Hi, thanks for providing this example. MJ Ray wrote: > For example, a PHPBB service page is about 20k, while PHPBB source is > 2.19MiB. You have a mighty uninteresting forum if people only look at a single page. Kidding aside, if 10% of all users download the source once and they average 10 pages, you have even portions (and a lot of forums seem to allow uploading images and stuff). Sure, if everyone visits one page and downloads the 2MB source, it's disproportionate, but it is but one amongst many abuse scenarios for such a service. Also, web space in the region of a few MB is a commodity - it would be ~20 Euros/per Month for virtual server that can hand out about .2-.25 of these 2MB source tarballs per second.
> Which services are you using? Webmail, some image galleries, a Zope instance with a limited number of users, Debian services, mostly. >>> and - as far as I can tell - it >>> is not generally accepted yet that providing a link to your project on >>> someone else's hosting site is sufficient [...] >> What does "generally accepted" mean to you here? > Ideally, in statement(s) by the licence author or a suitably large > number of users of the licence. Well, I'd recommend pestering the FSF for such a statement (or the reverse). They seem to be amenable to appending their FAQ. >> I propose that it is no different in the sense that "I put my code on >> Savannah for you to download" is no different than "I buy a server with >> provider Foo and put the code there". >> This is not legal advice, but I don't think you have a court case over a >> few days that the download is broken, be it for a typo in the URL or >> because you have to switch the provider of this service. > I agree that it doesn't matter whether the code is on Savannah or > provider Foo, as long as provider Foo isn't also the Program's host. > I propose that allowing users to interact remotely while not providing > an opportunity to receive the Corresponding Source for whatever reason > is a breach of the AGPL as written. In my experience, the judicial systems tend to restrict excessive license terms. In particular, the language of clause 13 "offer an opportunity to receive ... from a network server" does not look like a transient problem on that network server qualifies as a breach of the license. >>> If you know of general acceptance of hosting sites as sufficient, then >>> I suspect many people would be interested to learn of it. >> How can I know whether it is sufficient? If you want to, publish >> something that interests me under AGPL and sue me for violating your >> copyright when I deploy it and provide source via such a service. > I've got two formal disputes under way just now and that's two more > than I want, so I hope you don't mind if I don't take up your kind > offer immediately. No problem. Note that it also involves you writing software that interests me, so I can look forward to that. :) >> This is not legal advice, but you will lose that. Law is not >> mathematics, the scenario you are constructing here is not realistic. > I am a statistician more than a mathematician, so I am well aware of > probabilities and risk and so forth. Unless someone can show > evidence, there seems little reason to assert "you will lose that". Well, as a statistician, you are probably aware that people try to predict things using market models (or more simple betting with odds) with some success. But then, people also claimed to have done the math for the financial markets and got screwed by too many "not very probable" events being too correlated. Next we know some Utah company might claim that AGPL is unconstitutional. I think I'll stop here for this discussion. Thanks for elaborating on your arguments. I'm not quite convinced that you're arguments have a lot of traction, but most likely you are not that easily convinced either. Kind regards T. -- Thomas Viehmann, http://thomas.viehmann.net/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
