Re: License issue, and a possible solution

On Tue, Apr 27, 2004 at 08:29:50AM -0600, Kurt Fitzner wrote:
> Alan Coopersmith wrote:
> 
> >>Think of it this way.  Would people be paying as much for products 
> >>like Hummingbird or other commercial X servers if those companies had 
> >>to acknowledge that they stripped much of their code from XFree86?  
> >
> >If they are using code from XFree86, they already had to acknowledge it
> >in their copyright statements.  Acknowledging it even more wouldn't change
> >how much most people would be willing to pay for a product from any of the
> >commercial X-for-Windows vendors.
> 
> It's a perceived value, issue, I think.  What is the perceived value of 
> a product that Hummingbird produces, vs. the perceived value of a 
> product that XFree86 writes most of, and that Hummingbird slaps PC 
> compatibility onto?
> 
> >>In GPL, commercial companies wouldn't be able to use the code at all.
> >
> >Completely untrue.
> 
> <sigh> Ok, technically correct, yes.  But companies tend to shy away 
> from using code in commercial products that would require them to 
> publish the source code.  If people could compile and run Staroffice for 
> themselves, would there be any reason to use it over OpenOffice.org?
> 
> >>With BSD (and the similar 1.0 XFree license), commercial companies can 
> >>use the code, and hide teh fact to their customers.
> >
> >Again completely untrue.
> 
> And again, you are technically correct.  There's the "advertising 
> clause"ish type statement in old BSD and the XFree86 1.0.  But it's a 
> lot easier to hide the fact that XFree86 code is being used in a prudt 
> when all you need to do is stick some copyright notice somewhere in some 
> text file, rather than if you have to put something visible in an about 
> box or similar spot in the GUI.

Since the focus seems to be on the X server, I still don't see the
point.  I thought we already established that the problem with the new
license is how it is going to affect distributors of the thousands third
party X applications.  So far as I know, X servers do not pop up
windows or splash screens attributing third party authors.  At least
not that I have ever seen.  So, putting it in a text file in a
location where other credits are would satisfy even the new license.
And if they are truly excluding client libraries so that we do not
have the problem we have been discussing with the applications, then
what have they accomplished with the new licenses but to create a
boondoggle with significant legal risk to the distributors?


> >Sun already includes XFree86 copyright notices and acknowledgements of
> >various other forms in both Solaris and the Java Desktop System for 
> >Linux, so you can count us out of your conspiracy theory.
> >
> >A simple search for "XFree86" on http://www.sun.com will find pages 
> >crediting XFree86 in various ways, such as:
> >http://developers.sun.com/solaris/developer/support/driver/tools/video/video-index.html 
> 
> The above page, and several others in the search you suggested, talks 
> about the XF86 video driver porting kit, or mentions that XFree86 is "an 
> open source implementation of the X-window system".  I haven't gone 
> through the list exhaustively, but no page I've read It never actually 
> says that XF86 code is used inside the Sun X server.
> 
> Now, let me backpedal just a little.  That is, I don't really believe 
> that there is seriously a conspiracy - which is why I put "conspiracy 
> theory" in quotes (Canadians' attemps humour sometimes fly under 
> people's radar - as opposed to British humour which slaps people in the 
> face with herrings).  That being said, however, for all the companies to 
> migrate their support away from XF86, there has to be a perceived value 
> in the change.  And the only perceived value I can see is in the only 
> real difference, which is the license.  This means that putting 
> "Portions of the Sun (HP/Hummingbird/etc) X Server are based on code 
> from the XFree86 Project , Inc" in some about box was so much of a 
> problem for some people, that they were willing to fund the starting of 
> an entirely new project to get around it.

As I said above.  I think the new license accomplishes nothing for the
XFree86 project.  If the X server does not already have pop up
credits, then it does not require a pop up.


> Sun is in it to make money.  Publicly traded companies exist to make 
> money for the shareholders.  Therefore, it looks to me like someone at 
> Sun thought that the above about box would "cost" Sun more money than 
> funding a new project.
> 
> Perhaps I'm reading the situation entirely wrongly.  But while I believe 
> that the license change was unnecesary (and perhaps unwise), I also 
> believe that the reaction to it was completely out of proportion to the 
> change.
> 
> If I am reading this wrongly, I would very much appreciate knowing what 
> all the fuss is really about.
> 
> 	Kurt.

I disagree with you about the reaction.

I don't think it is about trying to cover up where the code really
came from to maximize profit at all.  Even Debian Linux rejected it
and they are one of the least commercialized distributions.  Probably
less than even Slackware.  I suspect that Slackware and some of the
others have just not yet realized the true legal risk this new license
poses.  I know that we didn't for a while.  Our first thought's were
the same as the argument you are making.  Then after seeing how almost
every major distributor jumped ship over to X.org, I re-read the new
license several times trying to understand why they reacted it so
strongly before it finally hit me.

The big companies probably saw it first because they have the money to
have their lawyers examine it.

If it truly is only about the server, then they have accomplished very
little to nothing since servers generally don't have pop ups anyway.
If it is about forcing it on applications, as the list of changed
files so far would seem to suggest, then it is going to be nearly
impossible for any of the vendors to guarantee compliance.  Whether it
was intended or not, I think it is a legal trap that would force a
split whether there were other deeper seeded reasons for it or not.

The more I studied the 1.1 license, the more it appeared to me to be
specifically designed to cause this split.  David Dawes's own comments
also strongly suggest he is happy about it and wants to bring XFree86
back to the old days of being used primarily by developers who
download it directly from the XFree86 site, with little to no
commercial influence about end user features.  However, there is a bit
of a problem with that attitude.  Even most software engineers I know
of run on a commercial Linux, BSD, or other Unix like distribution.
Not just the end users.  Very few people actually download something
as big as X windows and try to install it them self.  Most people, no
matter how much of a guru they are, do not want to have to create
their own distribution every time they install or upgrade their
system.  I will be quite surprised if XFree86 does not quickly fade
into the background if they stick with this new license.