From: nuny@bid.nes   
      
   On Monday, December 26, 2016 at 9:31:31 PM UTC-8, johnny1...@gmail.com wrote:   
   > On Saturday, December 17, 2016 at 10:24:17 AM UTC-6, Doc O'Leary wrote:   
   > >    
   > >    
   > > > It's pretty pointless to talk about "long-lasting" without agreeing   
   > > > what order of magnitude timeframe we are talking about.  A century?   
   > > > A billion years?   
   > >    
   > > It’s not even a question of that, but what *purpose* is to be served    
   > > by something lasting for a long time.  Yes, a tree can last for    
   > > thousands of years, but the “tech” of a rock can last even longer.   
      
     The "tech" of trees has lasted for millions of years too, but if the   
   "purpose" is to launch an organic satellite on a stage tree in two million   
   years, some DNA stuck in amber isn't going to be sufficient.   
      
   > I once posited, in a discussion of how buildings and structures could endure   
   > for geological ages with only automatic maintenance, that one good approach   
   > would be for the maintenance robots to replicate (with the necessary   
   > precautions against Darwin Error),   
      
     Then we have to worry about the robots degrading over time. That can be   
   handled by the robots just building "brand new" replacements every so often,   
   say at 3/4 of the lifetime of their internal component with the longest MTBF.   
      
     I don't mean just raping themselves for parts, I mean melt each other down,   
   refine and forge new parts and assemble them of course. So that means   
   maintaining forges and machine shops and semiconductor foundries for their   
   brains for X years, too.   
      
     The major "Darwin Error" is going to be in their data storage system- the   
   instruction sets they use to do maintenance on the machine itself, and on   
   themselves, and to build replacements. Triple redundancy provides adequate   
   data accuracy for current    
   systems intended to last years but for something intended to last millennia we   
   need something completely read-error proof.   
      
     Now we're back to the "store lots of data for millennia" thread.   
      
   > but for the structures themselves to be   
   > big, simple, and made of durable materials with long-lasting properties.   
      
     "Durable" is relative to what it has to endure.   
      
   > That is, stone.  The image I conjured up was that alien ruins made of heavy,   
   > chemically enduring types of rock with good compressive strength, and the   
   > robots would cut new stone blocks and replace damaged ones as needed.   
   >  Cyclopean Lovecraftian stone buildings, maintained by alien robots...   
      
     ...and if we built it of granite on Venus before its greenhouse effect ran   
   away, it would melt and erode to rubble soon afterward.   
      
     We could have built it of flint, chert, or the stuff the 1.8 billion-year   
   old Banded Iron Formation is made of... in the place that would later be known   
   as Chixculub...   
      
     Or the local star may get flare-y, or another star may get a bit too close   
   (that happens over dozens of millennia) and go nova, or disturb our Oort cloud   
   and start a rain of comets:   
      
   https://en.wikipedia.org/wiki/List_of_nearest_stars_and_brown_dw   
   rfs#/media/File:Near-stars-past-future-en.svg   
      
     For that matter, at about the same time as the Toba Genetic BottleNeck, a   
   faint red double star, Schotz's Star, apparently entered our Oort cloud:   
      
   http://www.centauri-dreams.org/?p=32602   
      
     and may have swapped a planet or two with Sol a la Campbell's _The Black   
   Star Passes_, which certainly would have made for "interesting times" a mere   
   70Kya.   
      
     How many environmental contingencies can we reasonably plan for?   
      
      
     Mark L. Fergerson   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)