Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Should babbage be able to create it's own tables? #4

Closed
pudo opened this issue Dec 4, 2015 · 5 comments
Closed

Should babbage be able to create it's own tables? #4

pudo opened this issue Dec 4, 2015 · 5 comments
Milestone

Comments

@pudo
Copy link
Contributor

pudo commented Dec 4, 2015

Right now, this functionality is not in the babbage package, but rather in whatever host app uses it (e.g. cubepress, spendb). While I don't think that babbage should only be useable with tables that it has created itself, adding the necessary code to quickly generate a table (eg. from JTS) could make the package more convenient to use.

@pudo
Copy link
Contributor Author

pudo commented Dec 4, 2015

Alternatively, babbage could come to depent on jts-sql and the two might have some secret protocols for how to talk about star schema (#3).

@pwalsh
Copy link
Member

pwalsh commented Dec 4, 2015

jts-sql (which we'll update extensively, soon, BTW) would not suit here: in Fiscal Data Package, we want to go from mapping > DB tables, and not resources > DB tables.

Can you link to the relevant code in SpenDB?

I like that babbage is agnostic to the creation of tables: not sue ATM what that means though: perhaps babbage.fiscal-data-package could be extended for the DB loading functionality.

@pudo
Copy link
Contributor Author

pudo commented Dec 4, 2015

The relevant code in spendb quite literally is jts-sql (that's where the lib comes from) :) This is because in spendb the tables generated are a function of the data uploaded, not the model applied - meaning users can change the model around afterwards without requiring the data to be re-loaded.

Of course this is less 'clean' than a normalized star schema approach, but it's the simplest thing and the best UX -- wondering if we can somehow retain the UX of model mutability even as we drag the aggregator towards a cleaner data model.

@pwalsh pwalsh added this to the Current milestone Dec 7, 2015
@pwalsh
Copy link
Member

pwalsh commented Dec 7, 2015

@pudo ( cc @akariv ) just thinking out loud, but in the FDP to babbage scenario, it seems we'd want a star schema, and that the tables created should be more or less ( :) ) consistent across all datasets. I think the table creation therefore possibly does belong with babbage.fiscal-data-package.

@pwalsh
Copy link
Member

pwalsh commented Dec 15, 2015

So, I think the answer is no, and we can close this issue. We are currently creating our own tables in the FDP - > babbage loader codebase. Maybe when that is done, we can all take a look. Perhaps there will be a reason to pull the table creation code out completely into a standalone lib.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants