On Tue, Apr 08, 2008 at 07:07:11PM +0200, Julien BLACHE wrote:
> > Ah, ok.  I though the hardware damage was just theoretical.  If it's
> > real, a warning would be appropriate IMO.
> 
> Believe me, it's not :) One of these days, you too will have to jump
> on the power button/plug at the rear of your scanner while at the same
> time shouting "WTF?" :)

:-)  In that case, while I still think this "I accept" window should
disappear, I think a warning window would be a good idea.  Like in
xvidtune, a warning that shows up every time the program is started.
Perhaps it should be possible to "not show this again", or something,
but by default it should.

The big difference between those is that the "I accept" only functions
to protect the author (and fails, but never mind that), while the
warning protects the user.  The latter is very acceptable; the former is
not IMO.  The whole point of free software is that we can remove
annoying "features".

> > So what is the issue exactly?  That he might be liable if the user
> > didn't accept the license?  Wouldn't that also be true for the rest of
> > the GPL'd software in Debian?
> 
> It is, indeed. The "NO WARRANTY" clause of the GPL is not valid in a
> number of countries. The GPL may be valid in full in the USA (though
> it has not been proven to date), but it for sure isn't pretty much
> anywhere else.

But does any legislation force a software provider to provide warrenty
by default?  I would expect that, assuming good faith, the default is
the same as when I give you something as a friend (as long as I don't
ask money.  If I do, it's good to be explicit about warrenty, I
suppose).

But maybe I shouldn't assume too much about sanity of legislation...

> >> accepting its license. In this case, the clickthrough makes it clear
> >> that the user accepted the license and is now bound by its terms.
> >
> > How does that make a difference?  The warrenty disclaimer is a notice,
> > not something the user accepts (AFAIK).  It just informs the user that
> > there is no warrenty.  I don't see the point in explicitly accepting
> > that.  If a user wants warrenty, they should get upstream (or whoever is
> > providing it) accept that they are providing it, not the other way
> > around.  But IANAL, and I could be wrong on this.
> 
> It's the same when you click "I accept" after reading (erm) an EULA
> while installing whatever proprietary crap (like, err, a SUN JDK for
> instance).

Ok, but that is because the license says you need to accept it before
you are allowed to use the software.  I'm not sure if they can actually
require that, but even if they can, the GPL explicitly mentions that no
acceptance is required to use the software.  And it says that adding
extra terms results in an undistributable piece of software.  So if the
warning really is a requirement (and I don't think it legally is, it's
just a request by upstream that Debian keeps it), then the software
can't be distributed at all.

> > I fully agree.  And I think these relationships are important.  But if
> > upstream is asking us to do things which we, as Debian, agree are not
> > the way to go, then IMO we should do things our way.  It's preferrable
> 
> I agree when it comes to things like license issues (cdrecord, ion3,
> mozilla) or equally important things. But this? Come on.

I think it is one big benefit of Debian that the users aren't harassed
with legal crap.  Setting a precedent of allowing authors to insert it
into Debian has the potential of making Debian *very much* worse IMO.
If every program would start doing this, I would seriously consider
forking the distribution.

> >> Anyway, I am amazed at the attention this clickthrough is getting.
> >
> > I think it has to do with what I wrote about how I feel about legal
> > stuff: I want to avoid it whenever possible.  With Debian, I can do that
> > (as a user) without getting in trouble.
> 
> You're a FLOSS developer and moreover a Debian Developer. You cannot
> avoid legal stuff, not one second.

As a user, I can.  When I need to install stuff on a Windows machine for
other people (I don't have any :-) ), I read the license.  When I
install a Debian package, I don't.   My sources.list only contains main,
and so I just know that things are good.  IMO this is actually a very
big benefit of Debian.  The feeling that you can trust the system, that
you don't have to read a license before installing something and still
knowing that you're not giving up your privacy, for example.

Sure I'm dealing with legal things all the time as a developer. But not
as a Debian user.  And that's very, very nice.

> That's unfortunate, but that's how it is today. (and it sucks, yes)

This reminds me of someone who said "With software, you have the dilemma
between following the law and being nice to people by sharing".  That's
true for him, because he uses non-free crap.  But it isn't for me.

In the same way, "legal stuff can't be avoided" is true for many people,
but not for Debian users (as long as they are in that role).

The attention of this bug should be explained (I think) by the fact that
(some) people really value that fact.  I do, anyway. :-)

> I am trying to protect both my users and my upstreams. This is a
> typical case: keeping the clickthrough protects both the user and
> upstream :)

Not at all.  A clickthrough which looks like a legal blurb to protect
the author does not at all protect the user.  A warning does.

Do you think there is even one single user who, when seeing a "please
accept this legal text" window, thinks "oh, I should be aware that this
software may be dangerous to use"?  It does however, for many people,
trigger a reaction of "wtf?  I thought I was using a system which didn't
give me such crap?"

Thanks,
Bas

-- 
I encourage people to send encrypted e-mail (see http://www.gnupg.org).
If you have problems reading my e-mail, use a better reader.
Please send the central message of e-mails as plain text
   in the message body, not as HTML and definitely not as MS Word.
Please do not use the MS Word format for attachments either.
For more information, see http://pcbcn10.phys.rug.nl/e-mail.html

Attachment: signature.asc
Description: Digital signature

Reply via email to