[petsc-dev] XDMF viewers

[Jed Brown]

On Sat, Dec 3, 2011 at 14:31, Tim Gallagher <tim.gallagher at gatech.edu>wrote:

> I was looking for the 'PETSc' way of doing it, something that lets you
> call VecView(vec, viewer) that generates the heavy data and XML file
> together. If there is not interest in that approach, I can just write the
> routines to make the XML in the code.


I don't know if Matt has that in PyLith, or if they use a less composable
interface.

Binary-appended VTK does work this way. You can write several vectors to a
file and they are held in memory until you close the file. The
implementation is memory-scalable in the sense that there are no gathers of
global state onto one process.

Unfortunately, you cannot write many time steps to the same file, so you
should usually create a separate file per step. (Lots of vis software likes
this, so it's not a terrible restriction, although I don't like it.)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20111203/ad521af2/attachment.html>