Re^3: Moose + ORM

Indeed. The idea would be to allow the client to specify how to "Describe" the database table as well use another RDMS.

But these things are not important right now and well past the scope of this discussion. This is not an RFC ... but rather a (hopefully) simple question whose answer will allow me to get to the point of submitting an RFC. =)

See Orion? ;)

Comment on Re^3: Moose + ORM

Replies are listed 'Best First'.
Re^4: Moose + ORM by Anonymous Monk on Jul 31, 2009 at 19:05 UTC
I like where you are headed. The basic problem which this addresses is Moose classes and database schemas getting out of sync. Unfortunately, it assumes that the database schema is always right, which is not necessarily true. I think the ideal solution is to make a Role that optionally injects attributes, or compares attributes and schema. I see the problem that in a group, sometimes people make changes to code and database, and there is no easy way to check whether the ORM mapping is valid. I had planned to do something with SQL::Translator to handle this, by making a Parser that can parse Moose class hierarchies. (To be honest, s/I had planned/It sure would be nice to have/g)	[reply]

Replies are listed 'Best First'.

Re^4: Moose + ORM
by Anonymous Monk on Jul 31, 2009 at 19:05 UTC

The basic problem which this addresses is Moose classes and database schemas getting out of sync.

Unfortunately, it assumes that the database schema is always right, which is not necessarily true.

I think the ideal solution is to make a Role that optionally injects attributes, or compares attributes and schema.

I see the problem that in a group, sometimes people make changes to code and database, and there is no easy way to check whether the ORM mapping is valid.

I had planned to do something with SQL::Translator to handle this, by making a Parser that can parse Moose class hierarchies. (To be honest, s/I had planned/It sure would be nice to have/g)

[reply]