> I know this is such a stereotypical "get off my lawn" statement but we've lost the art of software engineering.
Indeed. I am sure many of us here are burnt out on bloat. I am also sure many of us want to move to smaller stuff but cant simply because of industry momentum. BUT that doesn't mean the dream is dead, only that we must work towards those goals on our own time. I found Plan 9 and haven't looked back. I can rebuild the entire OS in seconds on a fast machine. Even my little Celeron J1900 can rebuild the OS for several supported architectures in minutes. I can share a USB device seamlessly across my network, PXE booted from a single disk without installing anything. Cat(1) is just 36 lines of C.
There's still hope. Just ignore the industry hype noise and put in the effort ourselves.
And just when we think we can't make software any more inefficient, slow, and bloated, they release things like Electron, where you ship an entire browser with your app! And then when we think it can't even get worse, we have Docker and containers where we ship the entire OS with the application.
I'm looking forward to when app developers ship you an entire computer in the mail to run their text editor.
The problem with Electron is that business-wise it is an excellent decision. You can get by with a few people to wrap the web app and integrate it with the OS, and then get updates pretty much for free.
Yet for the user it is bad -- bloated, slow, feels non-native, has specific bugs which are hard to address for the devs, etc.
I don't see any light for the desktop UI development unless there is some lightweight universal rendering engine. Tauri with WebView is somewhat promising, but it has problems on Linux and it is hard to target older systems.
It's a pretty OK example of a negative externality. A little like polluting: Just dumping your waste into the environment is business-wise an excellent decision. You avoid the cost and everyone else has to deal with the downsides.
Polluting is indeed an excellent business decision. The thing about apps is that all of them are polluting, just some of them are worse than others. And we tend to fill all available resources, so over time it only gets worse.
It's an excellent business decision... right up until your customers abandon you because you make bad quality software. Like many businesses have found time and again, deliberately sacrificing quality for profit is a short term gain for a long term loss.
there are quite a few examples of software built with electron that have very large user bases. this sounds like a personal vendetta against electron rather than meaningful insight.
Indeed. I am sure many of us here are burnt out on bloat. I am also sure many of us want to move to smaller stuff but cant simply because of industry momentum. BUT that doesn't mean the dream is dead, only that we must work towards those goals on our own time. I found Plan 9 and haven't looked back. I can rebuild the entire OS in seconds on a fast machine. Even my little Celeron J1900 can rebuild the OS for several supported architectures in minutes. I can share a USB device seamlessly across my network, PXE booted from a single disk without installing anything. Cat(1) is just 36 lines of C.
There's still hope. Just ignore the industry hype noise and put in the effort ourselves.