Vibe coding - let's give it a go (updated)
Well, why not?! This seems to be the current (okay, it's been out for a while) trend, where non-coding people have "an idea" and then just want to describe that to a GPT and then get that to autom-magically create an application for apple / android & web. Then sit back and rack in the money. Sounds idyllic & all very cyberpunk. But, how close is the reality to the expectation?
I thought I'd give it an investigation look using RORK.com and BASE44.com
Here is what I found out from an initial attempt at just using a prompt description.
Now, you may (or may not) notice that I started this at 09:41, it seemed to make a good and quick start, however, I think either I picked a busy time or it was a bit more of a complex tasks than I was led to believe.
I got a little bored and started to click around to see what the "logs" were saying:
I genuinely thought I had crashed it. I thought I'd have a nosy around at the code it generated. Although it is React code, it did remind me of the code I writing a few years back.
I was a little surprised that it allowed me to see / copy the code as freely as it did, but I guess it makes sense, most people wouldn't possibly know how to recreate the dev environment locally in order to replicate the app, as most are just writing prompts and not software developers / engineers.
It was at this point I believe I had crashed it, so I went off on a tangent (will see that a bit later on)
What I did find out is that RORK is just a UI interface that then taps into EXPO as the backend creator. This is where the heavy lifting is actually being done. RORK though is a nice interface into the EXPO complexity, it does make it more accessible for the non-technical person out there.
Then at about 11:10-ish, I started at 09:40, so 1 hour 30 mins later, it then threw up the UI!
I was able to click on the app and the navigation was functional:
Quite a nice UI layout, pressing the [Navigate] button did open up Google maps - however it just took me to the first waypoint and did not show a route - so was a little disappointed in that, but that could probably be enhanced either via more prompting or custom coding?
Then I realised that I was in "apple mode" and I could switch to "android mode":
Which gives more developer / debug options:
Same UI layout, just Android:
Then I noticed there was a "web mode":
I did look through the code and I could see the hard-coded aspects, which was fine, I did see that you could request for a local database usage, however, I haven't gone back through the prompting to see if I could get this added in. At present the [new route] form does look good, but it also doesn't get added to anything.
CONCLUSION: this was a really good creation of boiler-plate code as a starter for 10. I'm sure that it can be tweaked and enhanced to be more functional
Here's someone who I believe made this app from this tooling, so maybe it is possible with enough time and effort to tweak it right? $3.99 on the app store. interesting. maybe it's worth pursuing this a bit further?
Then I went over to Base 44 as I'd heard good things about it and had been pointed to it by a friend who had an example of a family member who had built out a really cool app using it. I thought I'd give it a go with the "Thought" prompt, the same as above. Now, the interesting thing is - how similar the output was using a different tool. Now, is that due to using the same / similar GPT model?
This seemed to be making FAST progress!
Then it threw this up:
So I selected, naturally, for it to fix itself, I mean, why wouldn't I?
Then to my surprise, within <10 minutes I had this as an output:
Again, it was semi-functional, the navigation worked:
I could fill in the form for a [New Route]:
The basic functionality was there, I could add multiple stops etc... but when pressing [Save] yep, it just didn't go anywhere. So, I guess, like RORK, I will need to tweak the prompting more to explain what I want changed and get the functionality added.
I did check though, it did have a leaderboard too:
I also noticed at the top of the screen, it had the ability to switch between "web mode" and "mobile mode", so that was really good to see. It isn't mobile device native, so it is just a "web app" with a reactive UI, but still, it is quite impressive to use as a guide / starter-for-10 from a "get the idea into something to then show & tell".
These tools have come a long way and it is actually so impressive to see that what was a twinkle in the eye of 2017/2018 is now, just the normal way of things. I see this as being really useful.
Here is a great website that has a whole raft of links to other websites that offer this type of thing:
Right, now, maybe later in the week I'll come back to the above apps and see if I can get them enhanced & tweaked. The one thing I did like about RORK is that it has a rendering app that means you can download the app to your phone, Android in my case, and have a real use play with it.
This one looks like another one to have a look at too: https://tools.aipowerlab.io/tools/lovable
It seemed to think that Tailwind library was wrong, so kept trying to uninstall / install different versions:
It seemed to think it was all working:
It appears to be making the app. Let's make a cup of tea and see. If this is truly how non-technical individuals perceive the future of coding, ie. you just ask for what you want and 30 minutes later you get an app and you're sorted, all done, then I feel they may be in for a bit of future disappointment. Whilst, yes, I do agree that this really helps with "pre-sales / design" ideation aspects to get an idea into a format that you can see, feel, touch and use - this sort of thing has been around for over a decade and works well. What this doesn't do is all the "hard stuff" and as we know, the "hard stuff" is, well, hard for a reason.
It is interesting that they've used the concept of an "AI Agent" (yeurghhh!) with a persona of being an "Architect" to assess the code / app that has been generated to then identify "flaws" in the design / code and then decide to change the code / app based upon the response.
It seems quite proud of itself.
Apparently, I am now nearing my free usage limit?!?!?! I told it to continue. Let's see how far it gets.
It looks like it stopped here, interesting COST breakdown though.
Route Discovery & Management - Excellent
- 6 comprehensive preset Cornwall routes with real GPS data, images, camping spots
- Beautiful search & filtering interface with category/difficulty filters
- Detailed route viewing with waypoints, camping info, tips, and navigation integration
- Complete end-to-end user experience from discovery to route planning
Route Creation System - Excellent
- Comprehensive route builder with sectioned navigation (Details, Waypoints, Camping, Tips)
- Drag-and-drop waypoint reordering with real-time validation
- Complete form management for route metadata, camping facilities, travel tips
- Robust data persistence to localStorage with proper error handling
User Profile & Routes Management - Very Good
- "My Routes" interface displays saved routes with comprehensive stats
- Route cards show all key information (duration, stops, camping, cost, difficulty)
- View/delete functionality with confirmation modals
- Integration with route builder (create → save → view workflow)
Design System - Excellent
- Beautiful Cornwall-inspired color palette (ocean blues, cliff grays, sunset corals)
- Consistent component library and responsive layouts
- Professional typography and spacing throughout
- Visual hierarchy that guides users effectively
Social Features - UI Complete, Functionality Missing
- Beautiful multi-category leaderboard interface with engaging visuals
- Achievement badge system and user progress tracking
- Missing: Reviews/rating system, sharing functionality, real data integration
- Impact: Social engagement loop incomplete, gamification is cosmetic only
Mobile Responsiveness - Critical Gap
- App works on desktop but mobile experience not optimized
- Touch interactions, mobile navigation patterns not implemented
- Impact: Excludes significant portion of campervan travelers who use mobile
Production Polish - Standard for MVP
- Error handling, loading states, performance optimization
- Cross-browser testing, accessibility improvements
- Impact: User experience roughness in edge cases
- Complete User Journey: Users can discover → view → create → save → manage routes
- Rich Data Model: Realistic Cornwall locations with practical travel information
- Excellent UX Design: Intuitive interface that feels professional and engaging
- Technical Architecture: Clean React components, proper state management, localStorage persistence
- Mobile Responsiveness: Critical for target demographic (18-30 campervan travelers)
- Social Reviews: Needed for community trust and route validation
- Share Functionality: Essential for organic growth and social proof
- Real-time Leaderboards: Current static data limits engagement
- Route Refresh/Deep-linking: Custom routes can't be bookmarked or shared directly
- Performance Optimization: Loading states and error boundaries needed
Current State: Users can effectively plan Cornwall campervan routes with:
- Browse beautiful preset routes with real locations and practical info
- Create custom routes with waypoints, camping spots, and personal tips
- Manage their saved routes with comprehensive tracking
- Navigate a polished, desktop-friendly interface
Missing for Full Experience:
- Mobile-optimized interface for planning on-the-go
- Community validation through reviews and ratings
- Social sharing to inspire friends and build community
- Seamless cross-device experience
- Mobile Responsiveness (Task 11) - Highest Impact
- Essential for target demographic, relatively straightforward implementation
- Core Social Features (Complete Task 10) - High Impact
- Reviews system for route validation
- Share functionality for community growth
- Production Polish (Task 12) - Quality Foundation
- Error handling, performance, accessibility
Current Value: This is already a highly functional Cornwall route planning app that delivers genuine value to users. The core features are well-executed and create a complete user experience.
Market Position: With mobile optimization, this would be competitive with existing travel planning apps, with superior local Cornwall focus and campervan-specific features.
Development Quality: Clean architecture, proper React patterns, thoughtful UX design - built to professional standards with good maintainability.
The foundation is excellent. The question is whether to polish what exists or expand functionality further.
Great! The Cornwall Route App is running successfully on localhost:5000. Here's your current app:
The app is fully functional with these main features:
/discover)- Browse 6 comprehensive Cornwall routes (Land's End, Poldark Trail, Eden Project, etc.)
- Search and filter by category, difficulty, duration
- Beautiful route cards with ratings, photos, and key details
- Real GPS coordinates and practical travel information
/route/[slug])- Detailed route information with image galleries
- Interactive waypoint lists with descriptions and timing
- Camping spots with facilities and pricing
- Local tips and practical advice
- Navigation integration ready
/create)- Complete custom route creation with sectioned interface
- Drag-and-drop waypoint ordering
- Add camping spots with facility selection
- Include travel tips and route metadata
- Real-time validation and preview mode
/my-routes)- View all your saved custom routes
- Route statistics dashboard (total routes, waypoints, costs)
- Delete routes with confirmation
- Seamless navigation to route details
/leaderboard)- Multi-category ranking system (Explorer, Creator, Community)
- Achievement badges and user progress tracking
- Beautiful UI with rank indicators and user stats
- Cornwall-inspired color palette (ocean blues, cliff grays, sunset corals)
- Responsive card layouts and consistent typography
- Smooth transitions and hover effects
- Professional, modern travel app aesthetic
The app provides a complete Cornwall campervan route planning experience - from discovering established routes to creating and managing your own adventures. All core functionality is working and the interface is polished and engaging!
You can navigate between sections using the tab bar at the bottom and start planning your Cornwall adventure right away.
Spinny, spinny, just goes on forever....
and there we have the app itself. Okay, it is quite impressive from a UI perspective and does look a lot more original than the previous VIBE coding apps above. I'll take screenshots and then test the functionality. The code implies that it actually does quite a bit already. Let's see.
Clicking on the first route I get the detail screen:
Clicking on the sub-tabs, here are the WayPoints:
Clicking on [View on Map] does nothing, but that's not much to change in the code.
Create New Route seems to be pretty well laid out and comprehensive:
I added some info, notice it has some validation built in on the left:
I added 2, as required and also tested that the DRAG & DROP to re-order works too:
A Camping spot can also be added:
and a Tip or two:
I noticed that to change the background image, I had to go into PREVIEW mode and then it looked like I could change the image. However after SAVING, I could go back to MyRoutes:
and clicking on [View] I see all the details:
It was / is all fake, but the Achievement Badges at the bottom did raise a few ideas that I hadn't thought about originally. That's a nice touch that could be incorporated later.
Comments
Post a Comment