I’ve been working for some months on a program for processing some data, and it’s now at a stage where rather than displaying information (stored using ActiveRecord) via the command-line, I’d like to display the processed information via a Rails app.
The first question I’m facing is whether I should have the data processing and the data display in the same application, or in separate applications. I’m tempted to have separate applications because the less code there is in the rails app, the less the chance of security issues (some of the data processing code is a little ugly and insecure for optimization reasons).
The second question is how to handle database schemas, if I split it into a rails display app and an offline data processor. Is it enough for the rails app to merely be given a fully populated database, or should I be providing the rails app with migrations and/or schema.rb as well?
Edit: The data processing application is a parser for Wikipedia’s Get to Philosophy phenomenon. For the English edition, it takes a 20GiB XML file and looks at how ~ 6 million pages link to each other.
Can’t really say whether to split applications or not without more context.
Re the second question, you can work fine without schema and migrations, back in the Rails 0.5 days with did fine without any of these fancy schmancy features.
Just be sure that you have a solid database schema that’ll equally work for both parts of your app (or both apps, depending on which way you go)