Time Line edit mode

[IanMantula]

I would like to see a way to edit the time line that provides more feedback about relationships and dependencies. I recently tried to re-organise the time-line for a particularly complex design. While I was able to achieve most of what I wanted to do, some of it was just too much trouble. 

I would like see:

• The name of each step along with a choice of relevant properties. 
• Feedback on why a step can't be moved beyond a specific point.
• Visual display of step dependencies.
• Visual display of error info for broken dependencies.

All of this would need to interactive so that changes are applied directly to the design.

[roambotics_scott]

Yeah - the Timeline should so at least expand into a dependency graph view (which is, IMO, what it really should be in the first place rather than trying to arbitrarily shoehorn it into one dimension).

[keqingsong]

Addressing your bullets:

• The name of each step along with a choice of relevant properties. 
  When you hover over a feature in the timeline, it tells you what that feature is and where it was applied to. You can edit features in the timeline by right-clicking on them and selecting edit feature. 
  
  
  
• Feedback on why a step can't be moved beyond a specific point.
  This is something we can do better. The timeline is a chronological history of your actions. It is not a browser tree where you can rearrange features and place them where you want them to occur. Sometimes you cannot move a feature to a future time because it can not occur at that time due to downstream dependencies. 
  
  
• Visual display of step dependencies.
• Visual display of error info for broken dependencies.
  In this model for example, if you drag the fillet feature after the shell, that'll affect the sketch since the sketch is based on what the model looked like before the change. The sketch is then highlighted in yellow, letting you know that a dependency is broken. 
  

You can learn about how you fix unhealthy features as well as timeline/performance best practices here: https://www.youtube.com/playlist?list=PLmA_xUT-8UlIh4hHJDCEDLPi3wQiMrPRY 

Does this help?

[IanMantula]

@keqingsong

Thanks for the feedback. 

To answer your comments on my bullets...

The name of each step along with a choice of relevant properties.

Yes I am aware of the ability to hover over a timeline feature, it's very useful in normal use. However it's not always visible, so if you have a timeline of 100 or more steps and you want to drag a feature from a scrolled part of the timeline it can be very difficult to place it exactly where you want. Dragging over each item and waiting for the tooltip to show is quite slow.

Perhaps it would be possible to have something like a bookmark or flag that you can use to temporarily highlight a feature so that it's easy to find?

Feedback on why a step can't be moved beyond a specific point

Yes it sometimes is difficult to understand why certain steps have to happen after others when there's no obvious relationship. The reason may be perfectly obvious to the internal wokings of F360, so some guidance would be nice!

Visual display of step dependencies and error info for broken dependencies

Again I'm familiar with the current behaviour, so this is what could work...

When you start to drag a timeline feature it highlights in a new colour all of the preceding features that the feature is dependent on if dragging backward. When dragging forward it highlights all of the features that are dependant on it. Of course I have no real idea if this is possible, but it could be achieved without a massive change to the basic UI.

The big question of course is should I be rearranging my time line to the extent that I have attempted? It was't for frivolous reasons, just that the design was more complex than expected and to simplify things I had to try a different timeline order. Now that I have nearly achieved that, I'm glad that I did it.

Ian

[kb9ydn]

For complex history based modelling (what we get in parametric mode vs. direct) I think a feature dependency graph would be far more useful than a chronological history line (most of the time at least).  The problem is that a dependency graph will be much larger and take up a lot of screen space whereas a single history line is just that, a single line.  So there are UI issues to deal with.

The other issue is that Fusion enforces the history order at the assembly level as well as the part level, and the timeline encompasses more than just features.  If it only included features a dependency graph would be easy, but it also includes commands like component moves and adding joints.  How these things would fit into a dependency graph is not clear to me.  Not that it couldn't be done, but I get the feeling this is sort of new territory in the CAD world.

As far as massive reorganization of a model, sometimes it's unavoidable.  I can say that with experience (in SWX) I've needed to do this less and less, sometimes things just evolve in ways you couldn't predict.

C|

[IanMantula]

OK, so I have commented as requested, what happens next?

Ian

[promm]

Thank you for your idea - this is getting archived due to lack of votes.  We do not have any plans to change our timeline behavior and it would take our community getting behind this idea for us to consider changes.

Cheers,

Mike Prom