In applications that deal with RESTful web APIs I frequently come across situations where the code looks like the folowing.

• API call to get number of "items" available
• Loop over a series of API calls to get those "items" in batches

The problem is that if I use a progress bar to track progress, naïve code suffers from a race condition if the number of "things" available changes between the call to get their count and the calls to fetch them.

This results in a lot of repetitious boilerplate code along the lines:

if count <= max_value:
    bar.update(count)

It's easy to forget to include the guard, which results in annoying non-deterministic failures that are hard to catch in testing.

It would be nice if the library could (optionally) include logic to ignore updates outside the expected range.