> > Found this Was having trouble but i guess this --> > https://docs.google.com/viewer?a=v&q=cache:v2lgkygxA_cJ:www.packtpub.com/sites/default/files/5467OS-Chapter-3-Database-Abstraction-Layer.pdf+&hl=en&pid=bl&srcid=ADGEESgEeuJ9YUV1Lwvy1CubPMg-92-ZM0Rtdvc14RQhuwGE32cSVmd-p35Tw21hW2R4PB9UcXbXLwTVrGRcthvusLvcdRDhnz8B2g-LsohKy64TyJcMX0yuCw5qiypnsPSxUyWHHoPT&sig=AHIEtbQPUlFmojpWCTBBJKHdwdR34aZeuw > > > explained it. DRY is broken however, which isn't a good thing as the > database gets bigger. >
The DAL does not inspect the database -- you have to tell it about your models for it to know what's in the database. This is necessary because the models can include many attributes that are not part of the database, such as defaults, validators, computed fields, etc. DRY is not broken. If you need to share models across applications, you can put them in a module and import them (only Linux, you can also use symbolic links to model files). If you don't need all the DAL-specific attributes of the models, you can also auto-import<http://web2py.com/books/default/chapter/29/6#Using-DAL-without-define-tables>the model definitions from another app via: db = DAL('sqlite://storage.sqlite', folder='path/to/app/databases', auto_import=True)) Note, if multiple apps are sharing a database, make sure for any given table within that database that migrations are being handled by at most one of the applications (i.e., for each table, set migrate to False in all but one app). Anthony
