[Carpet] Error in restriction correction corrected

[Erik Schnetter]

On February 26, I reported that I had corrected an error in Carpet's  
restriction operators.  This correction contained unfortunately an  
error, which I have now corrected.  I have also convinced myself that  
this error is not actually an error, and I have switched the default  
behaviour back to what it was before February 26.

To be specific, the question is whether restriction (copying fine to  
coarse data) should fill in grid points which are used for  
prolongation (interpolating coarse to fine data).  If this is done,  
then it is necessary to prolongate after restricting to keep the data  
consistent.  If this is not done, then there is a small region where  
both coarse and fine data exist, but may be different.

There is now a parameter  
"CarpetLib::omit_prolongation_points_when_restricting", which  
defaults to "no".

-erik

-- 
Erik Schnetter <schnetter at cct.lsu.edu>

My email is as private as my paper mail.  I therefore support encrypting
and signing email messages.  Get my PGP key from www.keyserver.net.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : /archives/developers/attachments/20060505/7e03c52f/attachment.pgp