sqlalchemy-bot

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * changed **milestone** from "fasttrack" to "tier 1"

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** not a bad idea but the irony is that while we are all about migrating databases, changing the structure of the alembic_version table itself is tricky....

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** maybe have this be enabled by a flag, "store_version_history", indicates the alembic_version table will be generated with additional columns up front. then have some other system...

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * added labels: **versioning model**

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * set **milestone** to "tier 1"

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** the API is not public right now and is hardcoded to how the current Boolean / Enum type work. this is the check: ``` # detect...

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** something has to be changed on the SQLAlchemy side for this to be nicer so lets at least document the workaround for now.

**Adrian ([@thiefmaster](https://github.com/thiefmaster)) wrote:** Thanks, I went with the second solution, just slightly different to be more flexible in the future (in case I ever want it for other things than...

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * changed **title** from "Incorrect constraint in new table when custom type" to "document this workaround for custom SchemaType + p"

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * added labels: **autogenerate - rendering**