PartitionedDatasetCollection master bug
This lists set of issues we need to address to make sure partitioned-dataset collection (PDC) i.e. the data type produced by IOSS reader is supported correctly.
Critical issues:
-
Selection support: since composite-id (CI), which we used for multiblock (MB) datasets, cannot be used with PDC in parallel as the structure is not same across ranks. We need to add support for "selectors" when defining selections. -
Multiblock inspector needs to needs to be updated to use Data Assembly and setup display properties using data-assembly paths. -
RedistributeDataSet filter load balancing (#20437 (closed)) -
RedistributeDataSet filter and global cell ids #20438 (closed) -
block name changes (#20439 (closed)) -
vtkPythonSelector and subsetting (#20488 (closed)) -
XML writers for PD/PDC don't seem to work as expected in parallel. (#20641 (closed)) -
parallel selection issue (#19395 (closed)) -
local rendering issue (#20654 (closed)) -
update chart representation to work with PDC instead of MBs (#20679) -
Plot Data doesn't support Partitioned Data Set Collections #21267 -
hover-on tooltips don't work (#20777 (closed)) -
partition balancer (#20802 (closed))
Important issues:
-
update plot selection over time to produce PDC instead of MB -
Group datasets filter (#20690 (closed)) -
Group timesteps filter (#20660 (closed))
Noncritical issues:
-
Saving a non distributed partitioned dataset segfault #20942 -
We need a new geometry filter that doesn't convert PDC to MB instead keeps it as PDC. -
We need a new mapper that replaces vtkCompositePolyDataMapper where properties are specified using data-assembly -
vtkAngularPeriodicFilter doesn't work with PDC -
ReaderAlgorithm bug: #20640
Edited by Spiros Tsalikis