Devs, how do you *really* feel about nocode?
I had a conversation the other day with what you would call a "traditional developer". You know, hammer out code kinda guy, no shortcuts.
We got to talking about nocode (think tools like Bubble) and he scoffed and told me:
"Yeah I tried them because the marketing is good, but once I got under the hood I realized it just isn't a robust option"
Curious to hear your opinion.
Trending on Indie Hackers
🔥 Roast My Landing Page
Being a front-end developer for more than 10 years and running a low code headless CMS, 👂 I heard that some developers and companies are scared of the no-code movement. It'll take over their jobs, they say.
No-code is not the future. Future is less-code, or, as some call it: low-code. And we all benefit from it.
No code allowed non-technical entrepreneurs and enthusiasts to realize their prototypes, MVPs, and even simple products without knowing how to code.
What does it mean for developers?
It's a win-win. New technologies are exciting! A positive-sum game.
Good take.
What do you think? - Is WordPress a no-code, low-code, or something entirely else? It has been a highly integrated LAMP (originally) but now with Elementor that is a GUI website builder and a variety of other plugins (such as e-commerce) it can be used as a no-code or you can open it up and write on top your own PHP ( so low-code).
@ArkSzklar WordPress is everything. You can build literary everything and anything with it. Even a full-blown WordPress-based SaaS. I think it is important for tools like WordPress to decide how you will use them, and for how long.
How to use it:
If you are going to use it to develop something complex that requires a lot of custom development, but you are not a developer - don't choose WordPres. It's not meant to be highly customizable. Of course there are successfull examples of hybrid WordPress projects, and you can make them too - but only you & your team yourself. If you are going to outsource development - consult with a reliable company about the stack before. There's a reason why React, Next.js, and Node exist.
When to stop:
I see a lot of projects fail to scale because of bluntly trying to use technology that no longer fits their needs. Even if you chose WordPress for your project (hopefully a small and simple one), set the boundaries. What milestone, when reached, will make you invest in a custom-developed solution?
People choose WordPress for all kinds of wrong reasons when none of them matters in their particular case. When we started talking to potential customers about Headless CMS BCMS, we often heard from them that WordPress is the most popular option, with hundreds of thousands of developers available on Upwork, etc. But still, they had a crappy slow WordPress website that they had been trying to improve and fix for years.
It's an illusion that using popular technologies like WordPress, React, WebFlow, Vue, Angular, Php, Ruby, or any other, will make your project more scalable and maintainable. More developers familiar with the stack doesn't mean anything.
Yep. That is an allure of a Swiss knife, you can even try to assemble an Ikea set with its Phillips drive, but can you defend yourself against an animal in the wild (the original use of a personal weapon) am not so sure, you need a Bowie knife.
But here is a thing, I can use my common sense to decide which tool to use for a job in real life, but with tech, you essentially have to build with it to really know it, there is not much room for common sense and imagination, especially when all is complex, takes time to learn and change all the time.
Agreed. I'd personally love working with a customer that used a no-code/low code to sketch out a solution to their problem. They'll have a deeper understanding of their problem and be able to speak less vaguely about it.
As a more traditional developer, my role would be to then scale the solution and make it production-ready. This change in emphasis for traditional devs may present a challenge to some but that's to be expected with any career. Our line of work requires constant learning to stay relevant.
It seems like a good idea, but my worry with no-code is that 1) it introduces lock-in and 2) you're likely to be limited by what the tools can do. My general feeling is if you are running a business and choose a no-code solution, you probably should use it for something super simple or else use it as initial "scaffolding" to get your business underway. Once you grow, it's probably a good idea to move away from it since you'll probably have more needs than it can provide, and you don't want to be stuck using some proprietary solution.
Unless it's open source, I'd be worried about a no-code company going under, changing the API one day, or jacking up the prices on you. You just don't have a lot of control.
Low code over no code IMO. Abstractions are always going to have their limits, so the more focused an abstraction can be, the better it will be.
Tools like bubble etc. aren't worth it for me personally as once you hit the limits you quickly lose all the time you saved
Where as tools like Algolia etc. have been much better abstractions IMO
Obviously different story if you're not a developer
I have been developing software since late 80s. And professionally since 1999.
nocode tools have been around since several decades ago.
The difference is that they are called nocode now :)
Is it a marketing thing ? probably.
ie. one of the earliest and clearest examples I can think of right now is Visual Basic in the mid 90s, and its drag and drop widgets. You could pretty much start up an app with almost no code. Or so many thought at the beginning.
