I'm taking a breather from working on openPDS. Here are a few notes on what we've been working on and a roadmap for going forward:
What we're doing (as of December 2, 2014):
Separating database requirements, so that files don't have to import sqlite3, pymongo, and psycopg2 all at the same time. #25 #27
Related to the above: The requirements.txt file should be auto-generated by start.py #25 #27
Decoupling the project from virtualenv. Seriously - people should be able to store their virtualen environments anywhere they want. #29
Testing. This should exist. #39
These are mostly derived from @bradmontgomery's #22
Things that I would do next:
(This is obviously subject to interpretation and debate)
make the code confirm to pep8 standards.
I'd remove access control, and replace with a user login. Contentious, I know, but access control creates some complications, and I don't think that the project is ready for it.
-- at the very least, access control should exist as a separate django app.
Remove visualizations and meetup
Create a connector to datahub
datahub notes
This is for the creation of a datahub backend, through which users would be able to directly access and edit their data.
Create an openPDS user on DataHub.
Automatically create accounts and openPDS repositories for new users, including passwords.
Create openPDS repositories for each new account
grant select, create, insert, update, delete, to openPDS
send the user the username/password, remove it from openPDS
Hi there,
I'm taking a breather from working on openPDS. Here are a few notes on what we've been working on and a roadmap for going forward:
What we're doing (as of December 2, 2014):
These are mostly derived from @bradmontgomery's #22
Things that I would do next:
(This is obviously subject to interpretation and debate)
datahub notes
This is for the creation of a datahub backend, through which users would be able to directly access and edit their data.