What accounts for the load time in Knime workflows? As I have been buildings (relative to my previous work) larger workflows I have noticed a significant increase in teh time to open a project.
This is just a wild guess but does the workflow load and cache the data for each node upon opening?
That will change in 2.0. The problem that you have run into is that the contained nodes get configured() upon opening - and that multiple times, which is expensive.
2.0 will have a different loading scheme that fixes the problem.
The data in the outports is restored at the latest time possible (on access). That is true for 1.3 and 2.0.
PS: The loading was one of reason (among others) for us to rewrite that particular piece of code.