2018 - I should have more time to post

  • I hope to retire mid-year from my decidedly non-nanotech related job, which will give me more time to devote to nanotech stuff. One task I'd like to do is try my hand at using some of the existing molecular modeling tools to determine their strong and weak points with respect to design of molecular machines. Probably someone has already done this, so that would be the first thing to check. As chemist Frank Westheimer once observed: "A couple of months in the laboratory can frequently save a couple of hours in the library."

  • Great,
    IMO one of the most advanced and interesting things that are currently going on
    are the SW projects of Christian Schafmeister.
    Here's a presentation:


    Then there also is:
    http://cadnano.org/
    Which I haven't yet looked into.


    As for "Nanoengineer-1", AFAIK it has been abandoned by the original developers.
    Eric Drexler somewhere did mention a few details about the exact reasons (can't find it right away).
    I think the development ran into a dead end probably due to a combination of
    * unready SW technology (especially GUI side) and
    * technical debt


    We really need functional reactive programming (FRP).
    As far as I can see this is the only way to cleanly separate logic from interface
    and have GUI programs composable instead of wheel-reinvent throwaway.
    (That would change orders of magnitude in productivity.)


    NE-1 was picked up by Bruce Allen ***
    http://moleculardynamicsstudio.blogspot.com/


    But:
    * I think this might now go in a bit different direction. Straying away from APM.
    It seems the focus went more to highly accurate predictions for research than for
    robust feature rich development & exploratory engineering.
    * NE1 is a conglomerate of deeply imperative base libraries => I'd wish for pure FRP wrappers. -- But that is not a goal.


    ----


    Nov. 2016 I had some mail traffic with Tom Moore ( http://machine-phase.blogspot.co.at/ )
    about APM related SW in general, installing NE-1 on Ubuntu, and what I would concretely wish for in particular.
    Here's a crop-out of what I wrote:


    "In the long run I want to see detail mipmapping, seamless zoom-out to bulk limit approximation,
    black box subsystems with physical geometry and logistic IO parameters (lazy deferrable!).
    And maybe polar logarithmic multi-scale view: ..."


    I think system level focus this is not really where *** is going.


    I also noted the major point that auto-equilibration is not scalable in NE-1 (major weak point)
    and pondered about the internal implementation and ways to improve this.
    Gotta look into the equilibration source-code sometime.


    ----


    "A couple of months in the laboratory can frequently save a couple of hours in the library."
    What a funny reverse formulation.
    It's about avoiding the tedious rediscovering of unknown knowns.


    But that sometimes works in reverse just as well.
    "A couple of months in the library can frequently save a couple of hours in the laboratory."
    Then it's about avoiding the tedious reconfirming of known unknowns.


    ----


    Other things coming to mind regarding molecular modelling software: nanohive, boinc, ...
    & the preparation for colour 3D printing of models I have done but with which's process I'm not very happy with yet.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!