xldenis / dynamit Goto Github PK

Infinite scroll is too aggressive in it's loading of content pages.

API needs to be restructured so that it will be more restful and place it's information in a better format.

Login flow would attach user id to redirect so that JS can know which user is logged in.

This is not optimised yet, but we are considering fixing it in the future.

COntent workers need to be worked on, so that they will fetch more data and keep the user's stream updated with the latest data.

Explicit user input needs to be accounted for so that we can check ourselves before we wreck ourselves.

Make a generic template that every posts follow.

We need to switch to postgres for our database and optimize our models to fully exploit the result of that (ie polymorphic source posts)

Sources need to be individually fetchable. The primary use of this would be in the search function. ie search 'twitter' -> only load twitter

Logging in from a user different than the first simply loads the first users account.

Rework score system. Current implementation is completely inadequate.

Instead of checking the log files to see what are the error on the server, find a way to make them visible more simply. Multiple solutions are possible :

• Launching two terminal at the end of the make file (but would not work on a server)
• Make a webpage that displays the content of the log files

Possibly slow down search, find a way to show so immediate results while the rest are still searched for.

Need a UI to implement preferences / Source management so that users can log in . out of various accounts.