On 5/8/05, Andreas Raab <andreas.raab at gmx.de> wrote:
> Actually no. It's the only license we require you to *accept* before you
> get the download.

Err... that usually (certainly by US law as far as I have followed it
during the last few years) is the license that will govern the
contract. What you accept on a click through is the valid license.

> Once you start Croquet you will find that a "jump to
> previous project" will get you into a vanilla 3.6 image [...]

Sure. But the click-through license is what counts.
> But I
> would be careful about the "Squeak is included" bit - by the same
> measures (having a pre-loaded image) one could claim that "Squeak is
> included with Seaside" that "Squeak is included with Tweak" etc.

Nope. Go download Seaside. Avi and friends don't offer a preloaded
image of Squeak with Seaside under their own license. Completely
different thing.

The deal with Croquet is:
- I click to accept a license (as far as it goes, click-through is
considered to be valid);
- I get software. An image and a virtual machine and a whole
collection of files. That is what VRI licensed to me. The works. Not
just whatever VRI considers to be part of Croquet, because how can I
know that as a simple Joe A. User?

The deal with Seaside is:
- I click to accept a license (is there a license on Seaside? whatever...);
- I get a package. I need to go to a separate place to download the
environment for that package, that place will present me with a
separate license (Squeak-L)

> The
> license of proprietary linux kernel drivers does not change even if they
> are "included" in Linux distributions (and even if you don't
> specifically accept their license before downloading the distro).
> 
Maybe, but you are stepping into very muddy waters if you take the
Linux kernel as an example. Heck, for most of Debian I don't even have
a click-through license...

But VRI *does* post a license. Without restricting parts.

> If you read Squeak-L carefully you'll find
> that VPRI does indeed adhere to all of the terms included so your
> finger-pointing might be totally off.

What Squeak-L? I'm arguing that - as an unsuspecting user - I'm only
bound by VRI's license. So I can fingerpoint all I want. VRI licensed
the package to me (that gigantic zip file with whatever is in there),
so I just put the blame there.

> Probably not. And I urge you guys to think about the consequences very,
> very seriously before you consider anything like the above. You could be
> making matters worse than they are.
> 
Dunnow. Again, for myself, I don't care. I'm currently doing a
government-sponsored open source project with Squeak, and nobody seems
to care about the small print, and not even about the license that we
will use for our own stuff, as long as it is somehow open source.

Open source + law just don't go together at the moment, so until we've
had some nice court cases, like around the Linux kernel, the trickier
things of the GPL, the enforceability of open source licenses at all,
etcetera, my stance is that I'll let myself be bound by the spirit of
a license, rather than by the letter.

But, if you put on your lawyer hat and want to satisfy the jerks over
at Debian Legal, it *is* a very interesting opportunity. Really,
there's no way arguing around the fact that when downloading
OpenCroquet, I receive a big ball of fluff that VRI licenses to me
under an extremely liberal license - if I would put up a Debian
package for OpenCroquet, it'd be immediately accepted...

I really wonder what the idea is behind this all, because somehow I
cannot believe that of all organizations, VRI has overlooked the small
matter of the SqueakL. And I'm also hesitant to believe that VRI wants
to keep a low profile around OpenCroquet (which is the ultimate
defence against licensing trickiness...). So if anything, I'm curious
;-)