Today Etienne, Ed, Rufus, Mark, Peter and I met up to start the first sprint of the new year. We began by clarifying the purpose of the sprint, today’s agenda and the project overall. We stated our aims as follows: we are not trying to re-do what is already available online, we are not getting into the detail of normalisation or disambiguation within a centralised database, and we are not intending to alter the academic culture overnight; however, we are going to improve the BibJSON facility for wider use, we are trying to determine how we can get more small groups and individuals involved, and we are identifying compelling, essential and simple reasons for people to support the project at this early stage before the ultimate global benefits can be realised. With this in mind, we got cracking.
Etienne, the newest member of the team, began coding pretty much straight away – he and Ed started working on MARC / RIS parsers. Peter and I started a huge list of FAQs for the website – Peter asking in-depth questions such as ‘do we want to create a single BibJSON collection for all the world’s metadata?’, me going for slightly less detail with ‘what is metadata?’ – and Mark assisted where he was needed. Peter hacked some datasets to go into the parsers and Mark got some coding done too. There was good progress made and we are set up well for tomorrow to crunch some issues.
One problem to be revisited is in relation to BibJSON, in having copies of the same record within different collections. If an object (the record) is held within multiple collections, there are separate copies of that object which could cause problems – for example, if a record is copied into several collections and then a typo is found, it can be a mammoth task trying to track down all erroneous copies and correct them… This issue is likely to be solved by creating Master / Slave relationship between copies. Also with regards to BibJSON, Etienne suggested providing a flat HTML version of collections (in addition to the javascript option), for easy use in departmental web pages.
All set for day 2…
Pingback: Unilever Centre for Molecular Informatics, Cambridge - BibSoup! A new OPEN approach to managing personal and group bibliographies « petermr's blog