Anoo S Pillai (3/18/2014)
A nice question, somewhat spoiled by a careless explanation.
You are setting a tough target Tom, Copied the explanation from BOL for un-ambiguity. Sad to see that it didn't work. Could you please ponder some ideas on a good explanation from SSC point of view.
Unfortunately BOL is sometimes not as clear or as accurate as it ought to be.
I think that "moves the record column with the largest width to another page in the ROW_OVERFLOW_DATA allocation unit, while maintaining a 24-byte pointer on the original page" could usefully be changed to read something along the lines of "moves columns to pages in the ROW_OVERFLOW_DATA allocation unit in order of decreasing column length, leaving 24-byte pointers in the original page, starting with the column with greatest width and stopping when the the resulting row with pointers is less than 8061 bytes". But maybe there's a shorter way to say it, or maye just leaving the original and adding "and repeats this with the widest column remaining in the row until the row is short enough" would be better.
Don't take "somewhat spoiled" as being anything like as strong as "spoiled" on its own, by the way. When I saw your reply I realised that I'm not sure whether "somewhat" changes meaning from place to place or not, would have expressed it differently if I'd realised that when I was writing it.