Archive of UserLand's first discussion group, started October 5, 1998.

Re: Guido and Richard

Author:A.M. Kuchling
Posted:9/8/2000; 9:04:14 AM
Topic:Guido and Richard
Msg #:21010 (In response to 20987)
Prev/Next:21009 / 21011

If there's any bogeyman here, it's CNRI management for requesting a license change in the first place. Chronology of events:

* Versions of Python up to 1.5.2 have an X11-inspired and GPL-compatible license that the community is happy with.

* In May 2000, GvR and 3 other developers (the Pythoneers) leave CNRI for BeOpen.

* Preparations continue for Python 1.6, now renumbered 2.0, aiming for a beta release July 1.

* A few hours before the 2.0b1 release, CNRI says "No, not without a different license." *Why* CNRI thinks a new license is needed has never been made clear.

* Weeks of negotiation follow between CNRI and BeOpen; happily, in the meantime development on the code continues.

* CNRI insists on a 1.6 release that contains the Python CVS tree as it was in May, when the Pythoneers left. (Python 2.0 contains the development from May until today.)

* CNRI's new license is longer and has various added terms, but doesn't really restrict what you can do with the code.

* BeOpen asks Stallman for an opinion on the GPL-compatibility of the CNRI license, because the Pythoneers think it's important. Stallman points out some problems; all but one of them are resolved by changing the CNRI license.

* One questionable term remains: CNRI and BeOpen think the term doesn't affect GPL-compatibility; Stallman and the FSF's lawyer think it does.

* The Pythoneers finally issue 2.0b1 on September 5. 2 months of delay have resulted from this licensing struggle!

* Someone raises the issue of the Python license on the debian-devel list. LinuxToday picks up the story, then Slashdot, and now Scripting News.

Note that RMS has not issued any broadside against Python, since he's perfectly aware that negotiations are still ongoing; this whole tempest came from the single debian-devel posting by someone else. This isn't an example of RMS being inflexible; instead it's an example of how clueless management can severely hamper technical work through legal wrangling and lack of understanding of the user community.



This page was archived on 6/13/2001; 4:56:34 PM.

© Copyright 1998-2001 UserLand Software, Inc.