So what do the previous posts regarding direct and indirect editing mean to product design?
As history-based modeling has been the dominant technology for many years it gets used for a wide variety of design. As history-free modeling becomes more robust and useful we now have two very viable and very different technologies to choose from. As this choice impacts the data or “intellectual property” of your company and will impact your product development process, the choice should be based on something more than just user preference.
With history-based technology, editing is typically indirect and the layer of intelligence used to drive the edit is created at the time the model is created. It’s all about creating a properly structured history tree. This environment naturally demands consistent modeling methodology and close attention to what and how the model may need to be changed in the future. The basic benefit of such an environment is that by appropriately recording every modeling step, anyone of the steps can be edited resulting in a different model. The basic challenge with this environment is that it requires consistency in modeling practices, can be very inflexible, can be difficult to learn and will require a good knowledge of the structure before utilizing the data. What you "see" is not what you get with history-based modeling. From a process point of view you need to make sure that the value you gain from a well structured history tree is worth the investment.
Direct editing will continue to be added to history-based modeling and will bring some added benefits, but as long as it is history-based it will be somewhat limited in capabilities.
If your process and products demand highly structured engineering data and this data will have a lifespan and utilization that will justify the investment, then history-based modeling may be a good fit.
With history-free technology, editing is typically direct. There is by default no data in the model that would describe how the model was created or that would drive one feature to influence another (parent/child). With history-free modeling what you see is what you get. It provides a very fast and flexible modeling environment with no requirement to understand design intent before beginning the modeling process. The basic challenge with history-free technology is in the editing of complex shapes. While the technology is rapidly improving, editing a shape that consists of B-spline or NURB surfaces with lots of rounds and filets can sometimes be difficult. Again, it goes back to topology. Keeping these surfaces “connected” during stretches and moves is not trivial. These complex shapes can certainly be modified with history-free technology, but it’s not as simple as changing a parameter and replaying the history, (assuming the history will support the edit).
We will continue to see more indirect editing capabilities in history-free modeling. PTC CoCreate has been using the D-Cube DCM technology for many years to drive its indirect editing capabilities. LEDAS has also developed very similar technology and you can now see this technology working in Rhino.
From a process point of view, if you just need to get designs done as fast as possible with as much flexibility as possible, then history-free technology may be the best fit. It is also a very lean technology. There is simply much less data involved with history-free technology. Large assembly in-context design is a natural fit.
The above comments have more to do with the use-model of the tool. From a use-model point of view there are many differences and these differences will certainly impact the design process. Of even greater importance than use-model is “data”. While both history-based and history-free technologies create solid B-Rep models, there is significant difference in the actual data that is being created. This data is valuable intellectual property (IP). This IP carries with it the results of a significant investment in product development. It is at the core of your product development and may have a lifespan measured in decades.
With history-free technology the 3D B-Rep solid is the master document. Without it you have nothing. With history-based technology the B-Rep model is just a by-product of the history tree. The history tree is the master document. In the context of history-based modeling, without the history tree you have nothing - but a “dumb” solid. The biggest and most significant issue regarding maintaining a history tree as the master document is that a history tree is proprietary. There is no industry standard for exchanging this data. As such it is likely that much “history-based” engineering IP (the history tree) is going to have a “relatively” short lifespan. Pay close attention to the IP of product development, the investment you put into it, and how it may impact your future product development.
It is interesting that some CAD companies talk about combining history-based technology with history-free technology. It will be interesting to see what they come up with. Today CAD data is either history-based or it is not. Once history is gone, it is gone. Recreating a history tree from a history-free model will be similar to recreating a photo album that has been destroyed in a fire. History-free and history-based CAD systems can certainly coexist and complement each other, but it must be well understood how this environment, and IP, will be managed.
Now that these two choices exist and are both very capable, people should take a close look at their product development process and consider which technology best fits their current and future needs. There are many characteristics to consider.