Difference between revisions of "TCM-July-2008"

From Termination-Portal.org
Jump to navigationJump to search
(To discuss)
Line 25: Line 25:
 
** RAM usage (unlimited I guess)
 
** RAM usage (unlimited I guess)
 
** CPU usage (defined by wall time, so one may use 16 cores? CPU time, so 60 seconds singlethreaded translates to about 4 seconds with 16 threads?)
 
** CPU usage (defined by wall time, so one may use 16 cores? CPU time, so 60 seconds singlethreaded translates to about 4 seconds with 16 threads?)
 +
** For CPU time limit the competition can run faster (if the tools run multithreaded), but there is no incentive to do so.
 +
** My opinion: Because the machines have 16 CPU cores each, multithreading should be encouraged (read: wall time!)
  
 
* details for Termination Workshop 2009 Leipzig
 
* details for Termination Workshop 2009 Leipzig

Revision as of 15:58, 15 July 2008

Date/Time/Location

15 July 2008, 6:30 pm, during RTA at Hagenberg

Agenda

  • current status of the competition test run
    • the "Torpa problem" (what can programs expect on $PATH, $PWD)
    • the "Jambox problem" (programs that write long proofs)
    • Q: test run for certified termination?
  • next steps
    • submission of new prover versions
    • submission of new problem (sets)
    • test runs (by submitters)
    • complete runs (by administrators)
    • evaluation (by committee) (how and when)
  • new category suggestions
    • complexity analysis (Georg Moser, Johannes Waldmann)
    • SN infinity (Hans Zantema, Jörg Endrullis)
    • we have more categories than people?
  • To discuss and clarify
    • RAM usage (unlimited I guess)
    • CPU usage (defined by wall time, so one may use 16 cores? CPU time, so 60 seconds singlethreaded translates to about 4 seconds with 16 threads?)
    • For CPU time limit the competition can run faster (if the tools run multithreaded), but there is no incentive to do so.
    • My opinion: Because the machines have 16 CPU cores each, multithreading should be encouraged (read: wall time!)
  • details for Termination Workshop 2009 Leipzig
    • date? (needs to be decided soon, to book the venue)
    • length/schedule? suggestion:
      • (day 1) sub-workshop on certified termination
      • (day 2, 3) termination workshop (if there are more submission, then already start on afternoon of day 1)
      • (day 4) would be possible to hold the IFIP WG 1.6 meeting
    • program committee? suggestion: informal workshop, no refereeing, small committee, include people from communities of termination of logic programming, imperative programming