(0009020)
Berk Geveci (administrator)
2007-09-17 11:27

Whether this is easy or not depends on the reader. For example, is there a way to make the Exodus reader to read all of the information again? I am assigning to Eric so that he can answer that question. If that works, we can probably add a button to the Exodus panel that resets the information. We will not extend this to all readers unless there is a REAL good reason to do so.

(0009045)
David Thompson (old) (reporter)
2007-09-18 02:20

This should be relatively simple for the reader... none of the truth tables/variable names change over time. I see two issues to resolve:

1. How does the rest of the ParaView pipeline "adjust" to the new time steps? If the number of time steps changes, any animation frames are off, the geometry/image caches must be erased, etc.

2. When using time sequence as mode shapes, it won't be the time steps that change (the ExodusII reader advertises a TIME_RANGE of [0,1] when mode shapes are on) but an "animatable parameter" that changes. Again, how will the rest of the ParaView pipeline accommodate the change?

(0009087)
Eric Stanton (reporter)
2007-09-18 13:53

Berk, did you want me to implement this since it looks like the exodus reader can handle it? Or was someone else going to? Doesn't matter to me.

(0009088)
Berk Geveci (administrator)
2007-09-18 13:56

Eric, feel free to implement this (if you have time). Please consult with Utkarsh on how to proceed.

(0009089)
Berk Geveci (administrator)
2007-09-18 13:56

Reminder sent to: Utkarsh Ayachit

Utkarsh, Eric will do this. Can you help him if necessary?

(0009178)
Alan Scott (manager)
2007-09-19 21:14

I was just talking to a user, and it sounds like this is a very, very popular feature. Any idea if this will be done for 3.2 or 3.4?

(0009179)
Eric Stanton (reporter)
2007-09-19 21:22

That depends on when we branch. But I think with a push we can get it in to 3.2.

(0009180)
Eric Stanton (reporter)
2007-09-19 21:34

One challenge is going to be how to test it...

(0009181)
Alan Scott (manager)
2007-09-19 21:46

Obviously, don't do anything dumb that hurts the release. But, it being in 3.2 would be a BIG thing.

As far as testing it, I have a user here that is creating an alegra run that will be very simple, not OUO, a few files output, write timesteps constantly for an hour or two, and that I can run. Then, I could let you try to read it as it is writing. Would that work?

(0009182)
Eric Stanton (reporter)
2007-09-19 21:47

Yea, that would be great. I'll let you know when its ready to be tested.

(0009385)
Eric Stanton (reporter)
2007-10-04 14:04

I've committed the fix to the trunk, and a patch for the 3.2 branch will be applied once next year's contract goes through.

(0009499)
Alan Scott (manager)
2007-10-17 16:02

I am pushing this one back until it is at least in the 3.2 branch.

alan

(0009562)
Ken Moreland (manager)
2007-10-25 13:08

This should be in the 3.2 branch now.

(0009611)
Alan Scott (manager)
2007-11-01 20:14

Nice!

Notes
(0009020) Berk Geveci (administrator) 2007-09-17 11:27	Whether this is easy or not depends on the reader. For example, is there a way to make the Exodus reader to read all of the information again? I am assigning to Eric so that he can answer that question. If that works, we can probably add a button to the Exodus panel that resets the information. We will not extend this to all readers unless there is a REAL good reason to do so.

(0009045) David Thompson (old) (reporter) 2007-09-18 02:20	This should be relatively simple for the reader... none of the truth tables/variable names change over time. I see two issues to resolve: 1. How does the rest of the ParaView pipeline "adjust" to the new time steps? If the number of time steps changes, any animation frames are off, the geometry/image caches must be erased, etc. 2. When using time sequence as mode shapes, it won't be the time steps that change (the ExodusII reader advertises a TIME_RANGE of [0,1] when mode shapes are on) but an "animatable parameter" that changes. Again, how will the rest of the ParaView pipeline accommodate the change?

