Hi and welcome back to “Trust the Process”.

After a more introductory issue last week, let’s talk a bit more about what I’ve actually worked on this week. Namely Morgenblau, my main product.

If you’re new here and have not heard of it yet, Morgenblau is essentially my ideal RSS reader with a tight ATProto integration that enables some cool (mainly social) features that weren’t possible in regular readers. But as you’ll read below, figuring out if that description is actually accurate for Morgenblau is quite a hard thing to do.

Process

As I already explained in the last issue, the progress on Morgenblau was practically non-existent over the summer. So to stop myself from procrastinating any further, I now only have one goal: Ship and deploy a first version as soon as possible with the goal of making it my daily driver and replacing NetNewsWire, my current RSS reader.

This week, I laid the foundation for that. The two main things I did were removing all the discovery-related code that was already there and a complete reset of the frontend part of the app (as always, after some time, I just wasn’t happy with the design anymore).

Tearing out discovery was mainly due to the fact that I don’t yet actually know how that feature should work, which led to the feature feeling quite uninspired and generic. Plus, since I already have quite a lot of sources in my current RSS reader, it isn’t required for me to start using Morgenblau to replace NetNewsWire. I envision discovery mainly solving the “cold start” problem for users that don’t use RSS yet and being the thing that might make “regular” people give RSS a shot in the first place.

What still stays though is the tight integration with a user’s PDS. All RSS feed subscriptions and article saves will still be saved there. This will enable a discovery feature in the future.

Which brings me to one core thought I had over the summer: So many cool Atmosphere apps are launching all the time but only very few of them seem to gain traction. In my opinion, this comes down to two reasons:

I’m sorry to say this, but a lot of them simply look completely vibe-coded and don’t feel as polished as they could be. I think this is due to the fact that a protocol like ATProto inherently is more interesting to work with for a backend developer than someone more frontend-focused. Which isn’t a bad thing per se; many apps are really impressive from a functional perspective, but in reality, many judge a book by its cover, which means an app has to mainly look good and have good UX.

The second reason is that many apps position themselves as “XY for ATProto”. The unfortunate thing is that outside of developers, almost nobody cares about this and it’s just not a selling point. I think we as an ecosystem have to do a better job of just building good apps and integrating the protocol where it makes life easier for the user (and thus makes the product better) instead of making it the central feature we advertise. I myself made this error too, as I still mainly pitch Morgenblau as “Your calm window into the Atmosphere”. Nobody knows what that’s supposed to mean.

This is why I also started to question what Morgenblau actually is. I think I will probably end up describing it as a “next-gen” RSS reader to a more technical audience (they know what RSS means), but I’m still figuring out how to market it to regular people.

My belief is that in today’s online culture, where many associate the web with just social media but are fed up with the algorithmic and engagement-driven nature of it, there is quite a big target group for a tool that helps them consume more intentionally. They don’t care that RSS or ATProto powers it. They just want an interface where they can consume the content they curated and care about in a calm and non-distracting way.

I quite like the term “slow media”, but I’m not sure what that means for Morgenblau. Is it a “slow reader”? Sounds like a performance issue honestly. And Morgenblau doesn’t only support blog posts, YouTube videos are also first-class citizens. Maybe you have a better idea?

On a related side note, inspired by Herman’s post “On reading books”, I disabled my YouTube watch history this week. Maybe you don’t know this, but this makes YouTube sooo much less addicting, because it will stop making personal recommendations and your home screen will be empty by default. Paired with an extension like Unhook, I now have a setup with almost 0 distractions, which is really nice. Until Morgenblau is ready, that will serve me quite well.

I also told you about my small budgeting app experiment last week. This week, I wanted to play around with the new cool kid on the AI block, Jev (well, not the newest anymore after the Opus and Sol/Luna release I guess, time moves too fast). I thought about how that might maybe even make a dedicated app obsolete, as I can just hand Jev my transactions in CSVs and it will automatically categorize them and cost practically nothing. And it doesn’t really need an interface either, which is why I started to build a little CLI as a prototype. I didn’t invest too much time into that yet, but so far, it works only meh. Not sure what the way forward is, I’ll let you know.

Finds

Cool things I discovered this week:

• Roman Opałka – Wikipedia: An artist who just painted increasing numbers for 46 years. Sounds silly, but the idea behind the project is quite interesting.
• LOOPWHEELER • A Type 7 Film: In the age of AI slop, seeing this Japanese man working on perfecting something as normal as T-shirts just makes me happy. Craft and especially caring are still such important aspects of any work.
• How I Vibed a Proof of Conway’s Conjecture — overreacted: Fascinating because it shows how a non-mathematician can now essentially prove mathematical theorems with some really cool and interesting techniques. It’s still far from easy though.
• UseMods: A collection of JS utilities I bookmarked for future projects.
• Grit your teeth and ship it: That one felt almost personal, but it was a great reminder to just let go, “grit your teeth and ship” the thing you made.
• Things I’ve been thinking lately: I generally like Alberto’s posts, so I thought I’d share his most recent piece.
• AI-generated posters don’t have to be horrible: Another proof that AI slop is not inevitable, if you just care enough and don’t accept the first solution.
• Blip is the fastest way to send files: Since switching to Android, this is my AirDrop alternative. I would say you should use it, even if you are on iOS (it works on all platforms). It’s just so incredibly fast.
• Search — a browser with nothing in the way: A minimal browser, maybe not so much for daily use, but it looks great in screenshots.
• Built by Designers: No Explanation Needed: Tools built by designers, there are some real gems here.
• Lunagraph: Design canvas that writes code, powered by Claude Code: One of the gems found on the site above. I have to try it out for my workflow.
• How to rage bait Americans: Christophe’s videos are always great, but this one was particularly interesting, and it shows again how bad social media is when it’s just engagement over everything.
• On reading books: Musings about an activity that we all should do more.
• The Amp Way: I started using Amp this week and their manifesto is quite interesting. Seems a bit too ambitious in the beginning, but after starting to use their product, I’m beginning to get it.

From the blog

I didn’t publish any posts since the last issue.

That’s it for this issue. Thanks for sticking around! If you have any questions, feedback, or just want to say hi, reply to this email and it will land straight in my inbox.

Until the next one, Dominik