From: fateman@gmail.com   
      
   On Thursday, June 21, 2018 at 11:16:04 PM UTC-7, Ralf Stephan wrote:   
   > This is using FLINT:   
   >   
   > sage: %time _=9^9^9   
   > CPU times: user 4.52 s, sys: 355 ms, total: 4.88 s   
   > Wall time: 4.88 s   
      
   why do you ask?  You hardly need a CAS to try to compute this (big) number.   
   It is one line in any Common Lisp implementation.   
   It is a simple library call in any number of libraries.   
   But that requires an answer to the question ...   
      
   And what would you do with it?   
      
   If you need the first few digits of it, here is a result   
   from Maxima   
      
   4.281247732335055b369693099  where the b signals the   
   beginning of the bigfloat exponent.   
   Of course, if you need more digits, like the first few thousand,   
   that is also available.  It appears to be about 0.00031 seconds   
   (averaged over 100 trials)   
      
      
   If you were to print this integer in full   
   out on paper, say 80 characters per line, 65 lines per page,   
   it would take over 71,000 pages.   
   on 20-pound bond paper, it would take 142 reams of paper at 5 pounds per ream.   
   Or about 711 pounds.  which is 322.5 kilograms.   
   (If you print on both sides of the paper, cut that in half.)   
      
   These comments should not be construed as a criticism of FLINT,   
   which I believe may indeed be faster than any other   
   comparable system for bignumber integer arithmetic.   
      
   Most people do not use a CAS for number theory.   
   If number theorists use a CAS, that's nice, especially   
   if they contribute nice software.  However, harnessing   
   others' efforts to satisfy their needs may be a   
   diversion of valuable resources to a small group.   
      
    Mr. Spock   
   said it nicely.  "The needs of the many outweigh   
   the needs of the few :)   
      
   RJF   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)