Ask HN: Opinion on low/ no code platforms (Flutterflow, Bubble, others?)
I'm need to build a ride sharing app using a low/no code platform. Hoping to use Claude for assistance, if it is an option. Looking for advice from folks who have used FF or Bubble (or both).
6 comments
Here is my experience, having worked on a competing low-code/no-code platform for several years:
- can be great for quickly getting a proof of concept up and running
- may suit your needs indefinitely if your scale remains small
- will likely scale very poorly as needs or applications grow due to the fact that they generally obfuscate computational complexity
- can be a lot more difficult to style or brand properly
- can be great for quickly getting a proof of concept up and running
- may suit your needs indefinitely if your scale remains small
- will likely scale very poorly as needs or applications grow due to the fact that they generally obfuscate computational complexity
- can be a lot more difficult to style or brand properly
Opinion.
Code to low code to no code.
A code can be converted to a low code. A low code can be migrates towards no code.
That's an ideal world scenario. These platforms and other platforms out there are heavily tightly coupled between front-end and backend hence making them unclean.
Why unclean? Migrating a code based builder(IDE) to a low code based builder (GUI with confoguration) to a no code based builder (DRAG AND DROP) ultimately towards design to code builders.
Unfortunately there isn't any.
Here is how I made my statements above.
An app is a giant state of rectangles of content with show or hide configuration. Each state is followed by an user interaction.
Now the problem is reduced to a UI.
How about building an UI that builds UI?
I have few things documented in past commits of my project called ui-editor.
I stopped working on it publicly due to time consumption of new commits. You can find the wiki or the read me in pat commits.
https://github.com/imvetri/ui-editor.
My opinion, those products don't scale or not maintainable in long run. Because those apps or the builders are code based. To have a clean solution! The builder should be a no code in first place.
Code to low code to no code.
A code can be converted to a low code. A low code can be migrates towards no code.
That's an ideal world scenario. These platforms and other platforms out there are heavily tightly coupled between front-end and backend hence making them unclean.
Why unclean? Migrating a code based builder(IDE) to a low code based builder (GUI with confoguration) to a no code based builder (DRAG AND DROP) ultimately towards design to code builders.
Unfortunately there isn't any.
Here is how I made my statements above.
An app is a giant state of rectangles of content with show or hide configuration. Each state is followed by an user interaction.
Now the problem is reduced to a UI.
How about building an UI that builds UI?
I have few things documented in past commits of my project called ui-editor.
I stopped working on it publicly due to time consumption of new commits. You can find the wiki or the read me in pat commits.
https://github.com/imvetri/ui-editor.
My opinion, those products don't scale or not maintainable in long run. Because those apps or the builders are code based. To have a clean solution! The builder should be a no code in first place.
My experience with low-to-no code as a developer is that even for just-above-trivial tasks the learning curve is still steep, but isn’t even in my area of expertise.
They are all basically wix-like constructors with basic advertised functionality working as designed, but quickly become convoluted, sometimes delusional, at anything beyond that. Also the fact that each year there’s a new set of them to check out.
An ideal low/no code platform could be useful imo, if done right. But real ones always suck at everything except their tutorial. You can watch an unimpressive tutorial and usually that’s all it does.
They are all basically wix-like constructors with basic advertised functionality working as designed, but quickly become convoluted, sometimes delusional, at anything beyond that. Also the fact that each year there’s a new set of them to check out.
An ideal low/no code platform could be useful imo, if done right. But real ones always suck at everything except their tutorial. You can watch an unimpressive tutorial and usually that’s all it does.
One of the Top 5 food delivery platforms in Malaysia used Bubble: https://play.google.com/store/apps/details?id=com.misirakyat...
They were hit with a pretty damn huge bill, which ended up with the tech bro community pointing and laughing at him for not using a proper platform.
So I think it work. It's been done in practice. But be careful where you're headed because low code tends to have a long free tier then suddenly spikes.
They were hit with a pretty damn huge bill, which ended up with the tech bro community pointing and laughing at him for not using a proper platform.
So I think it work. It's been done in practice. But be careful where you're headed because low code tends to have a long free tier then suddenly spikes.