I feel like this app, the Telegram channel that just posts all the casts being sent on Farcaster, is for me a confirmation that frontend/product development is uber difficult if you want to build a really great product.
The backend here is almost trivial if you know the tricks. I pulled in the snapchain repo, told Claude to extract the gossip part, and basically off I was. This was extremely straight forward. And it has remains consistently online for the last 24hrs. And I expect it to remain online.
However, then I started working on the website that'd show this stream of messages, it quickly dawned on me that it was impossible to build a visually appealing frontend that would continuously work really well, scale and be available to many people. Your brain bombards you with millions of possible questions. Where do I host this? What domain do I buy? How do I make this available? etc. I was working on the website and it had all of these edge cases and implementation details that I felt weren't important for the product to work. My intention was to express the level of engagement on Farcaster. Make it tangible how many people post.
So I had the idea for just posting it into Telegram and this solved so many problems for me where the stream of messages is now actually useful. You can go to that Telegram channel, it'll just work. You'll be able to scroll up or down indefinitely. It's smooth, the previews pop in neatly, everything is nicely animated. Your TG app won't crash despite the many messages sent there. I didn't have to register a domain name, I didn't have to add a Farcaster mini app file, or deal with the annoyance of of getting certain SDK details right. I'm literally just blasting all messages into the Telegram API. It took me minutes to do that. And it works. I can now genuinely switch between these two apps on my iPhone, and I can always be the fastest reply guy.
idk, the level of pain I'd have to go through to get this going elsewhere, e.g. on a website or a mini app, or a native app. All of these are 10x harder to achieve, especially when you just want to validate the idea.
I don't underappreciate anymore how complex it is to build a really really really good product. In a past life, I'd get freelance gigs with a 3 month timeline and a client that just "wants to have the product done by then." To me, this is not at all how a product is built. That lives off constant iteration. Not a one-off sprint to develop an non-validated brain fart.
E.g. with Kiwi News I've non-stop coded on building a really good app experience and I see how that product is now getting on par with some other apps which have larger teams. Importantly, it is to mention that I don't mean to compare myself on the level of e.g. success that others had to attract users (they're more successful). I'm purely speaking about the craft to build an app that feels very good in the hands to use. Meaning, it'll load quickly. All buttons etc. work well. Information is properly laid out on the screen etc.
In any event, this realization, that building really really good products is super hard and that it'll take a ton to make something world class also has made me deeply hopeless about ever managing to build something beautiful that'll stand the test of time. I just think that the world of economics doesn't want us to do this.
With Kiwi News it is such a case. With other ideas that I have too. For me the challenge is in the economics of this. I personally know that the more I invest to build a really beautiful thing, the higher my chances are that others will find it useful. I look around me and literally all popular things are beautiful. Or are all popular things made beautiful?
To me it is that all beautiful things eventually become more popular. But with this mindset, it also means that there's an endless amount of sisyphean work necessary to even get to validation. Like, you couldn't have just slapped together a bunch of website parts, called it Farcaster and then told people to use it in 2022. They've built Farcaster, and then the people came. This was honestly true when people started to realize the importance of the Farcaster hub 3 years ago. It was the same when they built frames, and it is the same now. Bluesky too had a really well built website/app. Their onboarding and everything worked fantastically well. They had built it, so people came.
I think you just have to build beautiful things
