In April 2005, I registered a domain name. I had a plan, and it was an ambitious one. I had read an O’Reilly book on XML (cover-to-cover) a couple of years beforehand, and it did something to my brain and I thought I could see through time. Weblogs were ascending fast, the web was exciting, and hoo boy did I think RSS was a neat idea!
The first manifestation of this was my first personal blog from around 2001, Minimalist Dong. This was both the name of the software, and the blog. Yeah, we used to roll our own blogs like cavemen, back in the day. We didn’t just get someone to host WordPress for us.
Here’s a screenshot of Minimalist Dong. It’s bare, and minimalist (by name, by nature), and does not have a lot going on.
See the sidebar though? Those other “sections” are just directories, with another copy of Minimalist Dong in them, and they’re generated from RSS feeds!
The thing about Minimalist Dong was that it wasn’t strictly just blog software, it was an RSS reader. Each post was a plain text file with some minor formatting information in it, and the PHP script would assemble them on load and (depending on how it was called) render that as HTML or … an RSS feed. Same code, same source.
So that was a fun diversion for a bit and then I lost interest after getting bogged down on the details of how to handle HTML in post bodies when reading feeds that I didn’t generate. There was a problem in the early days of RSS, you see. The format of the feed was well defined, but there wasn’t much consistency in what was actually enclosed in that feed. I’m sure things are much better now, right?
Right? Oh no.
Anyway … instead of having a good think, separating out style from content in my code, and inventing Markdown way too early in the timeline, the ADHD took hold as soon as it got hard and I moved on to other things, and stopped updating Minimalist Dong sometime in 2002.
Hubris
I still had all this in my head for a while though. It came to a head again in early 2005 when in a fit of impulsiveness I registered the domain feedtube.com and started work on the next iteration. I had this thought that since everything had an RSS feed, and we had some good aggregators and readers then why didn’t we just combine all the feeds we read into one “stream” or something? Then you could treat that one single stream as a feed in its own right, and re-publish that mixed up version as a new feed, and …
Yeah my brain went off on a wild ride here, and I’m not claiming to have originated anything previously unheard of. I was just reading a lot of stuff about what was going on, and other people definitely had these ideas too, as we’ll get to in a second.
I was just one guy with a domain name and a dream. I started writing some code to see what’d happen. Of course, once again the ADHD took charge and I ended up starting on a generalised, abstracted PHP framework to support this project, which in a moment of heightened self-awareness I named Hubris.
Hubris did not go anywhere. I abandoned it after a while when I realised that I was obsessively abstracting functionality to an extreme degree without actually writing anything that … did anything. If this thing had ever seen the light of day, using it would have involved writing a 10,000 line configuration file and calling a single function called do_everything().
It’s for the best we never went there.
Those Yahoo pricks
While Feedtube was technically a thing I considered myself to be working on for a while, it all went in the mental recycling bin in 2007 with the announcement and release of Yahoo Pipes. This is what I meant that I was not the only one thinking along these lines, I was just a random dude with no contacts or executive function who read stuff. Over at Yahoo they had people getting paid not to give up.
Here’s something from the initial description of Yahoo Pipes.
Pipes’ initial set of modules lets you assemble personalized information sources out of existing Web services and data feeds. Pipes outputs standard RSS 2.0, so you can subscribe to and read your pipes in your favorite aggregator. You can also create pipes that accept user input and run them on our servers as a kind of miniature Web application.
Pipes was launched with a description that pretty much matched exactly what was in my head. I wasn’t shattered or annoyed, I knew I’d slept on the idea and hadn’t even made any attempt to discuss it with anyone really. If anything, I was keen to use it and see how it went because obviously they had the resources to do it right, better than I ever could, and the entire world was going to jump on this now. The golden age of RSS everwhere was here, finally!
You remember Pipes, right? No? Oh. I guess it didn’t change everything after all. Well, maybe they did it wrong or something.
I wish I could tell you here what was wrong with Pipes. If there was a target audience for this product, it was me right? I had been looking for this thing for years, I wanted it. I tried to build it, so I’d be there day one with all my opinions and takes.
I never used it. Never logged in. Just thought “oh well, no point writing that now,” and moved on with other things. In 2015 they announced it was being shut down and it was the first time I’d even remembered it had been started up since that announcement.
Back on my bullshit
It’s now 2025, I’ve owned feedtube.com for 20 years and never really done anything with it. I tried out running a clickbait article site for a bit but I hate reading that kind of content so unsurprisingly I didn’t want to write it either. The main useful thing it did for a while was running an RSS to ActivityPub gateway, but the software I used for it wasn’t something that I liked working with, and I really didn’t like how it looked. Javascript isn’t really my jam either, so I didn’t want to get stuck into fixing it myself.
I don’t mind a bit of Python, however, and I’m feeling a bit rusty on that front … so I thought why not take another shot at this. I’d done some Django work previously, and it suited the way I thought and the way I worked.
So now this has happened. We have reached the point of a functional prototype of an RSS reader. It only took two decades!
What now?
Where’s the source Mike? Why can’t I see this on feedtube.com which you keep mentioning but not linking? What’s going on?
I don’t know where this goes. I still have a lot of thoughts bumping around in my head. The window of peak RSS feed availability is likely behind us, as even non-commercial news institutions are making excuses for why they’re dropping them. A lot of feeds are just suddenly ceasing to exist without even an announcement. It’s dire, but it’s not lost.
The initial prototype I have now was for me to figure out which combination of components would get this off the ground, so now it’s time to start fresh and build something maintainable and accessible. There’s a good possible 1.0 point for a modern, capable gateway that’d present posts and feeds via ActivityPub without any of the remix or combination features.
From there, I’m thinking still about that fundamental issue from way back in the 20th century that what is actually in the feeds is unpredictable and hard to reliably present to people. Maybe there are solutions to that. Not globally applicable ones, but there are things that help and maybe, just maybe, if posts could be run through a variable set of processors to produce a consistent result delivered as a new feed for a specific purpose. Like, imagine if you put a feed through a series of tubes that connected to other tubes …
If the feedtube.com in this paragraph is showing as a clickable link, you’ll know something has happened.
It’s good that I never registered a domain name for Hubris, though.