everbeek
commented
9 years ago I am not concerned that doing multiple child calls at a time for a single concept could be disruptive to a single instance of Biomixer, as well as for when multiple instances are running. It would increase server load, and could lead to browser side retry-churn.
The calls to get node children result in paged REST calls. That is, when we make them, we make a series of calls, each with a limited number of included results. For the UBERON artery example that I am fond of, browser load seems to be noticable. I feel like it could be improved by seeing if either the bulk facilities could take a request for a handful of child pages, or if it would help to perform up to, say, 5 at once. The current approach is to get them fully sequentially.