It had been a while since I had participated in a distributed computing project. Distributed computing was commonly used to refer especially to projects in which people around the world would contribute the use of their own computers, at night or during other slack time. This otherwise unused computing power was typically put to work on projects of a scientific and/or humanitarian nature. SETI (the Search for Extra-Terrestrial Intelligence) was an especially well-known and longstanding example of a distributed computing project. In the case of SETI@Home, as I recalled, users’ computers would be used to analyze radio-frequency records from various corners of space, so as to identify atypical bursts of radiation that might be emanating from living creatures.
I had been a volunteer for SETI in its early days. According to Wikipedia, it went public in May 1999; I was telling friends about it by December of that year. It was gratifying to think that I was pitching in on this grand project. It helped when distributed computing projects used cool screensavers or other graphical reinforcement of the idea that my little computer was doing funky advanced stuff that I did not understand.
I think every time I stopped doing distributed computing, it was because the software was not working like it was supposed to. It was seizing too much control over my machine, or making it work too hard, or not letting go when I needed to get some of my own work done. The last couple of times, I had been choosing my projects through the Berkeley Open Infrastructure for Network Computing (BOINC). BOINC offered software that would run any of several dozen distributed computing projects. BOINC’s software had been version 6.something, last time I checked; now it was 7.2.33. So possibly they had cleaned up their act, for my purposes.
It seemed that people had a variety of concerns about distributed computing. One that I don’t think had played a role in my own reactions, but maybe should have, was that a person could not necessarily be sure whether that software might be spying or otherwise violating privacy on one’s computer. In my own case, as I say, it had been more a matter of functionality and performance. I had decided to go with IBM’s World Community Grid, which seemed to be one of the few BOINC options drawing five-star ratings from a respectable number of voters. I hoped that this IBM nonprofit enterprise was playing the role of an honest broker in its allocation of distributed computing power to a selection of apparently needy and legitimate projects. Or at least I had my own cool member page on the World Community Grid website, telling me how many hours of computing power I had donated, and indicating which projects I had supported. It looked like Nutritious Rice for the World was one of the principal beneficiaries of my attentions, which (as I could now see) had ceased more than three years ago.
Anyway, the point of this post was just to introduce the possibility of distributed computing, to identify some possibilities, and to open the door for follow-up comments if I did decide to get back into the swim of things. I had spoken with a friend, just a few days earlier, who had never heard of SETI or BOINC or any of this; it seemed that there were probably others out there who would appreciate a brief public service announcement on the matter.
Which was all well and good; and yet, one day later, I found myself once again shutting down the operation and uninstalling BOINC. There were several reasons. One was that my desk was in my bedroom, and BOINC’s screensaver would light up the laptop monitor in the middle of the night and wake me. The screensaver would come into play even if I locked the machine. Since it was a laptop, I could not just turn the laptop off. I could close its lid, but I was trying to avoid making the lid sloppy from too much opening and closing. To similar effect, even restricting BOINC to 20% of CPU power would cause the laptop’s fan to roar at times during the night, waking me up. It rarely did that with my own workload. Another concern was that I tended to have Copernic Desktop Search (and sometimes other programs) running, and did not want its indexing (or other processes) to be deferred for the sake of the BOINC project. Probably the main point is that I still wasn’t at a point of confidence that BOINC would behave as a gentle, non-intrusive tool on my system.