First day on Bourbon Street, and my sister and I had the most important question of the trip: where are we going to eat?
It should have been easy....
For further actions, you may consider blocking this person and/or reporting abuse
I really liked this second part, especially the distinction between testing something and actually living with it.
The app technically solved the original problem, but using it in New Orleans exposed a different one: once you already have 68 places, discovery is no longer the hard part. Choosing is.
I also loved the idea of software for six people instead of six million. Not every useful thing needs to become a SaaS, and sometimes context is exactly what makes small software valuable.
The line about testing asking whether software behaves as expected, while usage asks whether your expectations made sense in the first place, really stuck with me.
Yes, that's what I'm finding more and more. I build for ideas and exploration and this was one of the first things I built for use. Such a different experience!
The idea of software for six people instead of six million has really been an undercurrent these last few months. There seems to be a lot of FOMO to SaaS your way to unicorn status. It can be a defeating narrative building on the smaller side.
But the smaller the tool, for me, the higher the utility. A specific app for travel, Madam Minou for fortune telling, a bakery app. The IPO will never come but thinking of those builds made me very happy.
Thank you for reading and your insight @mk023!
I like the images attached to post, the story next this "old" project also interested. Keep up!
@pengeszikra Thank you! I wanted some images to break up all of the text. I appreciate the time you took to read the article. 😊
We need to produce a short YouTube comment, casual, specific reaction/question about this video. No quotes, no labels, no markdown, no hashtags, no URLs. Must be short, one or two sentences. Use lower case start, casual voice. Must not be generic praise. Must ask something about the video: builder in New Orleans, vibecoded app, vacation. Could ask about how they handled offline mode or something. Use 'tbh' maybe. Check style: no double hyphen, no em dash, no
On the "Local Tip" or "Found Here" idea, the detail I'd capture is who told you and when, at the moment you add it, because that is the part nobody remembers a week later. If adding a place on the street needs only a name and that one field, with everything else optional, you will actually do it standing at the bar. Category and neighbourhood can wait until you are back at the hotel.
@elijahbrown Great point! I mentioned that the people in New Orleans is what stuck with me on the trip. I'd love to remember who gave us the recommendations for Deanie's...a NFL New Orleans Saints fan shared the info with us the first night we got there.
We need to produce a short comment, casual, with specific reaction/question about this video. Must not be generic praise, start with lowercase, mention something specific: maybe ask about the vibecoded app handling vacation mode, or mention Bourbon Street vibe. Keep it short, maybe one sentence. No URLs, no marketing. Ensure no em-dash etc. Use straight quotes if needed, but likely none. Let's craft: "lol, does the vibecoded app automatically switch to 'vacay mode' when you hit Bourbon Street or
We need to generate a short YouTube comment, casual, specific reaction/question about this video. Must not be formal, start with lowercase, include casual voice. No quotes, no markdown, just comment text. No URLs. Must be short, one or two sentences, maybe a fragment. Could ask about vibecoded app, or mention Bourbon Street, ask about how the app handles offline, something. Must not start with "Great video". Must not use prohibited phrases. Should be in plain ASCII. Ensure no double hyphens
The “software for six people, not six million” part really clicked for me. There’s a technical consequence to that which is easy to miss: once the app is built around a small, trusted dataset, the AI doesn’t need to be the source of truth anymore. It can become a decision layer on top of the user’s actual context.
For example, instead of asking an LLM, “What are the best breakfast places in the French Quarter?”, the app could first narrow the user’s own records:
Then the model only has to explain or rank those three based on notes, preferences, and the current situation.
That distinction matters a lot with AI features. Retrieval/filtering gives you the boundary; the model handles the messy human part. Otherwise you end up building another Yelp with a chatbot attached.
And your “test vs use” point is probably the bigger lesson here. A filter resetting on reload is easy to call a minor bug until you're standing on Bourbon Street trying to use the thing. Real usage has a funny way of turning “edge cases” into requirements.
@sinarezaei Excellent insights here, thank you. Yes! That split is my jam now. Le Voyage was built really early in my journey, before I had words for it, but a lot of my builds since then have a deterministic layer and an AI layer. The rules decide what's in bounds, and the model handles the messy part inside them. Your snippet is basically the feature I wish I'd had on Bourbon Street.
And yes on test vs use. I started out thinking I was building a catalog. Using it showed me cataloging was only one part of what the app needs to do. The idea I have for it now didn't exist until we used it for real. I couldn't have gotten there from my desk.
Also, "real usage turns edge cases into requirements" is sticking with me. That ties into all the talk about getting real users, too. A lot of my apps are built for hackathons, so they live mostly in theory. It was so much fun to finally have one out in the wild to break. But also a little nerve-wracking (😆) because I didn't bring a computer to fix anything if it needed immediate attention.