So Craig, it’s a little hard to quantify the overall end market, but the POS is easy to quantify. So the POS in the second quarter was approximately 320 million, and so a normal seasonal build in POS, or I guess I’ll call it an average seasonal build, would have carried it up from there. So if you look at the process that we use to estimate that, it’s heavily informed by the resale expectations from our distribution customers themselves—or our distribution partners themselves, and they look at sort of who wants what and their own trends, and so they give us that. We massage it a little bit and it turns into our point-of-sale forecast. It’s generally been pretty accurate, and so it was, I think, unexpected that it was weak. It took some time for our distribution partners to come to grips with what it was likely to be, so it took a couple of iterations to get it defined. And so it wound up by about mid-quarter, we had a pretty good grip on what it turned out to be, so I think everyone is continuing in that vein in Q4 but rather than reacting to it partway through the quarter, which we clearly saw going on, I think people have their arms firmly around it and are aggressively trying to get it adjusted in the early part of the fourth quarter, and we’re seeing that kind of behavior. So in terms of what the true end demand is, you really have to go in a segment at a time. So it’s clear that there was equipment built in the second quarter that didn’t sell through as rapidly as people expected it to. Probably the most extreme was in the solar inverter space, where there wound up being a lot of inverter inventory and the build plans for inverters turned down very sharply. How much you could subtract from the actual demand in Q2, it’s a little bit hard to do. Again, with many small customers as you often see in distribution, the visibility into the granularity of where all the stuff goes isn’t that good. So again, I think we could subtract some from what the overall Q2 demand was, but I don’t think it was a radical over-build, for example. I think it appears right now to have been a combination of a modest over-build and increased availability, allowing people to draw down inventories; and those things compounded have turned it into what appears to be a fairly sharp but likely contained inventory correction.
Craig Berger – FBR: Thanks for that detail, Mark. And then just how do we think about spending items as we begin 2012? Are you guys going to be holding the line, or do you have annual spending and salary increases? How do we think about the growth in OPEX?