Hi Jack,
This is very interesting. In fact, the main purpose of process icons and process containers in PixInsight is to serve as reusable and adaptable processing interfaces. This is a key part of the object-oriented architecture that is one of the fundamental paradigms in the design of PixInsight.
Of course we always try to reuse process icons in our processing work. However, in general we find that each image is a different case, requiring fine-tuning of many process parameters. Without doubt, having a good arsenal of process icons stored and well organized is of great help.
There are some processes that are especially sensitive to an image's particular characteristics. Perhaps the most clear case is SGBNR, whose parameters are extremely critical. Wavelets require also a good adaptation to the image to be efficient.
We want to develop the process icon system much further. For example, the next implementations will allow the user to write a descriptive text of arbitrary length for each icon. Having a process set well documented is essential when one reuses it, say several weeks after having created it.
There will be also much more flexibility to organize and edit process containers. Right now the interface to do this is quite rough. Individual process parameters will be editable in-line, without needing to open the associated processing window.
We'll include also a scripting system to enter a sequence of processes defined as plain text files, following an easy syntax. These scripts will be similar to the current files generated by Processing Report. Scripts will be an alternative way to view, edit and manage processes in PixInsight.
I haven't tried to make specific sets of process icons for some types of objects. However I think that with some restrictions this can be very useful. As you say, processes always have to be fine-tuned, but the skeleton of a process or some particular processing strategies can be usefully stored as a set of icons that can be reused.
Juan
