#quixotechem
I’m delighted to see that a major computational chemistry program (NWChem) has been released under a fully F/OSS Open Source software licence. There are many programs (“codes”) in compchem but few of them are F/OSS. The norm is either to be fully commercial, or allow free-to-academics. The main exceptions I knew about already were http://en.wikipedia.org/wiki/ABINIT and http://en.wikipedia.org/wiki/MPQC ; the former deals with solids rather than molecules so is outside the scope of Quixote. This means we can now – in principle – create completely open distributions for a community project.
Sujet : CCL:G: NWChem version 6.0 (Open Source) released
Date : mercredi 29 septembre 2010, 21:38:09
We are pleased to announce the release of NWChem version 6.0. This version
marks a transition of NWChem to an open-source software package. The software
is being released under the [Educational Community License 2.0] (ECL 2.0).
Users can download the source code and a select set of binaries from the new
open source web site http://www.nwchem-sw.org
New functionality, improvements, and bug fixes include:
* Greatly improved memory management for TCE four-index transformation,
CCSD(T), CR-EOMCCSD(T), and solver for EOMCCSD
* Performance and scalability improvements for TCE CCSD(T), CR-EOMCCSD(T), and
EOMCCSD
* TCE based static CCSD hyperpolarizabilities
* New exchange-correlation functionals available in the Gaussian DFT module
* Range-separated functionals: CAM-B3LYP, LC-BLYP, LC-PBE, LC-PBE0, BNL. These
functionals can also be used to perform TDDFT excited-state calculations
* SSB-D functional
* Double hybrid functionals (Semi-empirical hybrid DFT combined with
perturbative MP2)
* DFT response are now available for order 1 (linear response), single
frequency, electric field and mixed electric-magnetic field perturbations
* Greatly improved documentation for QM/MM simulations
* Spin-orbit now works with direct and distributed data approaches
* Plane-wave BAND module now has parallelization over k-points, AIMD, and
Spin-Orbit pseudopotentials
* Plane-wave modules have improved minimizers for metallic systems and
metadynamics capabilities
* Bug fix for DISP: Empirical long-range vdW contribution
* Bug fix for Hartree-Fock Exchange contributions in NMR
Please let us know if have any issues accessing the new website.
Best wishes,
Huub
__________________________________________________
Huub van Dam
Scientist
EMSL: Environmental Molecular Science Laboratory
Pacific Northwest National Laboratory
902 Battelle Boulevard
P.O. Box 999, MSIN K8-91
Richland, WA 99352 USA
Tel: 509-372-6441
Fax: 509-371-6445
Hubertus.vanDam : pnl.gov
www.emsl.pnl.gov
——————————————————-
——————————————————————————
Start uncovering the many advantages of virtual appliances
and start using them to simplify application deployment and
accelerate your shift to cloud computing.
http://p.sf.net/sfu/novell-sfdev2dev
_______________________________________________
Blueobelisk-discuss mailing list
Blueobelisk-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
|
|
|
|
Reply to all
The other major Open Source program besides MPQC and ABINIT is PSI3 (http://www.psicode.org), however, it’s development is quite closed, they only release their tarballs and have no code repositories, mailing lists or forums. Still, PSI3 has a couple of unique (for Open Source programs) and competetive (with respect to state-of-the-art closed source codes) features.
Also, one cannot praise MPQC enough, they (mostly Prof. Valeev) commit their code openly to their subversion repository at least once their papers get accepted, if not before. And Valeev is one of the leading persons in couple-cluster theory (and also major driving force behind psicode)
Uh, needless to say I am totally thrilled that NWChem went Open-Source (I hope that modified patent grant compared to the Apache-2.0 license is alright with all/most GNU/Linux distributions and not just OSI), and will see about pushing it to Debian/Ubuntu right away once Oktoberfest has finished. Thanks for sharing the information.
You should look at both MADNESS and GPAW as good examples of GPL codes in chemistry and material science. MADNESS is rather rudimentary in functionality but the computer science aspects – designed by NWChem progenitor Robert Harrison – are impressive.
MPQC is great but lacks a manual, user or developer list and one cannot have a reasonable expectation of support from the developers. I am a huge proponent of MPQC, but I cannot recommend it as a replacement for Gaussian, NWChem, GAMESS, etc. because of the lack of documentation.
I disagree with your analysis of PSI. Their SVN is not published on the Internet, but it does exist and they are willing to accept external developments if you have something to contribute. Have you approached them about contributing or are you just concluding they are closed based upon their web presence?
I think you are crazy for trying to get NWChem in Ubuntu (have you looked at the build system?) but I will try to help you if you’re serious. Emailing the user list is the best way to start this conversation.
Finally, I am a huge fan of Ed’s but his unrivaled authority is in integrals and explicitly-correlated methods. While he knows CC well, it is unfair to others to say he is a leader in the field.