Bruce Perens wrote: > I feel that the draft GPL change for the ASP issue chooses the wrong > balance point between the right to privacy and the right to modify. > > In my opinion, the correct solution would be to require publication > of derived works, with one-time notification to FSF
Why to the FSF? Not all GPL-licensed applications are copyrighted by the FSF. > of the publication URL, What to do when this URL changes? > when those works are deployed in the form of a service to others. > I think you can make the language about this limited enough that > there would be no more privacy issues than those that would be created > by the "source button". And how will this allow the users (clients) of said ASP-type application to know that the source must be made available to them upon request? > I am the creator of a GPL-ed embedded system: Busybox. I would not be > able to use the "source button" on busybox because of the system size > constraints. > Embedded systems are subject to the ASP problem - Busybox > is used for routers and appliance servers. > > But I really wouldn't want to use the source button on any of my software, > because I feel it breaks the efficiency of the software to drag its source > code around with each and every runnable copy. It feels awkward and kludgy. > Source should live on a well-known internet server where it can be easily > retrieved and archived. It doesn't need to be a ball and chain. OK, so BusyBox doesn't _need_ to include the 'fetch source' accessability option. > The source button also seems to be too interface-specific to me. I can > think of any number of services where there is not a direct user-interface > in the form of an HTML form, etc., but just a set of RPC calls. It > might be difficult to find a "source button" in that setting, even if > one exists. But do we want to remove this notion because a subset of this type of application can't (easily), or don't wish to include it? Emile -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
