> Another "we all hate cursors" article

Not necessarilly. I'm not disputing that cursors have their uses, and when you need one then you need one. However it's certainly worth being aware of the cost, and particularly so when writing a trigger. I don't think I'd recommend the route that the author described, as I think it could hide the fact that the real cost is in executing a complex operation n seperate times, and the cost of the cursor is small in comparison. I've certainly come across code like that. However if it's in a time critical piece of code (and unless I knew unequivocably otherwise I'd automatically assume this applied to a trigger) I'd devote some effort to making the stored procedure (to use the example in the article) accept a set of input items and process them in a single pass.