On 7/13/05, Sean Kellogg <[EMAIL PROTECTED]> wrote: > I am so confused. #1 allows a licensor to impose all manner of terms without > giving actual notice to the licensee, whereas #2 at least gives the licensee > a chance. The warranty provisions are a great example. The GPL rejects all > implied warranties, but doesn't tell a licensee it does so unless they go to > the trouble of reading the COPYING file. How does displaying the license > first and requiring folks say "yes, I understand" more problematic or > invasive?
Click-wrap that isn't trivial to circumvent is a sysadmin's nightmare (what kind of crap-ass software can't be updated site-wide without screenscraping GUI macro magic?) and it's not smart to require it where it's not necessary. If you can get people to behave decently with respect to the temptation to steal more than the vendor is offering, you don't need to club them with FUD. The statutory penalties for copying without license on a commercial scale are pretty steep, and the principal benefits of publication under the GPL can be tied quite satisfactorily to that need for license when modifying and/or copying. There's no sane boundary between making warranty disclaimers (practically legal no-ops at the retail level) binding and allowing enforcement of arbitration clauses on people who just intended to purchase a retail good or even pick up a free newspaper (which is surely bad public policy). So if your concerns relate to the body of the GPL rather than things that have to be hung on a right-to-use hook, why futz around with click-wrap? > Believe me, I understand the visceral reaction to click-wrap licenses. I have > had a lot of debates with law professors on the issue of whether click-wrap > licenses are a "good thing" since they postpone term presentation until far > after money has changed hands. But no one has presented a cogent argument > about how mandating that people actually agree to the terms of the GPL poses > a threat to the DFSG. It's not allowable under GPL section 6, it's inconvenient for important categories of users, and it's just plain stupid to do package-by-package anyway. Cheers, - Michael
