User: knarf Date: 2016/12/16 10:04 AM
Modified: /about/releases/ ET_2016_11_announcement.txt
Log: typos
File Changes:
Directory: /about/releases/ ===========================
File [modified]: ET_2016_11_announcement.txt Delta lines: +9 -4 =================================================================== --- about/releases/ET_2016_11_announcement.txt 2016-12-15 16:44:45 UTC (rev 1688) +++ about/releases/ET_2016_11_announcement.txt 2016-12-16 16:04:37 UTC (rev 1689) @@ -67,12 +67,12 @@ - Update linker flags to allow linking using newer versions of 'ar' - Add '.' to perl @INC directory to allow build using newer versions of Perl - Provide Cactus version as CACTUS capability to thorns - - Allow aligining interior (instead of just origin) of grid functions + - Allow aligning interior (instead of just origin) of grid functions - Output build timing information in VERBOSE mode * Formaline - Attempt to use fast methods where possible (hardlinks are available) - Only update files in thorns that changed - -> should make it noticibly faster on most systems + -> should make it noticeably faster on most systems * Simfactory - use of environment variables CACTUS_NUM_PROCS and CACTUS_NUM_THREADS in runscript of all machines. Update your private entries accordingly. @@ -89,7 +89,7 @@
=== Upcoming changes for the next releases
-The Tmunu paramer support_old_CalcTmunu_mechanism, will be removed after this +The Tmunu parameter support_old_CalcTmunu_mechanism, will be removed after this release. If you rely on this, your code is probably unnecessarily slow. Let us know if the removal would create a problem for you.
@@ -131,6 +131,7 @@ - Comet (#) - Draco (#) - Edison (#) +- Galileo - Gulob (#) - Gordon - Hydra (#) @@ -152,6 +153,10 @@ # ML_ADMConstraints: There seems to be a problem regarding calculating the constraints from the BSSN quantities. The exact problem is not known yet, but indications point towards a compiler problem (Intel in all cases). + In all cases the metric data (evolved with ML_BSSN) is identical to the test + data (to round-off error) and only the ML_ADMConstraints data differ + significantly from the test data. If output statements of relevant + quantities are inserted into the code, the differences disappear One consistent symptom is failing of the Dissipation and RotatingSymmetry testsuites (as they include output of the constraints using ML_ADMConstraints). Investigations on this matter will continue, @@ -175,5 +180,5 @@ Erik Schnetter Barry Wardell
-December 8th, 2016 +December 16th, 2016