Friday, January 15, 2010

Warren DeLano and the future of PyMOL

This blog is old and new news. The old news is that Warren passed away at the end of last year, after having successfully shown how OpenSource cheminformatics (and/or bioinformatics) software can be developed in a commercial setting (DeLano Scientific), and PyMol was a huge success. Warren had a SourceForge account (wdelano) for almost 10 years:

I had not blogged about it before as the news hit me hard. Surely, Warren knew a lot of people and I only was only one of many, but Warren's memory sticked well. I know Warren from the Jmol project, where we talked in the past of coming to an Open Specification for exchanging scenes between Jmol and PyMol. Around the end of my PhD contract we even briefly, but seriously, explored doing a post-doc in his group.

Anyway, lot's of people wrote up blogs (in arbitrary order: Rich, P212121, MacResearch, Jörg, MMB, Shirley, Derek, Wavefunction, Dan, Barry, and probably many more). They have set up a memorial fund which will focus on promoting the Open Source ideas of Warren, including an Award.

Yesterday, I was pinged about Schrödinger acquiring PyMol. The press release is, as usual, short on details, but those have become clearer during the day. Schrödinger is not new to Open Source cheminformatics, and has an product based on KNIME, which is now GPL, but also has a proprietary license for those who wish to license so.

But, unless I missed any other Open Source (-oriented) product, the acquisition of PyMol significantly changes the game for them: PyMol is a major Open Source product, bigger than KNIME at the moment, I'd guess. My immediate response to the acquisition is whether they acquired copyrights, and they did, according to this commit:

This is important as it puts Schrödinger in charge of license changes. Fortunately, they seem rather serious about the Open Source thing, and hired an active PyMol developer (Jason), and kept the existing Open Source license:

Therefore, congratulations to Schrödinger for getting seriously into the Open Source community, making them the next Dr Who of PyMol, and congratulations to the family of Warren in ensuring continued development of the PyMol project! It's hearth-warming to see that despite the bad times they are going through, and all they options they had with the PyMol code base, they find time for and strength in supporting Warren's ideas about the future of cheminformatics. My thoughts are with them!