Developer Meetings/2009-1

From openZIM
Jump to navigation Jump to search
Naming
  • the project's name is "openZIM" which means we provide an open source implementation of the "Zeno IMproved" file format
  • Decision: the file format will be renamed to "ZIM" to distinguish between the Zeno file format from Directmedia
Indexing and Searching
  • several possibilities can be implemented:
    • internal search index
    • external search index
    • using an external library such as Zeno, Sphinx, Xapian...
    • Decision: will be postponed until next developers meeting which will focus on that
Encoding
  • article texts in UTF-8
  • URLs in UTF-8
  • latin1 auto detection will be kept for backwards compatibility
  • problem: sorting multibyte characters
    • there should be libraries which are able to sort multibyte characters, including them
    • dropping QUnicode library
Namespaces


Header / Templating
  • discussion about how articles should be stored and viewed - some viewers need a dynamically created template (eg. web applications)
  • the ZIM creator should be able to create his own templates, CSS, javascript files etc.
  • Decision: the ZIM creator stores a template file inside the "-" namespace
  • the reader application can decide if it uses the template file to build the layout, but it is highly recommended to do so
  • Decision: the template article name is stored in the header, as well as the main page article name
Infrastructure