It just so happens that overwhelming amount of them do what they should not do - create actual apps (webapps in reality). We could thank Google or frameworks for this, but ultimately their incompetence leads to situations like this. Webdevs thinking of everything as a platform for their stuff that should be working at all times. If they were actual devs they would build proper native apps, think more about how devices actually supposed to work, and rely on cloud less.
I get that many could be static pages, but you’re comparing web vs desktop. And in that case, I prefer web most of the time. Why? It works the same everywhere, and I can probably access it just fine on my phone without having to get their mobile app, which probably has fewer features and more telemetry.
Web doesn’t make sense for everything, but it’s far better than desktop apps for relatively simple use cases. If the app isn’t performance sensitive and doesn’t need to store a ton of data, web is my preferred platform, especially since I’m a Linux user and would likely need to run the app through WINE instead.
Not desktop. Native. You can build native apps for smartphones and pretty much anything. Web comes into play when you decide you won’t build native version of what you want. In some cases that is guided by thoughts like “I know web will fit this project. I know the platform and will remember to keep memory and internet usage low”, but in most cases it’s about “no idea what that startup is about, but I know some AngularJS and they said I can use AWS so that’ll do”.
Similar thing is happening now with Unreal Engine 5. The difference between devs and webdevs became very similar to the difference between coding and vibe coding.
and doesn’t need to store a ton of data
I know by data you mean “data I care about as a dev” but that should also include data that is actually processed and saved on user’s device. And webapps are notoriously bad at keeping their caches and data usage low.
There is no difference, webdev is someone who solves problems with web tech, mobile dev is someone who does mobile apps, embedded dev works with low level components, etc. It’s all development, and often the two will mix.
For example, I worked at a startup using C for embedded stuff, Go for the server stuff, and web tech for the FE. Rust wasn’t out yet, so C was the best option for interfacing with the board components (in this case a math module for our signal processing), Go was a good mix of performance and ease of training new devs, and a web FE was the lowest barrier to getting our customers using our product (basically a high end IOT device). We built a small native FE for certain simulations, and eventually moved it to a server with a web FE.
At my current role, we build a reporting and simulation app for a niche industry. We do everything from simple forms to 3D rendering to simulations that take hours to complete (most are 15-45min). Our customers use crappy laptops, so doing the processing locally isn’t an option (they probably don’t have enough RAM anyway), so we’re going to need a server. Because of that, we decided to build it as a web app. We still have native components (some simulations use C++, another was Fortran until recently, etc), and they’re maintained by Ph.Ds in our field because the hard part isn’t the coding (our JS specialists could handle that, they’ve built a 3D app in the last few months with complex transformations and calculations due to business login needs), but knowing the math behind it all, hence the researchers.
Not all web apps are overengineered crap because they hired a dev team to build a static site, there are apps like the two I mentioned that do interesting things and happen to use web technologies.
I didn’t say that. There are always well engineered apps and things. But few. Compared to that, a lot of mainstream desktop apps are now web apps for no good reason. The actual reason is webdevs not challenging themselves to become something else or at least better.
If I read your case correctly, it’s basically “customers use crappy laptops -> we decided to make them use web browsers” which sounds insane to me because web content IS the reason why tons of otherwise unnecessary upgrades are done in recent 10 years or so. Office guys can’t use Chrome with just 8 GB of RAM because it will affect their business performance.
Not that I don’t believe your case doesn’t contain other specifics that make web a right choice. And I don’t need to know more of that. It’s just how it sounded to me.
it’s basically “customers use crappy laptops -> we decided to make them use web browsers” which sounds insane to me because web content IS the reason why tons of otherwise unnecessary upgrades are done in recent 10 years or so
We do processing on the cloud because our customers use crappy laptops. We use a web browser because:
convincing customers to get approval for our app is hard (corporate IT), whereas they already have a recent browser
it’s faster to get the result we want
our dev env (macOS) is different from our target platform (Windows); this is because it’s easier to work around our IT Dept than with them (and nobody on the team likes Windows), and our servers are Linux, and macOS is close enough for BE dev
audits are easier - just need to test the API, not the whole desktop app platform since auditors trust the browser
The same is true for my the other project I mentioned. If your business is largely B2B, then browser apps are a much easier sell than a desktop app. We’re getting a little pushback as we’re breaking into a new market, so we’re turning to Tauri to provide a desktop-app experience, a different keep the web app for the rest of our customers.
