For Approval: Transitive Grace Period Public Licence, v1.0

zooko zooko at zooko.com
Tue Feb 17 23:32:56 UTC 2009


On Feb 17, 2009, at 13:51 PM, Russ Nelson wrote:

> Here's how the discussion went:
...

Was this discussion in a conference call, or in private e-mail, or what?

Thank you for summarizing it for the list.

> A licenses code under the TGPPL.
> B chooses to take advantage of the delay.
> C observes B's contribution to the world of proprietary software,  
> and structures their actions to receive source in less than 12 months.
> B decides to renege on his promise to release source and decides to  
> violate the license.
> A decides that he doesn't mind that B is violating his copyright  
> (copyright is not like trademark; you can enforce selectively.)

That's interesting!  I think you are right that this could be an issue.

Something similar can happen with the GPL, can't it?

A licenses code under the GPL.
B chooses to use the code, produce a derived work from it, and ship  
the derived work while stating they will release the GPL'ed source  
code of their derived work any moment now.
C observes that B is producing a product derived from A's GPL'ed work  
and structures their actions assuming that B will indeed release the  
source code.
B decides to renege on his promise to release source code and decides  
to violate the GPL.
A decides that she doesn't mind that B is violating his copyright.

In fact, I think that this pattern has been playing out over and over  
with the GPL over the past years, especially in the market for  
embedded devices.

I think you are right that this sort of thing could be a problem with  
the TGPPL.  With the GPL then C can at least wait to see B's actual  
code release before making plans (and wise practitioners currently do  
this with such situations), but the TGPPL's grace period makes it  
harder for C to lay long-term plans.

On the other hand, people don't usually plan around one-time code- 
drop events.  They plan around the ongoing evolution of a codebase.   
This makes the problem that much worse: B can release the first  
version of his derived work under the GPL, then A can release new  
versions of her original work under GPL, then B can release a new  
version of his derived work, incorporating both A's changes and his  
own changes, and decide at that point to stop satisfying the GPL  
requirements.  Even if C was cautious and waited to see the first  
GPL'ed code release from B before committing their strategy to that  
product, then C may be surprised to learn that B has the legal  
ability to continue to release new versions of that product while  
withholding the source.

So, while I don't think this sort of problem is unique to TGPPL, I do  
think that it is important and I thank you for bringing it to my  
attention.  I would be interested in further concerns that you  
(plural) have and also ideas you may have for ameliorating such  
problems, with GPL, OSL, or TGPPL.

However, what does this have to do with the question of whether TGPPL  
is Open-Source-Definition-conformant?


> The board is not clear whether this could happen or not.  Clearly  
> this it is risky for OSI to allow its trademark (which MUST be  
> enforced consistently) to be used on software licensed in this  
> manner.  More thought is needed.  It's possible that years of  
> experience with this license may be needed before it can be approved.


I have long since stated on the allmydata.org project's web site that  
the TGPPL is an Open Source licence.  (See http://allmydata.org .)

I made that statement because TGPPL *is* an Open Source licence, and  
it is important for people who are considering using the  
allmydata.org source code to know that fact.

If you want to protect the trademark in the sense of prohibiting  
people from using it without your permission, then you'll need to  
either give me permission or sue me (and the other people who have  
released software under TGPPL).  I don't think you'll sue me, and if  
you did I don't think you would win, because the TGPPL is conformant  
with the Open Source Definition.  If you want to protect the  
trademark in the sense of prohibiting people from incorrectly using  
it on OSD-non-conformant software, then it would probably help rather  
than hinder your cause if you would go ahead and bless the use of the  
trademark on a conformant licence which its author has patiently and  
thoroughly submitted to your established Review Process.

Please include the Transitive Grace Period Public Licence, v1.0, on  
the list you maintain of Open Source Definition-conformant licences.

Thank you for your time.

Regards,

Zooko Wilcox-O'Hearn
---
Tahoe, the Least-Authority Filesystem -- http://allmydata.org
store your data: $10/month -- http://allmydata.com/?tracking=zsig



More information about the License-review mailing list