First Quixote Conference - 22nd-23rd March 2010

A meeting to create the first Open distributed repository for electronic simulations

Sponsored by the Hartree Centre

For a general overview of the Quixote Project, please follow the links on the Front Page.

Meeting's EtherPad is at http://okfnpad.org/quixotemeeting.

• Primary: build and populate a Chempound repository for outputs from electronic codes. Repository should search for chemical concepts, job meta-data and core physico-chemical properties.
  

• Subsidiary: build Core CML-XML parsers for 3- 8 commonly used codes in atomistic QM and planewave simulations. Use Lensfield to scrape directories for files.

• Chemistry (not QM methodology).
  

• practical achievement (running code and a working repository).
  

• Opensource/opendata without access control to the repository. Repoistry-agnostic (?SWORD,?OAI)

The meeting will have two main strands:

• use case: searching the repository for given concept over several filetypes
  

• preparation: several CML files (with index data) for several codes
  

• Files for data set (final CML, original TXT, index.rdf, etc.) aggregated under ORE
  

• Aggregation uploaded by SWORD (METS, ZIP, etc.. as agreed)
  

• index in simple name-value search (?RDF, ?Mongo)

• Use-case: disk populated with calculations of given kind where (say) suffix = FOO .
  

• Arbitrary directory structure.
  

• Lensfield trawls disk for *.foo
  

• JumboFooConverter/Parser converts to CML (failures => null files).
  

• conversion supported by FooDict.cml

The parsing need not be complete. The CML is used for specific data and indexing. The results COULD (if all goes well) be fed into the Chempound repo.
Small Parsers can be written on-the-fly for codes if the format is pre-explored. We do not have to do the whole job.

• The codes should all have mini-dictionaries (some will be large)
  

• The indexes will be normalized (e.g. dates, units, formulae, etc.). This means that users search a single conceptual model but the original data is not affected and can be retrieved as part of the ORE aggregation.

• Lensfield2
  

• JUMBO
  

• Moka
  

• Avogadro
  

• Chempound

NB: - this is a draft timetable and subject to change.

Monday 21st March
Arrival at Daresbury Park Hotel - check-in from 15:00
Tuesday 22nd March
Time	Activity
9:00	Taxis from Hotel to Lab
9:15-9:30	Tea/Coffee
9:30-9:50	Introduction to Conference and Quixote Project - Jens Thomas
9:50-10:10	Q5cost, a possible common format for Quantum Chemistry - Elda Rossi| q5cost.pdf
10:10-10:30	TURBOMOLE and CML - Semi-direct vs. parser approach - Andreas Gloess| talk.pdf
10:30-10:50	Avogadro & Kitware - Marcus Hanwell
10:50-11:10	Quantum Expresso, Moka & XML - Riccardo Sabatini
11:10-11:25	Tea/Coffee
11:25-12:30	Introduction to parsers, dictionaries, CML and databases - Peter Murray-Rust et al.
12:30-13:30	Lunch
13:30-15:30	Work on parsers/database
15:30-15:45	Tea/Coffee
15:45-18:30	Work on parsers/database
19:00	Dinner at the Ring O'Bells
Wednesday 23rd March
Time	Activity
8:30	Taxis from Hotel to Lab
9:00-9:30	Tea/Coffee
9:30-11:00	Work on parsers/database
11:00-11:15	Tea/Coffee
11:15-12:30	Work on parsers/database
12:30-13:30	Lunch
13:30-15:00	Work on parsers/database
15:00-15:15	Tea/Coffee
15:15-16:30	Work on parsers/database
16:30-17:00	Summing up

The meeting will take place at STFC Daresbury Laboratory.

The closest airports are Manchester and Liverpool.

For invited attendees, accommodation has been arranged for the nights of the 21st and 22nd at the Daresbury Park Hotel

For further information about the meeting, or if you are interested in attending, please contact Jens Thomas: jens.thomas@stfc.ac.uk