What do you mean? Webdevs are devs, just within a specific platform. And like any dev, they can suck or be great.
It just so happens that overwhelming amount of them do what they should not do - create actual apps (webapps in reality). We could thank Google or frameworks for this, but ultimately their incompetence leads to situations like this. Webdevs thinking of everything as a platform for their stuff that should be working at all times. If they were actual devs they would build proper native apps, think more about how devices actually supposed to work, and rely on cloud less.
What’s wrong with webapps?
I get that many could be static pages, but you’re comparing web vs desktop. And in that case, I prefer web most of the time. Why? It works the same everywhere, and I can probably access it just fine on my phone without having to get their mobile app, which probably has fewer features and more telemetry.
Web doesn’t make sense for everything, but it’s far better than desktop apps for relatively simple use cases. If the app isn’t performance sensitive and doesn’t need to store a ton of data, web is my preferred platform, especially since I’m a Linux user and would likely need to run the app through WINE instead.
Not desktop. Native. You can build native apps for smartphones and pretty much anything. Web comes into play when you decide you won’t build native version of what you want. In some cases that is guided by thoughts like “I know web will fit this project. I know the platform and will remember to keep memory and internet usage low”, but in most cases it’s about “no idea what that startup is about, but I know some AngularJS and they said I can use AWS so that’ll do”.
Similar thing is happening now with Unreal Engine 5. The difference between devs and webdevs became very similar to the difference between coding and vibe coding.
I know by data you mean “data I care about as a dev” but that should also include data that is actually processed and saved on user’s device. And webapps are notoriously bad at keeping their caches and data usage low.
There is no difference, webdev is someone who solves problems with web tech, mobile dev is someone who does mobile apps, embedded dev works with low level components, etc. It’s all development, and often the two will mix.
For example, I worked at a startup using C for embedded stuff, Go for the server stuff, and web tech for the FE. Rust wasn’t out yet, so C was the best option for interfacing with the board components (in this case a math module for our signal processing), Go was a good mix of performance and ease of training new devs, and a web FE was the lowest barrier to getting our customers using our product (basically a high end IOT device). We built a small native FE for certain simulations, and eventually moved it to a server with a web FE.
At my current role, we build a reporting and simulation app for a niche industry. We do everything from simple forms to 3D rendering to simulations that take hours to complete (most are 15-45min). Our customers use crappy laptops, so doing the processing locally isn’t an option (they probably don’t have enough RAM anyway), so we’re going to need a server. Because of that, we decided to build it as a web app. We still have native components (some simulations use C++, another was Fortran until recently, etc), and they’re maintained by Ph.Ds in our field because the hard part isn’t the coding (our JS specialists could handle that, they’ve built a 3D app in the last few months with complex transformations and calculations due to business login needs), but knowing the math behind it all, hence the researchers.
Not all web apps are overengineered crap because they hired a dev team to build a static site, there are apps like the two I mentioned that do interesting things and happen to use web technologies.
I didn’t say that. There are always well engineered apps and things. But few. Compared to that, a lot of mainstream desktop apps are now web apps for no good reason. The actual reason is webdevs not challenging themselves to become something else or at least better.
If I read your case correctly, it’s basically “customers use crappy laptops -> we decided to make them use web browsers” which sounds insane to me because web content IS the reason why tons of otherwise unnecessary upgrades are done in recent 10 years or so. Office guys can’t use Chrome with just 8 GB of RAM because it will affect their business performance.
Not that I don’t believe your case doesn’t contain other specifics that make web a right choice. And I don’t need to know more of that. It’s just how it sounded to me.
We do processing on the cloud because our customers use crappy laptops. We use a web browser because:
The same is true for my the other project I mentioned. If your business is largely B2B, then browser apps are a much easier sell than a desktop app. We’re getting a little pushback as we’re breaking into a new market, so we’re turning to Tauri to provide a desktop-app experience, a different keep the web app for the rest of our customers.