Another issue with the "starshot" plan is the lightsail might tend to fold up, crumple, or otherwise deform into a poorly- or non-functioning shape. Is that problem defeatable? If we imagine a 1-dimensional lightsail in a 2D universe for simplicity, the lightsail is then a curve-segment festooned with 1 or more deadweights. If there are no deadweights at the curve endpoints, we clearly lose since the curve will just bend upward to vertical and nonfunctioning position. If there are two equal deadweights at the two curve endpoints, then they will experience vertical downforce from the gees, and also tension force from the lightsail-curve. The vector sum of these two forces will always have an inward component if the curve has finite |derivative|, hence such a lightsail will always "fold" as the two deadweights move inward to meet each other. So: Also lose. If the curve has vertical derivative at the endpoints (such as a semicircle) then there might be hope for a (perhaps unstable) equilibrium configuration. And instability (if any) perhaps could be active-controlled (albeit the enormous accelerations we are talking about might present some difficulty for controllers to handle). In a 3D universe with a rotationally symmetric (e.g. hemispherical-shaped) lightsail and a deadweights festooned along its rim, this again could (I think) be designed to enjoy equilibrium and actively controlled to overcome instabilities. Also there now is an additional option of adding spin causing centrifugal outward forces, which could try to compensate for the inward tensional forces. So I actually think this particular problem is not inherently insurmountable, albeit it is annoying to have so many deadweights round rim that all play active control games, rather than just one, like in the press description involving just ONE "chip" . But anyhow, bottom line is I think the whole "starshot" proposal probably is total no-hope hogwash; it is just that this particular problem is not the reason it is. -- Warren D. Smith http://RangeVoting.org <-- add your endorsement (by clicking "endorse" as 1st step)