Specifically, my comments will be focused on using the Calibre Interactive (GUI) for xRC but it could apply to the other flows as well. We have many users of the GUI at my company including myself that have become very familiar with it's layout and features. However, I find that there is still one aspect of the GUI that is annoying. The structure of the GUI for xRC is global in nature i.e. the settings are not organized by any particular type of extraction.
Using the GUI for xRC is very nice mainly because most of the vast array of options that can be used to customize the desired final netlist are available somewhere in the menus, buttons or tabs. The problem is it can take a bit of navigating and setting of options before you are finally able to perform as an example a lumped capacitance extraction of a selected net that excludes stdcells. It is more challenging for the users that are new to the GUI or xRC.
Fortunately, Mentor has provided the ability to save the state of the GUI as a file (runsets) and then reload that file when that same flow is desired. They have also provided a means of customizing the GUI. At our company we do make heavy use of runsets but more than that we have a way of managing the runsets so that people can quickly run a specific type of flow.
I'd like to suggest that most users not just at our company (since we have in some ways worked around the problem already) would be able to more quickly perform extraction tasks if Mentor reorganized the GUI to be extraction task centric. Here are some possible benefits:
There would'nt be uncertainty whether you set all the right options for the given task.
This would eliminate the confusion that sometimes exists when you don't know whether an option applies to the type of task your doing.
There are many common extraction tasks such as those in chapter 6 of the xRC User's Manual that could be offered.
Options that are common to all extraction tasks would be separated so that no matter what you're doing you know you need to set those things first.
The environment variables that control extraction could also be grouped by extraction task and set to an appropriate setting.
Offer extensibility to developers so they have the ability to add a custom flow.
Let me know what you think.