Having done some weekend projects, you get it done in a weekend with usually two approaches:

1. Use "do it all for you tools" that do all the hard parts (e.g. Ruby on Rails+lots of gems, Django+lots of libraries, Drupal/Joomla + plugins, etc)

2. Cut scope to a huge degree (don't have any of those things, no users, just "if you make this request then X data is stored and you get redirected to a view").

This math-oriented pastebin was probably the second.

Almost to a weird extent, this is the only software that most of us will ever be able to write, code that we could write the MVP for in a day. Because anything else is too hard to see to completion. There's something to be said about going for that MVP and then iterating.

You need to focus on a single feature and relegate other issues either to existing services or for the next iteration.

Not "fully fledged" but "enough to be useful". After that you start having feedback that can guide gradual improvements.