Timeline bookmarks

[SGL-Design]

There have been several requests for this feature and/or search ability in the timeline but I'm not sure if I should create a new idea or upvote previous ones that have been archived.   

Here are links to older ones

https://forums.autodesk.com/t5/fusion-360-ideastation-request-a/timeline-bookmarks/idi-p/5821819

https://forums.autodesk.com/t5/fusion-360-ideastation-request-a/timeline-bookmarks-and-search/idi-p/...

https://forums.autodesk.com/t5/fusion-360-ideastation-request-a/timeline-markups/idi-p/7020545

https://forums.autodesk.com/t5/fusion-360-ideastation-request-a/find-in-timeline/idi-p/6454309

The title says it all, it would be great to have bookmarks with comments associated with the timeline.  I know using components can assist here, but even them some components can have large timeline histories and can be difficult to manage. In my application, I frequently go near the beginning of the timeline to upload more attached canvases.  going back and forth can be a hassle.

Thanks

[carloquinonez]

I get why Autodesk prioritizes features that are highly requested by the community, but that ends up screwing requests like this one (and all the rest of the ones you linked to). The reality is most parts designed in CAD are really simple and requests that only benefit a small fraction of the community are going to be ignored in favor of features requested by a large number of users. When I suggested this feature nearly 2 years ago, I didn't even get the perfunctory "Thanks for the idea - this is getting archived due to lack of votes..." comment from Autodesk.

Keeping my fingers crossed they listen this time. Good luck!

[SGL-Design]

You are absolutely right! I went to up vote all of the other related ones, but it certainly does mean that refining features like this are never very likely to occur. 

I do see the challenge that the Fusion Team has with these less common requests that may lack the 10 vote threshold on each individual request, even though the idea as a whole may have 30 votes across the multiple requests.   I'm not sure how I would prioritize these, but I when I face similar issues in my company I try to commit a portion of our development resources to refinement, say 15% of available time. This ensures that we are actually moving toward refining the details of a product rather than sitting at "good enough for most".   Of course generally you get a much lower return when working on that level of refinement. .

[Pedro_Bidarra]

The last one you posted is still active vote for that last one.

https://forums.autodesk.com/t5/fusion-360-ideastation-request-a/find-in-timeline/idi-p/6454309

[SGL-Design]

Thanks for pointing that out, up voted and telling others!

[carloquinonez]

The problem of complex timelines is especially acute when you're trying to leverage the design freedom of 3d-printing - complex geometries usually mean complex timelines. As engineers become more fluent with design for additive manufacturing (DFAM), they're able to aggregate more functionality into a single part, which inevitably leads to timelines becoming longer and longer. Eventually the cognitive overhead and computational costs of long timelines creates an artificial barrier to complexity.

What makes this problem worse is that bookmarking timelines is just a superficial solution. Ultimately, you have to start breaking up the timeline not just components, but across separate designs. But now we're talking about changing the standard CAD paradigm of "1 part = 1 file"...

If we can't even get traction on bandaid solutions to complexity (like a bookmark in the timeline), what hope is there for "real" solutions? This realization has me so discouraged that I don't bother reporting many of the other workarounds or ideas I have for handling highly complex models.

[kb9ydn]

I would rather they DON'T implement band-aid solutions for anything.  The problem with temporary solutions is that the complaining dies down long enough for people to forget about the problem, and eventually the temporary solution becomes permanent.

The issue with timeline complexity is a huge one though, that will have to be dealt with if Fusion is to ever become a serious contended in the CAD world.

C|

[SGL-Design]

I had never considered how 3d printing tends to greatly increase the potential complexity of any given part. Since 3d printing permits a single part to do the work of multiple separate components with traditional manufacturing and design techniques.  I have on occasion used fusions direct editing system to quickly make changes to a given part to make things fit, and while the time is greatly reduced, I feel that it essentially pollutes the model with features and changes that are difficult to replicate from scratch and difficult to understand when reviewing the model weeks later.   I am very much used to a standard, feature based approach with simpler models. I sometimes wonder if I am holding onto that paradigm needlessly and should almost think of it as doing whatever is required in the model with all available tools (  direct editing with polluted timeline history) until getting to a reachable point, and being satisfied with the outcome if it works correctly.  

For example, lets say I add a filet to a part, and days later I want to change the fillet radius. Coming from other parametric modelers, I would naturally attempt to find the fillet feature and edit it with a new value. which in fusion means using the timeline.  Alternatively, with Fusions direct editing capability I could simply select the fillet on the model with push/pull and move it to the new location. (Fillet may be a bad example, because it seems to actually change the original feature).    So maybe fusion is already operating in the more ideal paradigm, but I need to adapt my way of thinking about it. 

[SGL-Design]

kb9ydn, Any ideas on how they could fix the timeline issues?   I generally feel that we are more likely to get slow incremental improvement over getting a total redesign released in one go.   I actually have grown quite accustom to the timeline concept, and feel it really needs basic usability improvements. 

[carloquinonez]

Honestly, the timeline isn't enough to navigate the geometry specs for really complex parts. This isn't unique to Fusion360. The timeline in Fusion360 is basically the same thing as the Feature Tree in Solidworks or the Specification Tree in CATIA. What Fusion360 does differently is show the structure/organization of the design (the Browser) separately from the steps to produce the geometry (the Timeline). Most other CAD software basically lumps those together.

