What does a FOSsil Bank item look like?

Posted on 12 August 2014


I’ve started thinking – and talking – about what a new and improved FOSsil Bank might look like. A necessary consideration is the existing data.

There’s less than a megabyte of it, but it represents 1,646 different items. 1,340 of those are “works”, “sources” or “entries” – that is to say, creative works and those who create, modify and publish them. Each of these types has a different template, but they’re pretty similar in terms of content. In fact, the “entry” type is a consolidation of the “source” and “work” types, which are now deprecated.

This is an example entry, entry_green-comet.txt:

intro: ‘A sci-fi novel.’

site: ‘https://unglue.it/work/128673/’

general-details: “Gratis – or make a donation to support the author.\n\nEPUB, PDF. DRM free.” (ed: “\n\n” represents a page break)

sources: ‘[/source:unglue-it unglue.it]’

category: ‘[/category:literature Literature]’

genre: ‘[/genre:science-fiction Sci Fi]’

licence: ‘[/licence:by-sa CC BY-SA]’

copyright-details: “Uses an anomalous CC declaration: \n\nPublished under Creative Commons\nAttribution and Share-Alike\ncc-by-sa 2012” (ed: “\n\n” represents a page break)

code: green-comet

Two things are missing: the entry’s title (in this case, “Green Comet”) and its tags (in this case, “by-sa, literature, science-fiction, unglue-it, work”). I’m not sure where these are stored – it may be that they are not retrievable.

Tags mostly overlap with details already in the text. There are a few exceptions – for example, I sometimes forget or cannot be bothered to put a genre in the text, and only put it as a tag. The main other exception is that whether an item is a work or a source is identified using a tag.

In some cases, it is also identified in the heading of the text file, but for items of the entry type it is only found in the tags.