Persistence Best Practises
We made an interactive session on persistence
Plain SQLite
~40% of the audience use it
Toolchain, e.g. DB browser
No dependencies
Self made schema, under
Dev as full control, e.g. SQL queries
debuggable data(base)
CP and Loaders
A lot of boilerplate code
Reinventing the wheel (tools do the work already)
-(+) Need to write layer for CP, however close to CP with Cursors, etc.
no compile time checks
manual schema updates maintenance
SQL is another lang
SQL can get loooooong
Testablilty?
ORMs (Object/Relational Mappers
~10% of the audience use it
greenDAO (robust: around since 2011, fastest ORM) DbFlow () ORMLite Requery (quite new, beta, RX support)
Delta to SQLite
less biolerplate code
not reinventing the wheel (tools do the work already)
compile time checks
no to less sql
another lib you have to learn
potentially performance decrease (depends on the lib)
Realm
~ 4% of the audience use it
No ORM (?, technically it's an ORM to its own native database)
RX support
Docs
Tools
Others
~ 1% of the audience use it
Last updated