What's really interesting about CATIA is all the options is has for displaying this information. The default setting looks a lot like Solidworks or other CAD software, but it has a ton of options. Some options are visual - horizontal/vertical, text size, truncation behavior of long names, how inactive parts are displayed, etc. But there's an option to use completely different tree types - "Standard", "Structure", "Relational" and "Constructive Historic". You can see what I'm talking about here.

At the time I was using CATIA, I wasn't experienced enough to appreciate how important the tree could be, so I never really test-drove any of the different tree types, so I can't say if those types help with complex designs.

I deal with complex trees by breaking up designs into multiple designs and including links. This keeps the histories short and easy to modify, but it's incredibly clunky and I don't recommend it - if you're not dilligent about it or you structure the design poorly, you'll make your life a lot harder without any benefit. In fact, I'll keep all my work in a single file the first time I'm working on a new design to get a sense of how to structure it. Once I think I have a handle on how to breakup the design, I'll start over. Of course, this leads to other complications in the workflow because the rest of the CAD world assumes "1 part = 1 file", etc. For really tough problems like this, bandaid solutions are all we can hope for in the short term. So +1 to bookmarks!

You know the old joke about asking a farmer from the 1800's what he needs? The answer would probably be a stronger horse, faster horse, etc. They're not going to say "tractor". Right now, Autodesk is asking the engineers that grew up with conventional manufacturing, that implicitly accept "1 part = 1 file" and had KISS drilled into them from the first day of engineering class - "What do you need?"

I've come to conclude that CAD will only change once all of the current batch of engineers retire. I'm only half joking here. Engineers had a hard enough time accepting/understanding there's no such thing as an "assembly" file in Fusion - just imagine how a completely new way of designing parts would go over...

[kb9ydn]

I think @carloquinonez is pretty much spot on here.  The issue Fusion has with complex designs is that it doesn't allow you to easily break them down into smaller more manageable chunks.  And by break down, I mean segregation into chunks that have their own independent time lines.  The only thing it has right now is the concept of externally linked designs.  The issue with those is that they can't be edited in the context of their parent assembly; they have to be opened in their own separate work space.  This is desirable sometimes when you just want to get everything else out of the way and work on a more focused area, but it totally sucks when you need to see how a single part or assembly fits visually into the larger design while you're making changes to it.  In Fusion you just can't do it.

The other problem with Fusion is that by not creating a distinction between "parts" and "assemblies", it forces assemblies to be history based in exactly the same way that "parts" are.  This lets you do some interesting things but it creates a ton of extra overhead that most of the time just gets in the way.  When I'm designing something I really don't care about the order in which all of the components are added to the design.  When I need to change a feature of a component that was added early on, I don't want to literally go back in time to when that feature was created just so that it can be changed.  Of course it makes sense to roll back the history of that particular part, but not the entire rest of the design.

This is the root of the question "why doesn't Fusion have assemblies?"

As far as "files" are concerned (meaning files on some kind of storage medium like a hard disc); the concept of a "file" is just a storage mechanism.  Whether a design is stored in a single "file" or is distributed over 1000 "files" is completely immaterial.  What matters is that you can decompose a design into pieces that are easier to work on independently.  If those pieces need to be referenced by other designs, then they need some sort of reference-able construct, but not necessarily a "file" stored on a hard disk. 

Also, I think the points about the linear representation of a design are extremely important as well.  Being able to see the feature dependency structure in a tree format would be an IMMENSE help in complex designs.  Every history based CAD software should have the ability to display the feature history in multiple ways.  Sometimes a flat linear representation is good, like when you need to see where to relocate a sketch so that it can be used for multiple features.  But then sometimes you want to know how many different features are dependent on a particular sketch, so if that sketch is changed, what might be affected?  For this a tree structure is what you want.

Incidentally there are other idea requests for displaying the time line as a dependency tree.

C|

[carloquinonez]

For assemblies, can you turn off the history and link to designs for the individual parts (which can have their histories enabled)? I don't miss assemblies at all... Kinda glad their gone, it was just another abstraction, with it's own set of assumptions (and limitations). Assemblies, Multibody parts, assembly level features, etc are the some of the ways CAD software use to handle complex references/dependency trees. But none of those feel "right" to me, at some point I always find myself limited by those tools when designing for additive. The biggest reason I feel this way is modularity and reusability of the CAD designs.

[kb9ydn]

For assemblies, can you turn off the history and link to designs for the individual parts (which can have their histories enabled)?

Yes but they will be external links which means no in-context editing.  The other problem with no design history is that patterns are no longer parametric.  So if I create a pattern of components, I can't go back and change the spacing for example.

I don't miss assemblies at all... Kinda glad their gone, it was just another abstraction, with it's own set of assumptions (and limitations). Assemblies, Multibody parts, assembly level features, etc are the some of the ways CAD software use to handle complex references/dependency trees. But none of those feel "right" to me, at some point I always find myself limited by those tools when designing for additive. The biggest reason I feel this way is modularity and reusability of the CAD designs.

They are abstractions, but they're useful abstractions.  They're useful for managing design complexity, and for building intelligence into the design.  This does add some management overhead of course.  Whether you want this or not will depend on what types of things you're designing.  For typical machine design it's great.  For design of parts that require a lot of shaping, direct modeling will likely be faster.

C|