From OpenSQLCamp

Contents 1 Completely vertical versonable schema 1.1 why (my needs) 1.2 advantages 1.3 disadvantage 1.4 how

Completely vertical versonable schema

why (my needs)

• data comes in from many sources
  • varying levels of trust
  • trust can change at a later date
  • need to have the ability to slice out entire sources
• different ways to look at the data.
  • I need to be able to create complex views nearly on the fly.

advantages

• source every 'cell'
• trust for every 'cell'

disadvantage

• not a 'normal' looking schema
  • requires traning
  • helps to abstract the crazyness (code)
• slow slow slow
  • requires the use of summary tables to be useful (not the worst thing but yet another layer)

how

• in short blow everything apart
  • all data stored in vertical tables.
  • build relationships of two data points, by source and source instance, with trust and timestamps