(0009087) Eric Stanton (reporter) 2007-09-18 13:53	Berk, did you want me to implement this since it looks like the exodus reader can handle it? Or was someone else going to? Doesn't matter to me.

(0009088) Berk Geveci (administrator) 2007-09-18 13:56	Eric, feel free to implement this (if you have time). Please consult with Utkarsh on how to proceed.

(0009089) Berk Geveci (administrator) 2007-09-18 13:56	Reminder sent to: Utkarsh Ayachit Utkarsh, Eric will do this. Can you help him if necessary?

(0009178) Alan Scott (manager) 2007-09-19 21:14	I was just talking to a user, and it sounds like this is a very, very popular feature. Any idea if this will be done for 3.2 or 3.4?

(0009179) Eric Stanton (reporter) 2007-09-19 21:22	That depends on when we branch. But I think with a push we can get it in to 3.2.

(0009180) Eric Stanton (reporter) 2007-09-19 21:34	One challenge is going to be how to test it...

(0009181) Alan Scott (manager) 2007-09-19 21:46	Obviously, don't do anything dumb that hurts the release. But, it being in 3.2 would be a BIG thing. As far as testing it, I have a user here that is creating an alegra run that will be very simple, not OUO, a few files output, write timesteps constantly for an hour or two, and that I can run. Then, I could let you try to read it as it is writing. Would that work?

(0009182) Eric Stanton (reporter) 2007-09-19 21:47	Yea, that would be great. I'll let you know when its ready to be tested.

(0009385) Eric Stanton (reporter) 2007-10-04 14:04	I've committed the fix to the trunk, and a patch for the 3.2 branch will be applied once next year's contract goes through.

(0009499) Alan Scott (manager) 2007-10-17 16:02	I am pushing this one back until it is at least in the 3.2 branch. alan

(0009562) Ken Moreland (manager) 2007-10-25 13:08	This should be in the 3.2 branch now.

(0009611) Alan Scott (manager) 2007-11-01 20:14	Nice!

Issue History
Date Modified	Username	Field	Change
2007-09-11 20:42	Alan Scott	New Issue
2007-09-17 11:27	Berk Geveci	Note Added: 0009020
2007-09-17 11:27	Berk Geveci	Status	backlog => tabled
2007-09-17 11:27	Berk Geveci	Assigned To	=> Eric Stanton
2007-09-18 02:20	David Thompson (old)	Note Added: 0009045
2007-09-18 13:53	Eric Stanton	Note Added: 0009087
2007-09-18 13:56	Berk Geveci	Note Added: 0009088
2007-09-18 13:56	Berk Geveci	Note Added: 0009089
2007-09-19 21:14	Alan Scott	Note Added: 0009178
2007-09-19 21:22	Eric Stanton	Note Added: 0009179
2007-09-19 21:34	Eric Stanton	Note Added: 0009180
2007-09-19 21:46	Alan Scott	Note Added: 0009181
2007-09-19 21:47	Eric Stanton	Note Added: 0009182
2007-10-04 14:04	Eric Stanton	Status	tabled => @80@
2007-10-04 14:04	Eric Stanton	Resolution	open => fixed
2007-10-04 14:04	Eric Stanton	Note Added: 0009385
2007-10-17 16:02	Alan Scott	Status	@80@ => @20@
2007-10-17 16:02	Alan Scott	Resolution	fixed => reopened
2007-10-17 16:02	Alan Scott	Note Added: 0009499
2007-10-25 13:08	Ken Moreland	Status	@20@ => @80@
2007-10-25 13:08	Ken Moreland	Resolution	reopened => fixed
2007-10-25 13:08	Ken Moreland	Note Added: 0009562
2007-11-01 20:14	Alan Scott	Status	@80@ => closed
2007-11-01 20:14	Alan Scott	Note Added: 0009611
2011-06-16 13:10	Zack Galbreath	Category	=> (No Category)

Relationships