There are two questions there. On one hand, if there's demand, we'll address demand through existing assets and resources that we have. If they say, “I have a new application. I want to host it”, that can be as big as a breadbox or literally, a building. We have to assess that, so it's either a major project or we simply take part of an existing server. That's not a problem.
On service levels, going back to what Ron alluded to before about having set service-level expectations, that's what we're ramping up to try to set those as a standard as opposed to...simply what we inherited. Because, again, the premise when SSC was established was that we would...
And I have to tell you, this is the way it works in industry at very large scales. We used to call it “Your mess for less”, because we're basically going to keep doing the same thing, but through economies, costs will go down over time. That's basically the value proposition for containing the legacy environment until we get the future state stood up and everyone migrated over.