Should be obvious looking at the output of listobs. oct/nov 2007. still looking into this. Fixed Dec 3, 2007. Always been present, but never showed up until the last few months. Load on acc seems to have triggered this bug in MAW data avering. If your data is mislabeled, a careful uvcat, puthd the source name could fix it. But for mosaic'd observations the pointing center would probably be mis-averaged, and although the visibility data seem to be ok, the data processing would be affected. The advice is to flag the time range of the mislabeled source, since they are small portions of your track, but if you want to get the most out of the data, are careful uvcat/puthd combination may get you there.