so I thought I would fix ecto_mnesia up so it would work for Ecto 3. The adapter used to support migrations in the past. Given that Ecto was split up into two packages and I need stuff from there to make the migrations work, I am wondering should the migrations support be dropped, should I add ecto_sql as a dependency or should I copy the needed code from ecto_sql to ecto_mnesia?
So what do you think? Am I just wasting my time or is anyone interested in the adapter? And what about the migrations?
You already have Ecto3 support for Mnesia:
Afaik that migrations are not done yet, maybe you want to join efforts with that author?
lol thanks. I guess I wasted my time then. and yes, I can talk to them
Well its never wasted, you learned something in the journey for sure.
You can evaluate what its done already here:
And you may want to speak with @pascal to see if both of you can unite synergies?
I am using Memento:
For migrations I am rolling runtime migrations on the fly with versioned schemas. So each time a hit Mnesia to read from schema v1 I will upgrade on the fly to schema v2.
Hi there, with the original author, we’ve brought many updates to this adapter, which now support:
- multiple primary key
- (to-be-merged) support for foreign keys
- table creation helpers (migrations like)
Your feedback is welcome !