There is also a bit of a hype around these tools, and to say they would solve all your problems with no need for programming languages, or developers, is similar to the hype there is around with General Articial Intelligence, and having google engineers claiming there is already one.
Yes, you can create a complete SAAS with nocode 100%. But you won't be able to differentiate yourself that much from the competition that used the same or similar nocode tools.
Having said this, nocode tools have its place, and its benefit, but they are templated applications, that would solve a constrained number of problems.
If you need developers or not, that depends on the scope of the problem.
No code to me is complete and utter garbage. I might be a code "snob" but i think
its best just to learn how to code. Sure there is a learning curve, but it's very interesting subject to learn -- if you are that way inclined and you have patience .
When you write your own code, line by line, you have the flexibility to change any part and the flow of the app, but with a no code tool your hands are kind of tied behind your back. That's my take on the subject! There is no way in hell I could have created fabform.io with a no code solution. At least I'm not aware you could.
That's why we are moving away from building our MVP in Bubble and going to a fullstack deployment. Much easier to change and pivot.
I tried bubble because I found out it is the most advanced nocode tool out there as a dev, In my opinion there are benefits of having these tools in your toolbox, whenever I have an idea of an app to build I don't think I always have to build it with code specially when I just want to validate the demand for it, i have not seen any downsides of existence of these tools.
Always... It depends. We are writing reusable libs are just like no code but highly customizable. Honestly excel can probably do everything. Hard stuff most devs aren't good for either.
I think it has it uses, but I'm not fan of it as a long term solution. As others mentioned, it would be cool for MVP, landing pages and prototypes and it frees developers from a lot of wasted time. But as things become more serious I doubt no code solution will be able to handle growing complexity of the project.
I have been using a few no-code course platforms to run my courses over the years and I've wasted months (> 6) trying to figure out the right platform. I hopped on from one platform to another taking my courses and there was this one or few things always missing.
It's impossible to get all the features you want in a single place and then they say "Hey use Zapier and connect with XYZ for a certain feature" which in my opinion will break my bank once I go full throttle with my business. I do not want to pay 40$ a month for every tiny additional feature I need. And again, those integrations come with their own limitations.
Besides, these no-code platforms won't allow me to make basic UI changes like adding custom CSS to specific pages. Which is crazy in my opinion. The ones that sort of do lack other features that the initial platform offered. The security to content is minimal, users are sharing credentials, the reports are flawed, images are not optimized and I can't do anything about it.
When I upload an image on my current platform, for some reason, a 200KB image becomes 1 MB as I download it and check. A page with multiple images lags and there is nothing that I can do about it.
Other businesses in the same space, with their coded platforms, are able to move at a quick pace introducing new features and such.
With no code, when implementing my system, I feel so constrained. At every step they are like "you can't do this, you can't do that. Hey, use Zapier!" Man, I am done. So, done.
No-code may appear quick to production but not in the long run. It's the opposite in my opinion. As a developer, you'll always do better with writing your own code. Atleast you can tweak your application and make subtle changes without having that fear in your mind, "Oh God, I wonder if the platform is gonna let me make these changes."
Just had to make the rough decision of delaying product because no-code was not secure enough. No delayed 6 months while this non-tech founder learns all about full stack!
Highlighting the large security gaps in data, especially from the customer perspective, is a major red flag.
Right, security is crucial.
I wouldnt use no code. Unless for really simple internal tools that dont need to scale at all. Dont really see the point. Honestly many of these tools are almost as complicated to learn as a programming language (most devs never need to do any complicated stuff anyway for 99% its just glueing together libraries someone else made). Low code on the other hand is great.
I think nocode tools are cool.
Curiously, as my interest in building and bootstrapping has grown, so too has my appreciation for nocode tools. That may or may not be related.
In any event, I generally appreciate the democratization of tech creation tools. More than that I just like the nocode tools which take work off my plate, especially the pieces of work I don't enjoy. There's still plenty of code to be written and I'd love to have so many customers that some nocode part of the system needs reworking. :)
No code is good for very simple focused things. Turning a Google Sheet into a simple app for example.
The problem is when you have even mildly complex requirements then you find it is more time consuming in nocode land than with code. If you already know how to code that is.
I think no-code (as well as whatever it was called before this current iteration) is a precursor to AI driven development. The idea of software writing itself is so compelling that people try to create solutions in this realm even if the technology is not quite there yet. Is it there this time? I don't think so, but it does seem like its moving in the right direction.
