I recently spoke with Kristján Pétursson, the creator of a new RSS tool called Sponder and developer at Haha Moment. Short for transponder, Sponder allows you to create custom feeds tailored to your specific preferences that you can then import into your RSS reader (you should check it out!). The interview has been lightly edited for clarity.

Micah Blachman: What was your inspiration behind Sponder?

Kristján Pétursson: It started as a different idea: a long time ago, a friend of mine and I had this idea of an RSS social network but where everything is purely feeds. He actually texted me this morning, he forgot I was building Sponder and he’s building a similar thing now, too.

But the idea was that you subscribe to whatever you want. If you like it, you can publish it to your outgoing feed of things you find interesting. Then you could also publish things to your friends, like inbox feeds. So there would be a “things Kristján thought were interesting” feed and a “things Kristján thought Micah might like” feed. So I could publish in general or just to you.

I started intending to build that, but I realized that to get that working well, you needed somebody to be using a client for that product, and I didn’t want to get into a fight with everybody’s favorite RSS readers. People like what they like, I like what I like, and it felt like a very small product to say, “Abandon everything that you're doing now and come use this instead.” But next to that I had something I wanted to do. Most things I build are just because I want something and then I think that maybe another person wants this too.

There’s a blog I read called Bits About Money that’s about the intersection between tech and financial infrastructure. It’s fascinating in a way that I never want to actually have to do myself, but it’s fun to read about. I wanted to read its archive, but RSS feeds don’t do archives. They do what’s happening right now. I needed this thing that would put together the archive with the live feed, and then I could replay history at a compressed rate because this blog started 15 years ago and publishes once every month or two. I figured I’d read maybe an article every week or two until I caught up.

I thought that kind of tool was extremely specific and hard to build because I have to do web scraping. And so I thought, “Okay, let’s just get the system set up first to do some basic filtering.” Maybe other people can use that. That was actually the first version. It was just an extremely simple filter and the ability to make an RSS feed.

Then I started building in the things that I actually wanted that were harder to do—all of this difficult stuff under the hood that hopefully looks very simple on the front-end. Along the way, my friend mentioned, “Oh yeah, I wish I could skip reruns of podcasts.”

I thought, “Podcasts are RSS, I can do that.” I started putting on all these large features for what was, in theory, a small thing. But now Sponder does that! Once it had that feature, I felt like it was ready to push out onto Hacker News and other places because it was more than you could do with a simple extension to an existing system like FreshRSS. The rerun filtering involves a ton of historical analysis and statistics about what phrases tend to appear in titles when it‘s a rerun versus when it’s not. It felt bigger than somebody who just built something in an afternoon for their existing reader at that point, so I said, “Okay, I’ll give it to everybody else and see if they like it.” I’m still trying to get the word out there, though.

MB: I’m curious to know what made you decide to not have Sponder be an RSS reader with added capabilities?

KP: It’s mostly that I didn’t want to start that fight. It’s hard enough to get people to come and use something new. It’s even harder to get them to replace something they already have, whether or not they particularly like it. I’ve been meaning to replace Feedly for ages after they started cramming all their stuff into it.

But then, I know it’s going to be a 15-minute process where I download the OPML [a file format RSS readers use to exchange data] and then upload into my new reader. I know it’s not going to be hard, but I just haven’t wanted to bother spending those 15 minutes. And that’s a thing that I actively don't like using anymore. I just didn’t want to have to convince people to replace something even that they didn't like, but certainly something that they do like—if they’re already using FreshRSS or Feeder or whatever they’re happy with.

There are lots of good clients out there already. There are really good podcast clients all out there already too, and I really like when things are composable and just do their one job. It’s like the old Unix philosophy. I thought, “Okay, I see all these clients, whether they’re for RSS or podcasts or whatever, and they all have different features, right? They all organize your list in different ways. They all do different filtering. They all have their different stuff they chose to approach. But none of them do the entire gamut of every feature you could want.” But because it’s all just an XML file under the hood, Sponder can do whatever you want. Plus, you can pull from your sources that already exist. You can use it in your client that you already love. And it feels like an easier lift for me, and substantially less competition with things that are already out there and good.

MB: Where do you see Sponder as going in the future? Who do you feel is the target audience?

