[MOAB-dev] moab file format and solution in time

[Jed Brown]

Tim Tautges wrote:
> One other question - does anybody know how various file formats handle
> adapting mesh?  I know some formats use masking or element death
> markers, while others just write a whole new series of files when the
> mesh changes.  Does anybody try to write deltas to an initial or final
> mesh topology?

I'm not aware of such, though I'm sure someone has done a crappy job of
it and written a paper about how it's really essential.  If playing with
the mesh was an end in and of itself, then it would make sense to do
lightweight diffs.  This would make sense if you had a big parallel mesh
and were doing some quality improvement where you wanted to save every
update.  But *why* would you want to save it all?  To make a movie?  The
only realistic application I can think of is front-tracking with AMR,
where the physics is simple (few fields) and low-order discretization is
used.  Is this an audience you want to optimize for?  Otherwise don't
bother.


Jed

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 261 bytes
Desc: OpenPGP digital signature
URL: <http://lists.mcs.anl.gov/pipermail/moab-dev/attachments/20091119/53156908/attachment.pgp>