As much as I would be happier to just ignore it, there is something about the increasing Longhorn hysteria that is reminiscent of the depths Apple slid into in the mid-90s.

There were a succession of enticing technology demos promoted as seeds of totally new architectures, more than a couple of which almost survived deployment then in the process of their ultimate abandonment burnt many fans.

But the ask was always too big, just the same as it has always been with every other monolithic attempt at software over engineering.

The one thing we can count on from Microsoft is that they will eventually bring out something which they will tell us is Longhorn. They are too political to contemplate honest abandonment. But all they will ever deliver will be cherry picked features grafted onto their already long suffering underlying architcture.

Apple needed two things to break the cycle of unleliverable promises and finally produce the most significant new OS since IBM's VM: the only multi-achieving arse-kicker in the industry and a decade of somebody else somewhere else evolving the other half of their answer. (The fact that that "somebody else" was the same "arse kicker" on out placement also served to smooth a few wrinkles.)

One fact of life ignored by believers in seven day solutions is that there is a complexity ceiling which applies to engineering and logistics, beyond which more complexity can only be added by an evolutionary mechanism--variation plus selection. If it were not so serious it would be amusing that those who proclaim complexity as evidence for design are so far off target.

But back to Longhorn. My betting is that there will never be another truly new OS from Microsoft, or at least not while their market share remains significant. They will continue to lie about it and continue to fail to deliver. Intel and AMD all over again, yet even Intel has a more resilient basic culture to fall back on, IBM-style, than the isolates from Seattle.

Of course there is never an easier place to make a dishonest living than on the coat-tails of an empire in decline, so if that's your style don't let me discourage you. Just keep the dollars circulating.

Earlier this week I decided it was time to announce on NKS Forum the little project that has been keeping me distracted these last 3 months.

That announcement contains some disclaimers on possible browser issues where I am consciously pushing the boundaries, and it is also probably the best place for discussion, although comments are certainly welcome here.

Tick Tock might not be the kind of Class 4 system I prefer to hunt, but it may be a useful step towards doing some useful work in the field of evolving graphs which is increasingly theorised as being fundamental but largely neglected for detailed study because of perceived perceptual obstacles.

This may not strictly be a "note for nowhere else" but I am test running it here first because of the frequency with which assertions turn up on /. about the likes of the works of Shakespeare turning up in the digits of pi.

The importance is that even world renouned mathematicians get this wrong way to often. I just finally started reading Roger Penrose's Shadows and quickly found him confounding the number of algorithms with a countable number (even though he had just made mention of Cantor and "diagonal" in the same breath) in his effort to make a case that there are things human minds can do that are not computable. Meanwhile, in trying to make the opposite case, Wolfram cites the "oracle" argument which contains exactly the same misunderstanding.

While there are a whole lot of technical arguments that can be got into about orders of infinity, there is really only one distinction that matters to our everyday understanding of the world, that between aleph 0, countable infinity, and aleph 1 (and higher) which is uncountable.

I could just make the point that actualities are countable and possibilities are uncountable, but that would be putting the cart before the hourse.

First we need to recognise that any form of infinity is just a mathematical construct. We are not ever going to have to deal with actual infinities, but what is very useful is understanding limiting behaviour as something or other goes towards infinity. It is here that the difference between countable and uncountable infinities matters immensely.

countable

can be put into a one to one correspondence with the natural numbers

uncountable

cannot be put into a one to one correspondence with the natural numbers

The digits of pi, the keystrokes made by a room full of monkeys, the galaxies, stars, grains of sand, atoms are all countable. Each could be given a number.

Possible texts, possible algrithms, possible macromolecules, possible species, potential humans cannot be given a number. Each is a unique combination of a long sequence of units and every time you add one to the length of the base sequence, you multiply the possibilities by some factor greater than one, often much greater but even that does not matter except in the limit.

So if you struggle to count the possibilities of a small complex system, just making that system slightly larger multiplies those possibilities out of all proportion and counting them becomes orders of magnitude harder.

Even if, as I strongly suspect, our cosmos is but a semen stain on the sheets of some grand cosmological process going on beyond our event horizons, there is no "other earth" anywhere else. There is no other you. There is no other me! Fortunately!!

Hopefully this means that whatever had been preventing me posting here from my ISP (one of Australia's big 5) has finally gone away.

When such problems persist, it is all too easy to just forget about the possibility of posting anything.

Similarly, I'm becoming more and more aware just how easy it is for one's involvement in once all consuming online communities to just fade away without notice.

That awareness hasn't changed my development objectives, but it has again focused my attention on some of the deep impediments to truly empowering online communities.

Perl 6 was already very interesting even before Larry Wall's Apoclypse 5 brought us regexes we can build grammars on and that has sure stirred up the excitement level, as evidenced by this week's Perl 6 Digest (from which there doesn't seem to be a link to other recent P6P Digests, not that I could afford the time if there was).

To me, Perl 6 with it's strong linguistic roots has emerged as the great hope for tackling two major development objectives:

• integrating open ended functionality definition within a content management framework (and here I need to also keep in mind the need for CVS or Subversion equivalent functionality)
• enough analysis of natural (and other) language to make serious inroads on redundancy problems including both deliberate and inadvertent resubmission

If only there was time to draw an updated road map.

Arguably a bit more substantive than Mathematica co-developer Theo Gray's Periodic Table Table, Stephen Wolfram has invested much of his returns from Mathematica and his considerable intellectual ability into a long anticipated 1197 page tome that provides theoretical underpinnings for the notion that seemingly complex systems may be best understood by seeing the universe as an irreducable super massively parallel computer for which the cellular automata Wolfram has long been an authority on are a useful model.

The New York Times has an article on A New Kind of Science which I received links to from two unconnected directions within hours of its publication.

I am reminded by one of them that it 'hits the bookstores on Tuesday, May 14'.

A lateral though: time to proclaim the study of "paesthetics" being the destructive effects of ubiquity on human sensibilities.

PowerPoint, Excel, Word and other Micro$oft-ware aren't the only paesthetic topics, just some of the more relevant at /.

Another notable is the monospaced type written page and its infected brethered: e-mail, Usenet and IRC. (Monospaced is more sensible in a code intensive environement but the antithesis of readability when used for natural language.)

Moving further afield, students of paesthetics might interest themselves in money as a measure of value, or cream bricks, red roofing tiles and near white plaster walls as corrupters of architecture.
 

My mind is mostly scrambled between issues of community, of praxis and of theoria.

Slashdot provides a powerful exemplar of community and praxis, by which I here mean my 20 year old ideas for a Public Information Communications and Access system, which has morphed into my current efforts to pull together TransForum version 2.

Slashdot's Science section adds one more temptation to the range of sources which add data to my theoretical explorations of what in 2002 in Oz it has again become accepted to label "complex systems" but which I see more and more as part of the Big Picture.

My experience of community has been the best antidote for flawed notions of control and largely become something to be escaped from so I can focus more of my aging energies on the theoria and the praxis.

Maybe I should try to use this little corner of cyberspace to expose a few thoughts on subjects that might be better suited to this audience than to the handful of aging and mostly private TransForums which I sometimes use as sounding boards.

That might also depend on whether anybody else wants to join such an obscure discussion.