KP: It’s changing as I find new use cases, and I think it’ll change more as I try to introduce it to more people. You can probably tell from the demo UI—the fact that you can switch to a YAML mode [YAML is a human-readable format developers use to structure and organize data] if you just want to type code, I made this for technical people. I made this for myself and thought, Okay, maybe I can find a hundred other people like me.” I started to see that perhaps it could be broader pretty soon, though.

I have a bunch of friends who are doctors—my wife’s a doctor, and so I stole a handful of her friends and I play games with them pretty frequently. One of them mentioned that they use RSS to get their medical journal publications. But they’ll get a quarterly dump of 250 articles and probably care about 15 of them because they're doing cardiac electrophysiology. They don’t need all of the cardiology journal or even all of the general medicine journal, they just want electrophysiology. So I thought I could save doctors some time. They have enough on their plate, maybe I can filter for them.

When my friend mentioned podcasts with reruns and cross-promotions, things you didn’t quite sign up for, I thought I could do the analysis there, that’s an extremely broad thing. As much as I built Sponder to be a general and relatively technical tool, I feel if it gets any mass adoption—and by mass maybe more than a thousand people—the skipping podcast reruns sounds like maybe the most likely path to that. In the next phase or so I am going to try to make that good, easy, robust, and work obviously instead of being as finicky like it is now, because I could talk to just some normal human being, right? Normal human beings don’t want to come type YAML.

I could say, “Hey, are you tired of your podcast reruns?”

They might say, “Yeah, they’re kind of annoying.”

I’ll say, “For $2 a month, I can solve that for you. Put in the URL here, sign up. Here’s a new URL. You're finished.”

That seems like the best wide adoption path. I’ll be pleased if I can be useful for a hundred or two hundred people. I don’t expect it to be larger.

MB: How do you see Sponder as paving the way for future projects you might do with Haha Moment?

KP: The Sponder code itself took roughly three weeks to put together, and then added-on features like the hosting aspect turned into probably two months. But overall, it’s been about three months, because then you have to go set up the entire infrastructure. I put it on Google Cloud because they have a lot of good free tiers. Before Sponder is making me any money, I would like to avoid paying money just to keep it sustainable. But to do all that gave me all the tools I need to know how to push the link up to Google Cloud. Just having all that infrastructure work available to me now will be handy for the next things I do, and I expect them to go faster.

MB: What do you think people are getting wrong with RSS, and not really understanding that you think you could help people with through Sponder?

KP: That’s a cool kind of advocacy and education question. I don’t think people get anything wrong. I think they don’t know RSS is there. RSS was great between 2000 and 2010. Then Facebook, Twitter, Instagram, TikTok, and all of the current social media came and took over.

I think one of the big things that caused that shift is that RSS doesn’t natively have a shared nature, the actual networking part of it where you go to one spot and you see all the things your friends are doing and it grows from there. With RSS, you have to go to a site, wonder if it has an RSS feed, find the RSS feed, and put it your reader.

It’s just a bunch of steps that the average person or even probably the modal person doesn’t know is even an option. Just having people use RSS more is a massive outreach and education problem to solve.

MB: Do you think RSS is still thriving? From projects I’ve seen online, it seems like there are still people actively developing new RSS tools, like Sponder. But from your perspective, how do you see other people using it?

KP: I don’t feel like I have great transparency into what’s going on. From what I see, there seem to be two camps.

First, there are people probably like you and me, who subscribe to anywhere from 10 to 200 feeds and read stuff as it comes through whether or not they’re being completionist about what they read or just using it as a stream that goes by.

Second, there seems to be the more enterprise side of things where feeds roll up publicly available data and present it to a team that very much cares about it. For the security team in a large enterprise organization whose job it is to be on top of this stuff and react to it fast, for example, RSS is a good medium to push information.

And then, between the niche reader and the enterprise team, there doesn’t seem to be a middle ground. That middle ground has been eaten by more modern social media. I don’t know if it can ever really come back from, or if it will always be this almost-nostalgic technology, the way people look at having a fancy car with a manual transmission, or riding expensive and fast race horses. We know that horses are not a part of the modern infrastructure. We invented engines and now horses are gone. So why do we even still have horses? Just because people like them and people race them.

It feels to me like that's where RSS is. People like it, but most people don’t think about it. I expect that to continue to be true.

MB: I think it’s interesting to compare it to email, a legacy format which is still thriving. People are still using email—maybe some people have switched mostly to using other communication formats, but email is certainly not going away. I think it’s interesting to see that RSS is losing while there are still so many people using email.

KP: The first mover advantage is huge in a new space. Facebook wasn’t quite the first mover in social media—there was MySpace and some things before that. But various things happened: MySpace failed, Facebook took over. Email was after the older technologies, like the Bulletin Board System, which was a little more like RSS in the feel that you were dialing in somewhere to get the things that are there, instead of something being sent to you at your home.

I think email was the first one that seemed to just become the common denominator. It was a way that you could have a box where people could dump things in. And it felt very much the same as physical mail, in that you have a mailbox and people dump things into your mailbox. And we all hate the spam that gets dumped into both of those places, so there are obvious downsides. But email was there first, and it was an understandable metaphor for people. Now it is so widespread, and so relied on by the companies you sign up for and the companies that advertise to you.

It’s such a large ecosystem that the momentum there is going to make it stay alive. Then we went and started putting all of our login and authentication into email because it was available. Now everybody is fundamentally bound to their email address as their identity online. If they lose access to it, they also lose access to their bank and to all these other crazy things. I don’t think it's going away, but I do wish that email was more pull-based and less push-based.

I want to sit down and read my RSS articles when I have a down minute and I’m just relaxing, as opposed to somebody deciding that I need to buy their new product and emailing me, and that causing a ping on my watch. It makes me reasonably upset every time that happens.

I mostly turn off notifications for everything. I will not listen unless I’ve gone to look for the thing. So I really want a pull society, to the extent that anybody’s going towards the pull model that RSS matches. But nobody’s replacing email at this point.

MB: It’s interesting—you were mentioning the social network you thought about creating, and I think that has interesting parallels to email, in that you have an inbox and you are filtering out everything you don’t want. You’re left with only the things you are actually interested in, which is one of the pros of RSS. I think a hybrid between the two might be something that people would want, to remedy the shortcomings of both email and RSS.

KP: Maybe, because RSS definitely lacks the social aspect in large part. There are a couple of services that let you mention people when you publish things [webmentions]. But that’s why my friend and I had this idea and we liked how you had an inbox feed, but it’s not pushing anything at you. You’re going and checking the box when you’re curious.

Contrast that to email, where advertisers are telling me they have new pants and I need to buy pants. I don’t need to buy pants. When I have a hole in mine, that’s when I’ll go buy pants. It makes me a little sad that I don‘t think anything will ever unseat email—but we can still have a little bit of optimism.

MB: What would you advise to someone who wants to do more with RSS?

KP: I think I would encourage people to think about how to replace their stuff. Look at what you are consuming, think about what parts of it you actually want to consume versus having somebody throw a thing in your face, and then find where you can get that in an RSS feed.

I’m building out a few tiny tools in Sponder to help you do that, to find RSS feeds that might be harder to find, and it’ll try the 15 places it might be, and then maybe that’ll get someone to replace TikTok.

But you should just figure out what parts of your content make you happy and then find an RSS feed for it. And then you don’t have to think about it anymore ever again. You’ll get the things that make you happy, and then you’re done and can ignore everything else. Nothing will ever be added without you taking more action and saying, “Oh, I thought of a new thing that makes me happy, and I’m going to find a feed.”

MB: You had mentioned in your email that you’re working on a puzzle game—what is it and what was the inspiration?

KP: The inspiration was Pips, in The New York Times—it popped up recently, and it’s a fun game. I’ve been doing Sudokus and a bunch of variants of Sudoku for years, when I’m sitting at the bus stop and want to think—rather than when I’m sitting at the bus stop and don’t want to think, which is when I read RSS articles.

Pips looked really interesting to me, but I couldn’t quite tell if the puzzles were arranged in such a way that there was always a purely logical path to place the pieces. My wife would usually beat me in terms of time, because she was more willing to place a domino and see if it worked out—whereas I wanted to start from axioms and never place a piece wrong, which is what you can do in Sudoku.

I'm calling my game Scintil, short for scintillate. It’s basically Pips, but you use colorful triangles. The shapes are more interesting, but also they can overlap and mix colors. I think it’ll be interesting. Other than that, it's the same regional constraint idea of “this has to be red” or “these have to be all different colors” or something like that. It’s very clearly Pips-inspired, you’re just building a stained glass window out of overlapping pieces instead of dominoes.

I found my interview with Kristján to be really fascinating, because I too wish that more things were pull- rather than push-based. It’s a little ironic, then, that this is an email newsletter, but hopefully you’re glad it comes into your inbox! (Of course, there’s always an RSS feed available, too.)