No-code is very interesting for small projects, prototypes, internal tools, or projects that don't require much more than entering and visualizing data. It is essential to point out that, despite the current hype, these tools have been around for a long time.
Most experienced developers understand that software rapidly evolves into a state where the ability to customize the code is essential. And nothing has yet been invented that is more efficient for customizing software than code (there are some alternatives like flowcharts, etc., but these solutions quickly become much more complex than code due to the difficulty of debugging, visualizing the structure, and understanding the architecture).
I believe that low-code is more advantageous in this regard, especially when it comes to generating and maintaining code based on battle-tested platforms. Today I have a low-code tool on the market (for Laravel/PHP developers), Vemto (https://vemto.app). Most developers who test the tool like it because it brings many of the facilities of No-code but still generates clean code aligned with Laravel guidelines.
In this way, developers have the advantages of both approaches.
No-code and low-code have their goods, but I think low-code is more advantageous for long-term projects.
Love it, I use Airtable and it's automations a lot. Really good if you want to prototype something fast.
It depends on what you think of as being no-code. Is Squarespace no-code? I run Bear Blog, would that be considered no-code?
We will always be making abstractions in software, sometimes for the better, sometimes for the worse. If you're taking about stringing together disparate sets of services in order to create a functional whole, then you may as well learn to code it as you're thinking like a coder without actually writing any code. It'll be cheaper and faster (and significantly mode stable and inexpensive in the long run) if you code it yourself.
However, if you're just building a simple proof-of-concept, I don't think there's any issue with using "no-code" tools to do it.
I believe that open-source low-code platforms are the way to go to mitigate such kind of issues. I am currently building an open source, low-code product, ByteChef, so let's see.
I'm a traditional developer but I never used nocode tools not because I dislike it, but just because I never needed it. As any programming language, nocode it's just a tool: If you can accomplish your objective with, why don't use?
I don't think nocode will replace code programming, it's just a degree more in the high level languages. Did you ever seen a low level language being replaced (efficiently) by a high level language?
Nocode is perfect for simple applications that don't have complex business logic. A good example of a nocode implementation is CRUD, but most applications need CRUD and other types of logic too.
As I am building a #nocode blogging platform (docswrite.com), I can say #Nocode tools are not new.
There are many old players in Nocode like WordPress, Magento, OpenCart etc
Yes, there are situations when you restricted what you can not do but no code gives the quick momentum you need. Once you are profitable you can always rebuild your platform from scratch.
This comment was deleted a year ago.
I am strongly against the no-code moment. With no code whatever you built for MVP is not user friendly and it will die when you want to add more features.
There are use cases where nocode platforms will do a much better job than building it from scratch. Mainly because of how cheap and fast it is to build with no code for certain things.
Said that I have to admit that those use cases are usually some simple projects and/or something very common.
As soon as you need some very specific things, some API integration, or an extended level of customisation, nocode will become a nightmare.
I gave Webflow a go, absolute nightmare. The learning curve would’ve been a factor but I found it harder to work with than just plain old cold.
Definitely useful for small businesses that need a webpage but for me, the sites they produce feel like how a car would feel if you made it but only did the bolts up half the way. Super shaky and rattles heaps but less work than making sure every single bolt was done the entire way down.
No Code is great for technically minded people...
But it doesn't really help those who don't know how to code to begin with
The average person has no knowledge of databases and workflows.
NoCode actually benefits developers the most as it allows them to save time on the tedious areas and focus on the product
Love it -- can't imagine spending time on internal dashboards and a bunch of other things. Let me work on the creative stuff
Personally I can't stand it lol
No code is supposed to help you build faster. But unless your product is in the happy path, it's normally not that easy anyway.
At that point, why use a no code tool when you can create it in the same time with actual code?
I'm ok with no code. As long as your expectations are in line.
For example, I think most people should use Squarespace or Wix for simple websites. Will it scale if you want to add new features? No. But does it do simple websites well? For sure.
I do not support the idea that no-code will completely substitute the developer, I see those tools as an assistant, a colleague for $15/month. Therefore, I propose to think of no-code as a code generator that generates code integrated into all modern engineering practices: CI / CD, PR and others.
I think that no code may be useful for some kind of stuff, not for developing a SAAS or so but for automating some tasks. For example, using n8n for automating certain processes.
For developing a complete solution I would say it is a waste of time. You will probably need to do things that you can't .
In a bad scenario you would be limited. In the worst you would have to recreate all the service using code because of the lack of possibilities and problems on growing your solution.
