I remember learning someplace that it takes the average person 45 minutes to regain mental focus after they've been interrupted. I think of this often when I'm in "the code zone," am asked to break away in order to do something silly and then can't remember what I was doing when I return to my desk.
I wish I could find that study - I could pass it out to managers everywhere as an argument for things like telecommuting and office doors and against things like pointless meetings, loud employees, and stupid questions...
Re:DeMarco...
Herkemer on 2002-10-07T21:28:29
Yep, it's Chapter 10 in my edition. I've got the page marked and often make people who interupt me read the section on "Flow" before I will help them.
*grin*
Re:DeMarco...
ziggy on 2002-10-08T01:16:05
Yep. It's one of DeMarco's long standing rants. He talks about it in his latest book Slack. It's intended to be a quick read, and focuses at some of the higher level issues in management, especially management in software development. (i.e., something you can give your skeptical manager and hope he'll learn something from it.)When he gets to the topic of task switching, DeMarco asserts that there's a 15% overhead incurred when switching tasks. He proves it by repeated assertion: it must be true because he's been saying it for years and no one has really refuted the point.
;-) The other way to look at it is that the idea of zero-overhead task switching is also a proof-by-repeated-assertion. However, simple analysis demonstrates that task switching does have a cost (especially when focusing on intellectual activities), so given the two choices -- 0% and 15%, 15% is likely to be closer to the truth (if only because it accepts that there is a cost to pay.)
Re:DeMarco...
james on 2002-10-08T11:01:24
I've just finished reading Slack, and its been left on the desk of my indirect boss. He gets it, but others at that particular level of management don't.
I'm not sure it really adds much to Peopleware, but it is written as a management-handbook type thing, rather than a Knowledge Workers Unite! type thing, which probably will earn it some brownie points with non-technical managers.
Overall a pretty good book. Something I did at a previous job was get the development team a secretary, and the time it saved for everybody was enormous. DeMarco explains *why* it worked and how to repeat it. This is a good thing:-)
