(0024346)
Robert Maynard (developer)
2010-12-31 11:54

ParaView Commit SHA: c96853dcbcc

After talking with Utkarsh we decided that with how rarely we change default values it seemed reasonable that the python state would not include them.

(0024361)
Pat Marion (reporter)
2011-01-03 11:34

Commit c96853dcbcc needs to be reverted. For saving python state, the CaptureAllProperties flag must be True.

I think that solving this bug requires writing a routine that checks if a property's current value matches the default_value defined in xml.

(0024368)
Robert Maynard (developer)
2011-01-03 16:29

I have reverted commit c96853dcbcc.

(0031138)
Andy Bauer (developer)
2013-07-12 17:22

Note that Kyle fixed this already for saving state. The option is available under Save Full State in the Settings menu. I'm working on enabling this feature for the generated co-processing scripts.

(0031142)
Andy Bauer (developer)
2013-07-16 23:00

To control whether or not default values are included in the catalyst and spatio-temporal generated scripts, set the option in the settings menu. The default value in the settings menu is to cull the default values from the generated script.

(0031336)
Utkarsh Ayachit (administrator)
2013-07-26 10:00

SUMMARY
---------------------------------------------
Topics merged into master:
        10965_cull_default_values
        14179-add-PythonQt-plugin
        cth_markers_on_by_default
(VTK) fix-catalyst
(VTK) update_vtk_master

(0032021)
Andy Bauer (developer)
2013-12-18 11:20

Steps to reproduce the current incorrect behaviour:
1) start trace
2) create wavelet, apply
3) create cut filter, apply
4) stop trace
5) save file
now repeat all steps with save full state in python switched from the previous setting.

(0032032)
Utkarsh Ayachit (administrator)
2013-12-19 16:56

I think there's lot of confusion with this bug.

Andy/Alan, "Trace" is never supposed to change based on whether "Save Full State in Python" setting is checked or not. "Trace" is always a log of the the changed properties, so it's only going to record the values that changed.

The only thing this setting (the "Save Full State in Python" setting from the Settings dialog) affects is when one tries to "Save State" as Python (accessible from File->Save State menu; don't forget to choose the "*.py" file extension). With that you'll indeed get the "full" or "trimmed" states files as expected.

(0032035)
Utkarsh Ayachit (administrator)
2013-12-20 08:26

Topics merged into master:
        14286_fix_orientation_widget
(VTK) 14426-huge-png
        14464_fix_specular_highlights
        14466_fix_calculator_panel
        disable-unified-bindings-by-default
        fix-collaboration
        fix_parallel_amrconnectivity
        gio-cleanup-and-warning-fixes
        pvweb-enhancements-events-protocols-docs

(0032056)
Alan Scott (manager)
2013-12-24 17:45

Fascinating bug. As per Utkarsh's last note, this isn't a bug at all, since Kens' writeup is confusing. So, I am not sure what to test.

With regards to Trace State not culling default values (summary of this bug), Utkarsh's latest work has greatly reduced the amount of default state placed in the trace. Very nice. There are issues with the new trace. (Basically, can_ex2 = FindSource( "can.ex2" )is being incorrectly shoved into the trace.) I am dealing with it in bug number 14214. This is basically fixed.

With regards to Save Full State as Python doesn't effect trace, this is correct. This is not a bug.

With regards to saving state as python (description of this bug), and playing with the Save Full State as Python setting, this now works correctly. The non Save Full State is greatly reduced.

Tested master, local server or pvbatch, Linux.

