[Paraview] Misleading facts about Catalyst

[Kolja Petersen]

Heya,
I'd need some more reliable input on the Catalyst design.

Last week I asked how the server side can notify the client of recent
changes. Utkarsh wrote this wouldn't be supported by Paraview. After I
clarified that exactly this is happening during a Catalyst Live simulation,
the discussion was mute. I am still hoping for an answer how to notify the
client of any changes that might happen on the server side for example as a
result of some network activity.

Another example, some days before, I was wondering why Catalyst datasets of
considerable size take so long from the Insitu side to the Live
Visualization side, given that the Catalyst User Guide gives the impression
that everything is designed for high throughput. Someone insisted that
everything is transported through MPI until I mentioned that Paraview is
not designed for a dynamically added MPI communicator. Only then it was
explained that Catalyst data exchange is going through TCP...

The Catalyst development seems increasingly problematic. Have the Catalyst
developers left Kitware, or what is going on? We are trying to work on
improved solutions, but with the recent quality of comments we can't see
this happening.
Thank you
Kolja


<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
Virenfrei.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20171127/7fa5da4b/attachment.html>