[If this is a second posting, I apologize; I drafted (and thought I
sent) this reply already, but now can find no sign of it]
On Aug 4, 2008, at 4:01 AM, Igor Telezhinsky wrote:
I use fipy revision 2718.
Actually, I don't believe you are. This is a bad design of SVN, in my
opinion (although I kind of understand why it works that way). `svn
update` reports the latest revision in the entire repository, not the
latest revision in your working copy:
guyer% svn up
At revision 2718.
guyer% svn info
Path: .
URL: svn+ssh://[EMAIL PROTECTED]/usr/local/svn-fipy-repos/trunk
Repository Root: svn+ssh://[EMAIL PROTECTED]/usr/local/svn-fipy-repos
Repository UUID: d80e17d7-ff13-0410-a124-85740d801063
Revision: 2718
Node Kind: directory
Schedule: normal
Last Changed Author: wd15
Last Changed Rev: 2709
Last Changed Date: 2008-07-28 10:01:05 -0400 (Mon, 28 Jul 2008)
so, although the repository has revisions up to 2718, the latest
revision in the trunk/ is only 2709. Since, in your other message, you
said you were running STABLE
guyer% svn info http://matforge.org/svn/fipy/tags/STABLE
Path: STABLE
URL: http://matforge.org/svn/fipy/tags/STABLE
Repository Root: http://matforge.org/svn/fipy
Repository UUID: d80e17d7-ff13-0410-a124-85740d801063
Revision: 2718
Node Kind: directory
Last Changed Author: guyer
Last Changed Rev: 2473
Last Changed Date: 2008-02-08 14:26:37 -0500 (Fri, 08 Feb 2008)
I would guess that you're actually running FiPy 1.2, at svn revision
2473. This would be consistent with the syntax changes you had to make.
By the way, in the solution there are values of order 1e-322
positive
as well as negative ones. I guess I can count them as 0, can I?
I would certainly think so. If, however, your entire solution vector
is of this order of magnitude, I would again question the scaling of
your problem.
