On Feb 16, 2010, at 8:39 AM, Egon Willighagen wrote:
> http://blueobelisk.stackexchange.com/questions/231/what-formats-fall-into-open-specification
I've updated that with a list of structure file formats which I consider open:
- SMILES
- The MDL connection table formats
- NCBI's ASN.1 definition for structures
- PDB
- SMILES
- SMD
(Plus one, "CEX", of which I can find very few traces of.)
I think the SMD case is very interesting, from several viewpoints. It's an
effort from a number of different people/companies to make a new connection
file format, based on the Molfile format.
By no means do we claim that the SMD format presented here covers
all information explicitly that will ever be used in computer
chemistry. But it is designed in an entirely modular form, thus
permitting definition and addition of completely new information
without affecting the existing structures. This unique feature
provides an extreme flexibility that ensures a nearly unlimited
upward compatibility with respect to future developments.
It was an open definition, with input from companies and individuals. You could
get the spec (by sending a request by mail - I assume the spec was published as
well). There were plans to start a more permanent caretaker organization, which
would be in contact with the national and international standards organizations.
It didn't go anywhere.
On question I have for those BO people developing the concept of "open
{standard,specification}" is: would the SMD specification be considered open?
It had a spec, it was published, different people were involved, there were
plans for an organization to oversee the spec. But they disappeared, along with
a number of other proposed structure file formats.
If it's not open, why not? Was it ever open? What made it closed?
Another question I have is what difference would a clear definition make? Would
the OB people stop supporting SD files? Or would someone sit down (perhaps
clean-room like) and write an "OpenCT" definition which is independent from the
MDL format description?
(Since that's the bit of magic which seems to make OpenSMILES acceptable -
despite the lack of any implementations!)
Or, my hope is that the OB people agree with my statement that SMILES and the
MDL connection table formats specifications are open; or at least give me a
clear statement why they aren't.
Andrew
[email protected]
------------------------------------------------------------------------------
SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
http://p.sf.net/sfu/solaris-dev2dev
_______________________________________________
Blueobelisk-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
