On Tue, Feb 16, 2010 at 10:33 PM, Andrew Dalke <[email protected]>wrote:
>
> > I am happy to draft some principles which address what an Open
> Specification is.
>
> "in the context of Blue Obelisk", yes? Since "Open Specification" and
> variations already have existing meanings which I have no problems with:
>
>
Yes, in the context of Blue Obelisk.
It is clear that there is a divergence of opinions and we are unlikely to
come to complete agreement. This is common in Open movements and has
happened in Open Source and Open Access. There is no completely agreeement
on Open Access and there is unlikely to be. But there are subfields of
practice that operate well and consistently. In this vein I think the search
for agreement on "Open Specifications" is likely to be difficult and
certainly will take time.
My position is that of someone who implements semantic chemistry aimed at
community use. It is freely given. It is based on the premise that semantic
interoperability requires:
* a formal specification that is implementable
* reference code to illustrate this specification
* examples (units tests) against which the code can be run and the
specification tested
* validation criteria, implementable by machine.
This is roughly the view that the W3C would take about their specifications
(recommendations).
Some people feel this is overkill for chemistry and that loose, mutable code
and specifications will allow greater innovation and interoperability. Some
believe that forking of specifcations and reference code is a good thing.
I'm not forcing people out of this view.
Note that specifications here are similar to language specifications (e.g.
Perl, Python). There is no support for mutating these and the community
relies on consistency for their compilers. I believe that chemical
specifications are effectively chemical compilers and should have the same
consistency.
I do not believe that trying to draft legal specifications for
specifications will be helpful. We have spent huge amounts of discussion on
Open Access and Open Data and the leagl approach has been of relatively
little use. What has been useful has been the development of community norms
and this is increasingly accepted in those working in OA and OD.
There *are* community norms for specifications in informatics - there are
also community norms in bioscience. There are few norms in chemistry and
that's part of what we are trying to find.
I'll think about this and return at some time.
P.
--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
------------------------------------------------------------------------------
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
