RES: [Paraview] Readers compatibility list

rnelias rnelias at nacad.ufrj.br
Mon Oct 23 12:34:28 EDT 2006


Hi Berk, thanks for you reply.

I've been following the great improvements that the PV team has been doing
with the software and, I guess, that all PV users are concious and recognize
that PV is a great software in the development phase and that it has bugs
(as any other software) and that new and old features have been, massively,
developed everyday. My idea with the "compatibility list" would be a place
to gather information about PV readers and their compatibility according to
the PV version. It could be started as a table in the PV wiki's page where
anyone could make a compatibility checkmark in this table. Something like:


PV version 2.x

             VTK   PLOT3D   EXODUS   ENSIGHT
Filter X:     X      X        X 
Filter Y:            X        X         X
Filter Z:     X      X

In the future, new PV users that are looking for the file format which best
suit their needs could use this table. Furthermore, according to the PV
evolution and the problems worked out, this table would be changed
accordingly.

I know that Ensight's readers have been reimplemented in the newer PV
versions and that some features are not available yet, for example: I'm able
to use volume rendering when I load my data in parallel using the Ensight
SOS format, but I can't produce animations with such datasets (which are
simple)since it's not supported yet (we could put it in a table to future
users -- besides the bug track). 

I only know such issues related to Ensight's format, firstly because I'm
still using Ensight's formats and secondly because I read the maillist
frequently and sometimes post my own problems here. Imagine, for example, a
new PV user willing to use the Ensight's format (or any other format), this
person would face the same problems than I (and others) without having the
chance to know about them in advance.

Anyway, we could try to start something like a "compatibility list" on some
wiki's page as Randall did.

Thanks a lot for your feedback

Renato.


-----Mensagem original-----
De: Berk Geveci [mailto:berk.geveci at kitware.com] 
Enviada em: segunda-feira, 23 de outubro de 2006 12:35
Para: Renato N. Elias
Cc: paraview at paraview.org
Assunto: Re: [Paraview] Readers compatibility list

Hi Renato,

I don't have a compatibility list.

Some of the issues you have been running into is due to the transition to
the new multi-block code in VTK. Although many of the filters worked out of
the box after the move to the new code, some require changes to handle the
new data structures. The readers that show these problems are:
* Ensight
* Plot3D
* XML multiblock/AMR

We are working on improving the multi-block support. Meanwhile, I will
introduce a filter that can convert a multi-block dataset to a simple
dataset by extracting one block. Using that filter, you will be able to
extract one block and apply all available filters as before. This is a
temporary solution. As the multi-block support grows, more and more filters
will become available. There are more filters avaible for multiblock
datasets in 2.6. I will try to generate a list of those that do not work
before the release.

As for lookmarks, we are rewriting them completely in paraview 3. They will
work for all readers. If you have issues with the lookmark implementation in
paraview 2, post a bug report (or point me to one if it exists) and I will
see if it can be fixed before 2.6.

As for volume rendering, you will be able to volume render image and
unstructured data  as long as they are "simple" datasets. If you have a
multi-block dataset produced by one of the readers above, you will have to
apply a filter to convert it to a simple dataset.

As for parallel reading, the readers that I know support parallel reading
are:
* Parallel XML readers
* EnSight SOS reader
* XDMF reader (structured and partitioned unstructured)
* Exodus reader
* Spyplot reader

There may be others. If a reader does not support parallel reading, paraview
will take care of redistributing data over MPI. This behaviour is a little
different depending on the data type. For polydata and unstructured grids,
the first node will read all the data and redistribute using MPI. For
structured datasets (image, rectilinear and structured grids), all processes
will read all the data and then crop it.

I hope this helps. If you are aware of any bugs in any of these, let me
know.

-berk

On 10/21/06, Renato N. Elias <rnelias at nacad.ufrj.br> wrote:
>
> Hey guys,
>
> Is there a ParaView compatibility list according to the reader used?
>
> For example:
>
> VTK Format: all PV filters supported, parallel reading, lookmarks
> Exodus: filter X, Y, Z, volume rendering, parallel reading, etc...
> Plot3D: filter W, K, L (partially supported)
> Ensight: ...
> ...
> ...
>
> It would be helpful when deciding which format should be adopted to 
> use with ParaView.
>
> maybe someone in PV team already have something like that.
>
> Regards
>
> Renato N. Elias
> ===============================================
> PhD student - http://www.nacad.ufrj.br/~rnelias High Performance 
> Computing Center Federal University of Rio de Janeiro Rio de Janeiro, 
> Brazil
> +55(21) 2562-8080
>
> _______________________________________________
> ParaView mailing list
> ParaView at paraview.org
> http://www.paraview.org/mailman/listinfo/paraview
>


--
 Berk Geveci
 Kitware Inc.
 28 Corporate Drive
 Clifton Park, NY, 12309



More information about the ParaView mailing list