History versus Feature versus Parametric modeler: what are the differences ?

[mbraga0001]

New to Autodesk Fusion, but not new to CAD industry.

I appreciate if someone could give me some tips about these concepts (or the differences among then) and what kind of modeler Fusion is.

In the mean time, I understand in a ... parametric-feature-history-based modeler (as I think Autodesk Fusion is), one can see and change the model by its creation structure. The consequence of this is ... HUGE.

For example, allow me to make an analogy with a program (from a developer perspective): in this analogy the executable is the model and the code is the creation structure. Once you change the code (creation structure), you need to re-compile to get the executable (model). In order to change, you need access to each and all constructions elements, that can be ... a lot.

Before I go further ... Am I in the right direction ?

Thank you

Marcio

[jakefowler · Accepted solution]

Hi Marcio, and welcome to the Fusion 360 community!

Yes, I think you understand the fundamental concept of history-based/parametric modelling (these terms are interchangeable in general use). Fusion 360 records each feature used (and their associated inputs) in the timeline running along the bottom of the window. As you suggest in the programming analogy, this is effectively a ‘recipe’ for your model. You can replay this at any time using the ‘play’ button at the left of the timeline, and watch the model build itself from scratch, to see this recipe in action.

There are many advantages to this approach, one of the most useful ones being that you can change the inputs/values for an earlier operation (e.g. in response to a design change request), and the related features will update automatically. Without this history, changing an existing region of the model will in most cases require you to manually rebuild many connected downstream operations too.

Regarding your concern that changes require access to the entire history: generally, when you make a change, only features later in the history require updating (an incremental build, if you like :)). So changing a recent feature should recalculate the model more quickly than changing a feature very early in the timeline.

Fusion 360 actually started off as a history-free modeler, and when we decided to add history, we implemented it in a way that aims to avoid adding complexity & fragility to the modelling process. For example, in Fusion 360, when you make a change that can’t be resolved automatically, we use a cached version of your previous geometry to prevent the model from failing outright. We also try to intelligently interpret modelling changes (for example, moving a face will (where possible) change the inputs to the feature used to create that body, rather than adding an extra ‘move face’ feature to the timeline unnecessarily).

Garin from our product management team put together a video that does a good job of explaining the advantages of Fusion 360's history-based workflow:

Essentially, we’re aiming to offer the best of both worlds: the freedom of history-free modelling along with the time-saving benefits of history-based modelling. This approach is based on some new technology, and we’re certainly not finished working on it yet: we’re still adding features and fixing issues in the workflow to get this working how we’d like it. And since this is all brand new, we’re also heavily soliciting feedback from customers, much of it from this forum and the IdeaStation, to help guide our future development direction.

Hope this answers some of your questions, but please let us know if you have any more!

Kind regards,

Jake

[mbraga0001]

Thank you, very much, Jake for the explanation.

Allow me to suggest to use History-based design to constrast with Direct design.

Both solid capabilities (Features and Parametric) can be used with these 2 design styles (Direct and History-based).

Congratulations: Fusion is becoming impressive.