set a flag on snippets that would fork the evaluation of the snippet into a thread pool and then allow the weaving of the responses at page render time. This would be a benefit to systems that need to fork multiple long running threads (e.g., ad servers, etc.) to pull data in.