Really interesting to see how something built “pro bono” can still gain that kind of traction. Goes to show that distribution and solving a real problem matter more than anything else. It also feels like as more web apps are being built, the challenge is less about creating them and more about getting them in front of the right users. I have been exploring https://unstore.io recently, which focuses on curated and safer web apps, and it has been a nice way to discover projects like this outside the usual channels. Curious what ended up being your biggest driver for those first users.

Great explanation, especially the way the example shows how quickly things break once real usage increases. One thing I find interesting is how different apps handle that jump from small to real scale. A lot of it only becomes visible once people actually start using the product, not during initial development. I have been browsing a lot of these kinds of apps on https://unstore.io and it is interesting to see how different projects deal with scaling challenges in practice.

Nice explanation, especially the distinction between access and refresh tokens. A lot of people gloss over that. One thing I keep noticing when working with modern web apps is how much auth complexity affects how quickly people can actually ship and iterate. It is interesting seeing how many indie apps solve this differently. I have been browsing a bunch of these on https://unstore.io and there are lots of small apps handling auth in creative ways

This is such an important reminder. It’s easy to treat semantics as optional when everything “works,” but the long-term cost (especially for accessibility) adds up quickly. What’s interesting is how this extends beyond code quality into how we discover and use web apps too. A lot of tools are technically accessible, but still hard to find or revisit in a meaningful way. Feels like there’s a broader gap between building things correctly and making them truly usable in practice. Some platforms are starting to explore that layer (like https://unstore.io), but it still feels early. Really appreciated how clearly you broke this down.

This is a great breakdown, especially the part about how modern frontend architectures make scraping much harder than it initially seems. It really shows how different today’s web apps are compared to more static sites. One interesting side effect of this is that discovering tools or extracting useful workflows from existing apps becomes less straightforward as everything gets more dynamic and protected. I’ve seen some early attempts to approach this from a different angle, like https://unstore.io, where web apps are surfaced more intentionally instead of being discovered indirectly. Feels like as the web evolves, we might need better ways to navigate what’s already built rather than relying on workarounds like scraping.

Nice comparison, especially around how Expo Router simplifies file-based navigation. It feels like a lot of these tooling decisions are getting easier over time. Something I’ve noticed though is that once you’re past these choices and actually ship your app, the next challenge becomes distribution. The stack is clear, but getting users in front of what you built is still pretty fragmented. I’ve been exploring that problem a bit and came across https://unstore.io, which is trying to make web apps easier to browse in one place. Feels like we’ve made building apps much smoother, but discovery is still catching up.

This is a point that gets overlooked a lot. It's easier than ever to build an AI demo, but reliability, maintenance, security, and user experience are what determine whether something survives beyond the initial launch. One thing I've noticed while browsing new web apps on platforms like https://unstore.io that there are countless AI-powered prototypes being released every week, but only a small percentage feel production-ready. The difference is rarely the model itself. It's everything around it. The post does a good job highlighting why "it works on my machine" is still a long way from a sustainable product.

Nice summary, especially useful for anyone trying to figure out the world of open source. Your observation about the problem of information overload certainly rang true because of the numerous tools and repositories available that make it difficult to navigate without some sort of framework. The other thing that was interesting to see here is your focus on platforms that could provide some sort of structure in such a fragmented space. Similarly, I found https://unstore.io to be quite different from all of this because it focuses on a slightly different way of discovering tools, in a manner similar to an app store. It is fascinating to see how the discovery process depends not only on the quality of tools but also on the medium used to discover them. What do you think would prevail in the future, curated platforms or a more discovery-based approach?

Really enjoyed this breakdown, especially the deployment pitfalls. The node_modules issue and the caching problem with Nginx felt very familiar from my own experience. I also like the philosophy behind your tool site. Keeping things ad free, without forced logins and with a focus on privacy is something a lot of tool platforms have drifted away from. One thing I have been noticing while working on side projects is how fragmented tool discovery still is. Most of the time you rely on blog lists or Product Hunt launches, but there is not really a simple way to browse web apps in a more structured way. I recently came across https://unstore.io which is trying to approach this more like an app store for web tools, which feels quite aligned with what you are building. Curious how you are thinking about distribution long term. Are you focusing mainly on SEO or experimenting with other channels as well.

Really like how you broke this down, especially the focus on speed and avoiding over engineering early on. One thing I’ve been experimenting with lately is reducing setup time even more so I can focus almost entirely on validating the idea. Been using simpler approaches like https://unstore.io to launch small apps quickly without getting stuck in the build phase too long. Feels like the faster you can get something live, the better decisions you make. Curious if you think 7 days is becoming the new standard or if even shorter cycles make sense?