I’ve grown a bit restless with Brimstone (the software that’s running this site) at the moment. I’ve been dribbling in features here and there to keep myself occupied, but I think it needs a bit of an overhaul both visually and in the background.
TODO
- Change the CSS and page structure from Bootswatch to Materialize I’ve never enjoyed doing my own CSS other than a few small bits, so this move is largely just for the visual appeal and a vector for removing JS
- Remove Admin Section reliance on JavaScript I’ve never enjoyed writing JavaScript either. During the restructuring I’m going to be changing the admin section to have minimal JavaScript, and elegant rollback. Currently the features that rely on JS are posting Notes and Replies, and favourites. Since this is the bulk of interactions with the software, I think it needs changing a bit. I like some shiny interactive features, but these shouldn’t be necessary to create particular forms of content.
- Condense Notes and Posts into single data structure The bulk of my content is written as Notes, which are a shameless emulation of Tweets and were actually the first type of content I produced. Posts are the longer, blog-post style, pieces of content. There’s literally no other reason other than my excitement and stupid categorisation of content that they need to be separate. I’ll do content inference instead; pieces with a title get published with an ‘Article’ template, and those without get a ‘Note’ template. With a location field, it means I can also infer check-ins as well as add richer data to all posts. This might not work for later content types such as exercise data as they don’t share many common fields. Syndication templates can also be produced.
- Restructure the URLs to content and implement redirects for old Tweet citations
- Feature to import all tweets between certain dates to create a backlog of content
- Implement a basic exercise tracker
- Implement a basic spending tracker
- Produce summaries of my content (heavily inspired by Rhiaro
- Implement proper user management to let people ‘install’ the software themselves, without editing files to create a user.
- Release the source on Github.
That’s a lot to do, and I’m thinking of dedicating an evening a week to the project. We’ll see how it goes. For now, you get bonus points if you understand the relationship between the title and the content of this post.