Issue History
Date Modified	Username	Field	Change
2010-07-08 16:11	Ken Moreland	New Issue
2010-09-01 11:34	Utkarsh Ayachit	Target Version	3.10 => 3.10.shortlist
2010-12-31 10:11	Robert Maynard	Assigned To	=> Robert Maynard
2010-12-31 10:11	Robert Maynard	Status	backlog => tabled
2010-12-31 11:54	Robert Maynard	Note Added: 0024346
2010-12-31 11:54	Robert Maynard	Status	tabled => @80@
2010-12-31 11:54	Robert Maynard	Fixed in Version	=> 3.10.shortlist
2010-12-31 11:54	Robert Maynard	Resolution	open => fixed
2011-01-03 11:34	Pat Marion	Note Added: 0024361
2011-01-03 11:34	Pat Marion	Status	@80@ => @20@
2011-01-03 11:34	Pat Marion	Resolution	fixed => reopened
2011-01-03 14:20	Pat Marion	Status	@20@ => tabled
2011-01-03 16:29	Robert Maynard	Note Added: 0024368
2011-06-16 13:10	Zack Galbreath	Category	Feature Request => Feature
2013-01-08 21:33	Alan Scott	Project	=> Sandia
2013-01-08 21:33	Alan Scott	Type	=> incorrect functionality
2013-03-18 15:09	Utkarsh Ayachit	Relationship added	related to 0013891
2013-03-18 15:09	Utkarsh Ayachit	Project	Sandia => Sandia-Catalyst
2013-03-18 15:09	Utkarsh Ayachit	Assigned To	Robert Maynard =>
2013-07-12 17:20	Andy Bauer	Assigned To	=> Andy Bauer
2013-07-12 17:21	Andy Bauer	Status	backlog => todo
2013-07-12 17:22	Andy Bauer	Note Added: 0031138
2013-07-12 17:22	Andy Bauer	Status	todo => active development
2013-07-16 23:00	Andy Bauer	Topic Name	=> 10965_cull_default_values
2013-07-16 23:00	Andy Bauer	Note Added: 0031142
2013-07-16 23:00	Andy Bauer	Status	active development => gatekeeper review
2013-07-16 23:00	Andy Bauer	Fixed in Version	=> git-next
2013-07-16 23:00	Andy Bauer	Resolution	reopened => fixed
2013-07-26 10:00	Utkarsh Ayachit	Fixed in Version	git-next => git-master
2013-07-26 10:00	Utkarsh Ayachit	Status	gatekeeper review => customer review
2013-07-26 10:00	Utkarsh Ayachit	Note Added: 0031336
2013-11-01 13:18	Utkarsh Ayachit	Fixed in Version	git-master => 4.1
2013-12-18 11:17	Andy Bauer	Assigned To	Andy Bauer => Utkarsh Ayachit
2013-12-18 11:17	Andy Bauer	Status	customer review => todo
2013-12-18 11:20	Andy Bauer	Note Added: 0032021
2013-12-19 16:51	Utkarsh Ayachit	Description Updated
2013-12-19 16:56	Utkarsh Ayachit	Note Added: 0032032
2013-12-19 16:56	Utkarsh Ayachit	Status	todo => gatekeeper review
2013-12-20 08:26	Utkarsh Ayachit	Fixed in Version	4.1 => git-master
2013-12-20 08:26	Utkarsh Ayachit	Status	gatekeeper review => customer review
2013-12-20 08:26	Utkarsh Ayachit	Note Added: 0032035
2013-12-24 17:45	Alan Scott	Note Added: 0032056
2013-12-24 17:45	Alan Scott	Status	customer review => closed
2014-01-11 08:39	Utkarsh Ayachit	Fixed in Version	git-master => 4.1

Notes
(0024346) Robert Maynard (developer) 2010-12-31 11:54	ParaView Commit SHA: c96853dcbcc After talking with Utkarsh we decided that with how rarely we change default values it seemed reasonable that the python state would not include them.

(0024361) Pat Marion (reporter) 2011-01-03 11:34	Commit c96853dcbcc needs to be reverted. For saving python state, the CaptureAllProperties flag must be True. I think that solving this bug requires writing a routine that checks if a property's current value matches the default_value defined in xml.

(0024368) Robert Maynard (developer) 2011-01-03 16:29	I have reverted commit c96853dcbcc.

(0031138) Andy Bauer (developer) 2013-07-12 17:22	Note that Kyle fixed this already for saving state. The option is available under Save Full State in the Settings menu. I'm working on enabling this feature for the generated co-processing scripts.

(0031142) Andy Bauer (developer) 2013-07-16 23:00	To control whether or not default values are included in the catalyst and spatio-temporal generated scripts, set the option in the settings menu. The default value in the settings menu is to cull the default values from the generated script.

(0031336) Utkarsh Ayachit (administrator) 2013-07-26 10:00	SUMMARY --------------------------------------------- Topics merged into master: 10965_cull_default_values 14179-add-PythonQt-plugin cth_markers_on_by_default (VTK) fix-catalyst (VTK) update_vtk_master

(0032021) Andy Bauer (developer) 2013-12-18 11:20	Steps to reproduce the current incorrect behaviour: 1) start trace 2) create wavelet, apply 3) create cut filter, apply 4) stop trace 5) save file now repeat all steps with save full state in python switched from the previous setting.

(0032032) Utkarsh Ayachit (administrator) 2013-12-19 16:56	I think there's lot of confusion with this bug. Andy/Alan, "Trace" is never supposed to change based on whether "Save Full State in Python" setting is checked or not. "Trace" is always a log of the the changed properties, so it's only going to record the values that changed. The only thing this setting (the "Save Full State in Python" setting from the Settings dialog) affects is when one tries to "Save State" as Python (accessible from File->Save State menu; don't forget to choose the "*.py" file extension). With that you'll indeed get the "full" or "trimmed" states files as expected.

(0032035) Utkarsh Ayachit (administrator) 2013-12-20 08:26	Topics merged into master: 14286_fix_orientation_widget (VTK) 14426-huge-png 14464_fix_specular_highlights 14466_fix_calculator_panel disable-unified-bindings-by-default fix-collaboration fix_parallel_amrconnectivity gio-cleanup-and-warning-fixes pvweb-enhancements-events-protocols-docs

(0032056) Alan Scott (manager) 2013-12-24 17:45	Fascinating bug. As per Utkarsh's last note, this isn't a bug at all, since Kens' writeup is confusing. So, I am not sure what to test. With regards to Trace State not culling default values (summary of this bug), Utkarsh's latest work has greatly reduced the amount of default state placed in the trace. Very nice. There are issues with the new trace. (Basically, can_ex2 = FindSource( "can.ex2" )is being incorrectly shoved into the trace.) I am dealing with it in bug number 14214. This is basically fixed. With regards to Save Full State as Python doesn't effect trace, this is correct. This is not a bug. With regards to saving state as python (description of this bug), and playing with the Save Full State as Python setting, this now works correctly. The non Save Full State is greatly reduced. Tested master, local server or pvbatch, Linux.