https://github.com/saulshanabrook/cerebral-orm
https://github.com/saulshanabrook/cerebral-orm
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/saulshanabrook/cerebral-orm
- Owner: saulshanabrook
- License: mit
- Created: 2016-07-06T11:59:40.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2016-07-06T12:16:26.000Z (over 9 years ago)
- Last Synced: 2025-06-20T13:11:02.705Z (4 months ago)
- Size: 1000 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# cerebral-orm
Cerebral is great, but it's not enough. It would be great to get more
features for building rich client side applications out of the box. In my mind,
these include:
* Offline application, that will sync when it gets online. Like
Google Keep on your phone. The user of the application shouldn't
have to worry about what is local and what is synced. It should only
present itself to the user, if it can't sync for some reason.
* Typed and valid state, at all times. In the backend, we have SQL to
help us make sure we can't save invalid data. Now, we have a mix of
runtime validation (tcomb), compile time type checking (typescript),
and data validation of state (action types). Instead, we should
be able to say "All Users have these fields. If it does not, it is not valid"
* Easy to implement relationional data. Right now, we have to think *a lot* when
about how to store our data (indexed by key), how to retrieve it, and how to
reference other data. We have different methods for retrieval and for updating.
We have no standard specification for how to achieve this.
Overall, I want to make an idiot proof system. Or at least, not require so much
from the developer. Cerebral is a solid base, it's time to start building on top.