This really doesn't answer my question. Does the resource share actually work? I think no. Otherwise there would be no way that collatz(0) would run before MW(100,000). I just don't understand what else I could change. And I hate to say I told you so, but MW is down again and my computer sat idle until I noticed it this morning. I would figure that there is a software guy somewhere reading this and thinking that he knows how to get this to work. Maybe if all of those people at Berkeley stopping watching that muffin-top girl up in the tree because they are going to cut it down for another building would apply the share to the gpu I could stop ranting, until then I will keep testing.

---

Resource share is a long term average, over several months. It's not intended to be minute by minute.

Getting work from a project(s) to tide you over outages is the job of the 2 caches in Network usage.

Resource share is a long term average, over several months. It's not intended to be minute by minute.

Getting work from a project(s) to tide you over outages is the job of the 2 caches in Network usage.

The 12 MW tasks that I keep in cache are completed in about 12-13 minutes. This is about 0.01 days. So what boinc needs is less tasks in collatz than MW, so I need to set my collatz cache to store less that 0.01 days worth of work?

---