That is to say in the previous post that it would be perspicacious to address the manner of making the "in memory tree view" of the database consistent with the extant "persistent datastore" in an efficient manner rather than to make assumptions that there are no other influences on the "persistent datastore" (unless, of course, one has taken steps to ensure that there can be no other other mutators of that persistent datastore). In other words, reliance on assumptions of fact being true, when they are actually false, will cause erroneous results.