My experience so far, as a senior developer, with "low code" tools is that in the long run I'm wasting a lot more time trying to go around their limitations and understand their systems than I would have spent simply implementing it myself.
What would you say about a low-code or better generated-code tool that lest you build in an instant basic functionality but the output is actual react/node (or your stack here) code and you can take control at any point?
If you can build it with no code, it's probably not a very innovative software/product. It's more than like easily replicated by anyone else who can afford a no code platform description.
Maybe it's viable to spin up a very rough MVP to test a proof of concept, but I think it puts you in a box if you are trying to creatively solve customer problems.
I'm definitely not afraid they'll take my job and everyone will suddenly turn into a developer. Literally the entire worlds population can type already, but only a fraction is capable of programming.
Once the tools get good enough to do real work, developers will be the ones using them again.
Also those visual tool will run into the same issue as architecture diagrams do. As soon as a system gets complicated enough to need a diagram to give an overview, the diagram will be too big to give a quick overview.
It's just a mismatch.
For what it's worth, a lot of what we mean when we say no-code is just 'better website builders'. I think they'll continue to get better over time and replace a lot of the mundane work that is currently done by software developers—and designers for that matter. I don't think it will be overnight, and it won't be winner-takes-all, but it is definitely happening.
Just another tool shrug - if they save me time, I'll use them. But most of the time, they're just widgets.
It doesn't matter to the customer what you use to build there tool. Personally if I am working on a big project I will code it because it makes it less dependent. For ex if something goes wrong with no code architecture my business also faces the impact.
No does can get 80% of the task done, thereby taking work off the developer's plate. However the 20% of the use cases that they can't handle take 80% of the developer's time to do with real code.
No code tools have their place. I think it can help kill the need for a full-stack do-it-all dev. Developers who want to ship/prototype faster, can focus on their strengths and use no code to backfill their skill gap.
For example, my core strength is native Android development and strong NodeJs skill. So I use Bubble for the dashboard of my recent projects and wrote Firebase functions that supplement Bubble limitations.
I can crank out performant native mobile apps faster than I can read any no-code manual on mobile app development. But Bubble and Webflow are a thousand miles ahead of my Frontend skill.
Every tool has its place. For some types of business, no-code tools allow non-technical entrepreneurs to set up the necessary technical support structures around their core offering without hiring a developer, which is super neat! Except for certain niches, however, no-code is always going to have a scaling problem, and doesn't allow total blue-sky creativity in solving your customer's problems.
My own saas was originally just targeted at devs, but I'm learning to embrace the no-code folks. They make great customers! Essentially they are sophisticated tool-users that are willing to invest time and money in learning new tools.
Devs tend to be resistant to spending money on anything, since there is this feeling that "I could build this myself." No-code folks understand the value of a good tool, and are happy to pay for it.
As a developer, I think using them to prototype an idea is great, but you will be limited by what they can do as a long-term solution. If your idea takes off, you can always redo it in code and move users to the new system.
I think that's where there is heartburn... "redoing" things. I get the reasons for nocode getting to market quick though. It's a tradeoff.
Yes, the goal would be to test an idea quickly and then move to development to add additional features. As a developer myself, I always start coding but it really depends on what you are experienced with.
My main concern is "what happens with testing?". Even worst: Is everything locked in a visual diagram on a remote server? After years of programming and maintaining code it is very hard to imagine a workflow without continuous integration, testing and a git repository
A whole different story is with "low-code". We need more tools to get a lot of work done with less coding
Great point, and is something that I think about too when discussing nocode. These platforms have a vested interest in making a walled garden.
This comment was deleted a year ago.
This comment was deleted a year ago.
I don't agree, kinda - I think that some well made no code tools can actually accelerate the learning of concepts or technologies, that helps to "invest in yourself" as you put it.
I used a backend as a service platform that accelerated my learning of the backend, by removing all the unnecessary complexities.
The same goes with Webflow that lets you deal with real CSS properties and can accelerate the learning for someone who might want to switch to "real code" in the future.
Studying technologies and frameworks from 0 often makes people overlook some important concepts regarding "the big picture" of how something is supposed to work. That's how you get questions on reddit or stack overflow from people studying programming for a year (and some hired junior developers too) among the lines of "what is an API?".
I think no-code helps bring down information barriers and it abstracts a lot of boilerplate code, although the current tools have many dumb flaws and are lacking some very basic functionalities (e.g. there is not a single app builder right now that is "good enough", but we'll get there eventually)
This comment was deleted a year ago.