First, a feed reader

A basic feed reader is pretty straightforward: fetch RSS and Atom feeds and display them. That’s it.

Marking items read is optional: Facebook timelines just display them in (kind of!) reverse-chron, and you just stop reading when you run into something you’ve already seen. (Facebook sabotages this by shuffling things around when, I guess, they have new comments on them, but that’s a different issue.)

Being able to add and delete feeds is optional, if it allows reading an OPML file.

Parsing feeds can be an exercise in edge cases, but luckily there’s already XML::Feeds so that work has been taken care of.

Second, a blog publisher

Publishing a (static) blog is an easy next step: the feed reader is already formatting posts. The publishing side just has to read entries from something other than an XML file. I’m going to start with Plerd-style markdown files, because that will allow me to start writing this blog with the intention of being able to import it.

Third, blog comments

I haven’t decided how to implement this yet. Webmentions are a possibility, but so would be to start work on the default browser client. Of course, the most basic method would be to have a simple HTML form that does a POST, but without some kind of authen/authz system that would be pretty open to abuse. We’ll see.


Comment? Email it to me. (I'll assume I can publish it unless you say otherwise)


Next post: Building a minimal feed reader

Previous post: A first post on Wirebird