Back when I first started attending maths conferences, laptops barely existed. In those days, if you got bored or lost during a seminar or conference presentation, the normal thing to do was to drift off, daydream or discreetly work on your own problems on the conference-supplied pad of paper. It seemed important – perhaps out of courtesy to the speaker – not to make it obvious that you had tuned out, and so working on your own overhead transparencies or actually falling asleep was frowned upon, though neither was completely unknown.
Nowadays of course, almost everyone brings their laptop to a conference and so now the temptations of email, web-browsing or finishing up corrections to a paper are ever-present. However it is impossible to use a laptop discreetly and so it is immediately obvious, in particular to the speaker, that you’ve probably stopped paying attention to the talk. In some talks that I’ve been to, half the audience is working on their laptop after the first ten minutes.
What is the appropriate etiquette relating to laptop use during seminars? Is it disrespectful and off-putting to the speaker, or just a sensible use of valuable time? Of course the issue is clouded by the fact that there are perfectly legitimate reasons to use a laptop during a seminar – during Terry Tao’s plenary at the Aust MS conference, the guy in front of me was using his laptop to look up definitions or follow up on details on various things that cropped up during the talk. And if Michael starts “live-blogging” the Phan systems seminar, then he’ll need to be using a laptop for that.
Overall, I’m just not sure what to think.
On the other hand, some people obviously know exactly what they think. At last years ACCMCC meeting in Auckland, Doron Zeilberger gave one of the plenary talks. Doron, who is well known for having many strongly held opinions, started his talk by announcing that laptop use during his seminar was forbidden, and ordered everyone there to close their laptops or leave. My former PhD supervisor Brendan McKay and a certain prominent NZ mathematician who I shall name only as M were well known at the conference for seemingly being permanently immersed in their laptops. Luckily Brendan was there at the start of the talk to hear the warning, but for some reason M was about ten minutes late in arriving. We watched him with bated breath, and sure enough, after a few minutes the urge overcame him and out came his laptop, at which point a fired-up Doron strode up the steps to confront him, and shut it down. As M plaintively remarked later “I was only doing what he told us to do and looking at his website”.
Just as I finished writing this, I realized that one of Doron’s opinions gives an explicit description of the conduct he expects from his audience, and also mentions this same episode, including a link to the video.
(This talk was also notable for being one of the loudest seminars that I have ever attended. Doron is a very passionate speaker talking about things that he cares about deeply, and he doesn’t skimp on the volume.)
SymOmega 
I think a lecturer has the right to establish conditions s/he expects the audience to adhere to. If these conditions are unusual, s/he should provide a chance for anyone who does not like them to leave.
It is also clear that any behavior that distracts other audience members, is completely out of place. Arguably the lecturer has a positive duty to bring any such behavior to a halt, but s/he can handle the matter gracefully, and need not take it as a personal affront. (It would be a shame if passionate attachment to the subject was mistakenly interpreted as an over-developed sense of self-importance.)
On the other hand, if someone wants my undivided attention for 50 minutes, then they have obligations to me. They should be interesting and clear for the full 50 minutes, and they had better finish on time.
In the real world we often attend seminars out of a sense of social duty, and in advance we may have neither a clear idea of the actual material that will be presented nor evan an inkling of the level at which it will be presented. (Think, for example, of most plenary lectures at most conferences.) In this case it seem perfectly reasonable to undertake some quiet and non-distracting amusement, for example web browsing.
At some point, there must be a threshold where having a laptop open is deemed reasonable. Take for example a seminar in a small room in front of three people. In this case, it would be clear that no one of the three people should be occupied with a laptop for the duration of the presentation. Similarly, if all of the people in a room of sixty were using a laptop, almost every speaker would feel that they were wasting their time. However, in a room of 1000 people, if 10 were using a laptop, it would cause no issue (to most speakers). So how do we measure this? Is it as simple as measuring the proportion of people on laptops, or is there a nonlinear relationship here? This might be one for our other coffee drinking colleagues who turn coffee into models.