Laundry List   -  23 Feb 2005

\item (*) H_T ContFrac cost-vs-mres tests

\item (*) Add force monitoring for HMC. Currently, we have energy
contributions for each monomial. More useful is contributions to force
for each monomial. This monitor is simply  norm2(ds_u)

\item (*) Bagel/QMP-2 tests for IBM Bluegene/L and NERSC SP. {\em B\'alint: BAGEL QMP2 Runs on BlueGene/L. QMP-2 is a big but fixed hit. NERSC Still to be Done}

\item (*) PARTFILE I/O in QIO, for efficient I/O on QCDOC.

Based on force monitoring tests:
  \item Fractional power monomials??? 
  \item Sexton-Weingarten integration for gauge piece ala Degrand???
  \item SAP for force term???

\item Stout-link deriv. for HMC. How do we add stout smearing for the
actions??? Is this a new  "state" ?? We can't overload on the arg types
to the factory calls.

\item Aniso precond. and even-odd aniso precond. for linops
\item Aniso precond. and even-odd aniso precond. for HMC

\item Aniso support in GaugeActs - its nearly there, only IO.

\item Make generic update "harness" that also works for pure-gauge
heatbath.

\item Make heat-bath production ready. Need over-relaxation. Need I/O !

\item Need some kind of reader/writer (??) for ILDG meta-data conventions???
BJ points out management stuff not workable. Could do physics portion.

\item Make xcheck that uses XMLDiff for complete regression tests.

\item A. Hart says the subset-ted reunit is slow - should use
QDP's profile and isolate.


DONE:

\item Change check_PROGRAMS to bin_PROGRAMS and install binaries
in chroma/mainprogs/main

\item Simple aniso for DWF - fix weirdness

\item 4D Chrono Predictor Done. (23/Feb/2005) - 5D Chrono Predictor Done (24/Feb 2004) 

\item BAGEL Running on Bluegene/L using QCDOC target and QMP_v2. IBM needs checking/porting still (potential XLC/isms). Detailed report to write.