[cig-commits] r14993 - short/3D/PyLith/trunk
brad at geodynamics.org
brad at geodynamics.org
Thu May 14 09:43:47 PDT 2009
Author: brad
Date: 2009-05-14 09:43:46 -0700 (Thu, 14 May 2009)
New Revision: 14993
Modified:
short/3D/PyLith/trunk/TODO
Log:
Worked on TODO list.
Modified: short/3D/PyLith/trunk/TODO
===================================================================
--- short/3D/PyLith/trunk/TODO 2009-05-14 03:17:42 UTC (rev 14992)
+++ short/3D/PyLith/trunk/TODO 2009-05-14 16:43:46 UTC (rev 14993)
@@ -2,19 +2,59 @@
CURRENT ISSUES/PRIORITIES
======================================================================
+Matt
+ Memory model
+ field split
+
+Brad
+ Test precompute geometry
+ Nondimensionalization
+ cleanup
+ full-scale testing
+ test uniform refinement
+
+Charles
+ 3-D Power-law rheology
+ C++
+ libtests
+ Python
+ pytests
+ full-scale test (axial compression/extension?)
+ 2-D Maxwell viscoelastic rheology
+ C++
+ libtests
+ Python
+ pytests
+ full-scale test
+ Initial stress, strain
+ 3-D Maxwell model
+ 3-D Generalized Maxwell model
+ 3-D Power-low model
+ 2-D Maxwell model
+ tests
+ Manual
+ Materials
+ Savage-Prescott benchmark
+
-1. Fix Tabrez's bug. [Matt]
1. SNES
+ Cleanup SolutionFields.
+ Remove time history stuff. [TEST]
+ Replace solveSolnVec with solution. [TEST]
+
Power-law rheology
2. Nondimensionalization
- Add Quadrature::computeGeometry(cell, coordinatesCell)
- do computation on 1 cell
- much less memory than holding all info in Fields
- keep computeGeometruy(mesh, cell) and retrieveGeometry(cell)
+ Add units to spatialdata. [UPDATE TESTING]
+ Add Quadrature::computeGeometry(coordinatesCell)
+ Use #define PRECOMPUTE_GEOMETRY [TEST]
+
+ MeshIO: coordinates and coordinates_dimensioned [UPDATE TESTING]
+
DataWriteVTK needs dimensionalizer
Done coordinates in output (nondimensional instead of dimensioned)
@@ -22,8 +62,6 @@
(nondimensional instead of dimensioned) const slip rate in output
(slip rate not slip)
- Add ability to use symmetric matrix.
-
Ask constraints if a block matrix is okay. If okay and matrix type
is "unknown" (not set by user), then use "baij" rather than
"aij". Do this in Python.
@@ -54,6 +92,7 @@
7. two fault (static, parallel)
8. gravity w/initial stress and strain
+ Gravity with faults, Neumann BC.
4. Add missing unit tests
@@ -113,8 +152,7 @@
5. Tidy up
- Add error message if number of time steps computed by time
- stepping scheme is 0.
+ Replace memcpy() calls with loops.
How do we determine the orientation for a fault in a 1-D mesh? We
assume normaldir is +1.0, but cohesive cells could be created so
@@ -160,8 +198,6 @@
6. Memory model [Matt]
playpen/memcheck/*
- Must add nested stages so that logging totals are correct
-
7. 2-D Plane strain Maxwell viscoelastic rheology [Charles]
8. 2-D Plane strain Generalized Maxwell viscoelastic rheology [Charles]
@@ -207,7 +243,6 @@
Uniform refinement (debug, check, all cell types)
-Cuthill-McKee ordering of vertices
======================================================================
KNOWN DEFICIENCIES
More information about the CIG-COMMITS
mailing list