Meeting notice: 09-018-98; 7:30 pm.

Note: New Meeting Place

        -- NE-43 is being rehabbed.  This week we will
        meet instead at the Royal East at 782 Main St.
        Cambridge (a block down from Mass ave.) Same
        time.


Suggested topic:  Fry's Bottleneck


Chris Fry points out that

        -- NT seems to assume very large programs with
very low error rates

        -- that we don't begin to know how to write
programs that perform reliably on these levels

        -- that there is no very scalable solution to
the problem of writing bugfree software in sight

        -- and that nonetheless, most of us assume that
the fabrication problems involved with NT are
predominantly about controlling atoms, and if those
were solved the appropriate controlware would come
along in a reasonable time.  He asks: Is this bravado
or realism?

Biology seems to question the first premise, since it
offers examples of complicated entities being built out
of very short programs.  The human program only has
100,000 instructions in it, which seems like a
manageable number from the pov of bug control.

However, it is possible that these short codes are
tightly adapted to a specific and limited set of atoms
and assembly procedures, such that if we were to try to
expand these operations with nonbiological materials
and new twists like Drexler's 'machine-phase chemistry'
we would end up back in Fry's Bottleneck: being forced
to write programs too big for our quality maintenance
britches.


Mike Newhall passes this on:


                Want to Balance a Battleship
                     on Your Head?


        Nanotubes--tiny carbon straws just billionths
        of a meter in diameter--are thought to be some
        of the toughest stuff ever made. But their
        small size makes these mighty tubes hard to
        test. Now physicists say that simply by
        studying the shape of the tubes as they bend,
        they can tell how they hold up under large
        compressive forces. They conclude, in the
        current Physical Review Letters, that nanotubes
        can take a hundred times the stress of any
        other known fiber without buckling.

        Press on a drinking straw from both ends, and
        at some point it will buckle. Lacking any way
        to grip and crush a single nanotube, physicist
        Daniel Wagner and colleagues at the Weizmann
        Institute of Science in Rehovot, Israel, tested
        them en masse. The team threw a dash of
        nanotubes into a liquid polymer resin and let
        it harden. As the plastic hardened, it shrunk,
        putting tremendous pressure on the embedded
        tubes. Near the surface, some of the tubes
        happened to be fixed at both ends but free in
        the middle. These tubes buckled, forming loops
        and kinks that stuck out like the fibers on a
        rug.

        The arc and shape of the buckles depends on the
        compressive strength of the tube, Wagner says.
        Using transmission electron microscopy, the
        team imaged individual tubes and sized up the
        kinks. From that they were able to calculate
        that the tubes could withstand a compressive
        stress of roughly a hundred billion Pascals
        (equivalent to the pressure you would feel
        balancing a battleship on your head) without
        buckling. That's about what theory predicted,
        Wagner says, "but the numbers are so gigantic
        you're almost afraid to publish them."

        "What this shows is that the tubes are indeed
        far stronger than anything else," says Jerzy
        Bernholc, a physicist at North Carolina State
        University in Raleigh. It also indicates that
        the tubes are sticking strongly to the plastic,
        otherwise the plastic wouldn't be able to bend
        them, he points out. "What you have is a highly
        reinforced polymer." If someone can figure out
        how to make tons of tubes cheaply, Bernholc
        says, combining them with polymers could make
        superstrong materials for products such as
        bullet-proof vests.

                --DAVID KESTENBAUM

Announcement Archive:
http://world.std.com/~fhapgood/nsgpage.html.

hapgood@pobox.com