[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > What if IBM had left the patent clauses out entirely? Would we not > then be saying that the license was clearly DFSG free? Yet such a > license would grant *fewer* rights than does this one.
In general, it's not possible to know about what patents cover a work of software. This is because (a) the existence of a patent is secret until it's approved (b) software is just an abstract collection of energy -- it's what it symbolizes that makes it tread on a patent, or not (c) patents are written in a fashion which does not make their applicability to a work of software immediately evident. So, we're forced to go on the basis of what we know. If we have no reason to believe that patents are an issue, we proceed on that basis. If we have reason to believe that patents are an issue, we proceed on that basis. If, for example, we have every indication that Jikes is not constrained by any significant patents, then it doesn't really matter what the license says about patents. But if we at some point find out that it is constrained by some patents we'll have to clean up the resulting mess. The simplest assumption, on seeing that patent clause, was that it would be significant. However, the additional information which has been presented here (that the license is intended for re-use on other pieces of software) makes it less likely to be a problem. At the moment, I think that this license would be ok for debian inclusion (but that we should toss anything out of debian that IBM asks us not to distribute under our terms). -- Raul
