Frank Mittelbach <[EMAIL PROTECTED]> wrote: > Walter Landry writes: > > 5a1 is not a free alternative. 5a2 approaches that, but it has to > > cover _every_ occasion where 5a1 fails, not just most of them. > > I don't think it is acceptable that you take a list of "or"s, judge > each of them individually and conclude that each of them is not 100% > therefore the whole can't be either. Not 5a2 has to fullfil DSFG but > 5a. > > if i distribute FOO (that does runtime info to users ie is > interactive in this sense) under lppl you can of course use parts of > foo in "grep" since "FOO" neq "grep" there is no requirement for at > all. > > if you want to distribute a variant FOO as a replacement for the > original FOO then it would be "interactive" as well, thus the > requirement in 5a2 would be reasonable where would that conflict > with DSFG?
What if I want to distribute a non-interactive replacement for FOO? That is, suppose someone writes a crappy version of grep that spits out this information everytime you run it. I can't take those announcements out. GPL 2c, on the other hand, does allow it. > ps your suggested rewrite is by no means similar to GPL 2c since 2c > depends on an external fact about the modified program "being > interactive or not" while your rewrite makes the applicability a > decision of the author of the Derived Work which has no binding at > all I know it is not the same, but I couldn't come up with something better. I'm open to suggestions. Regards, Walter Landry [EMAIL PROTECTED]
