The thing is that Electron apps don’t even look good compared to native apps. They’re slow and janky and, once you’ve seen a few of them, your impression is “the company didn’t care enough to build a native app”. In that sense, an Electron webpage in an app has the same connotations as AI artwork on a Substack essay: it looks slick if you’ve never seen one before, but cheap and shoddy if you know what it is.
I’m still convinced Electron only exists because there’s a huge surplus of mediocre web devs.
Electron solves hardly any problem that QT, GTK or all those other UI frameworks didn’t already solve 20 years ago. But for QT you need at least a few developers with passing knowledge of something other than js and css. And those guys are expensive.
OR, it is a huge conspiracy by Micron et al to increase demand for memory modules.
If you’re talking about an app that exist solely as Electron, then you might be right. But the primary benefit of Electron is that you can distribute your already existing webapp as a downloadable app, which reduces the amount of maintenance significantly.
Also, when it comes to UI diversity and customization, nothing beats HTML+CSS.
And as you mentioned, there’s a looot of webdevs. Electron empowers those people to easily create applications. Which they did, they created many useful apps. An application that isn’t perfect resource usage-wise is often much better than no application at all.
Think of Minecraft. Java is arguably the worst language to use for a chunk-based 3D game. But it’s still better than no Minecraft at all.
Writing the same app in Javascript for a DOM browser is quicker and easier than writing it in a native GUI toolkit, as you don’t need to think about life cycles or memory management as much. Of course, nothing comes for free, so the cost is borne by the userbase needing more RAM and faster CPUs to get the same work done. Which is philosophically in the same tradition of offloading negative externalities as dumping toxic waste in rivers.
I mean “didn’t care enough” == pay an entire separate dev team to develop for potentially each OS (at least windows probably) in languages unfamiliar to the web team. Then separately maintain that desktop app in perpetuity while the backend changes and drifts. This all for a vanishingly small subset of the userbase that could have used the website with identical functionality. It’s a similar issue with mobile apps.
I fully see the appeal of electron apps even if it’s not perfectly native feeling. I mean vscode is pretty universally liked n that’s electron. Also, it being a web app is a huge boon for anyone needing a code editor in browser.
Perhaps frameworks like Tauri will supplant electron soon enough for better performance and native feel.
The thing is that Electron apps don’t even look good compared to native apps. They’re slow and janky and, once you’ve seen a few of them, your impression is “the company didn’t care enough to build a native app”. In that sense, an Electron webpage in an app has the same connotations as AI artwork on a Substack essay: it looks slick if you’ve never seen one before, but cheap and shoddy if you know what it is.
I’m still convinced Electron only exists because there’s a huge surplus of mediocre web devs.
Electron solves hardly any problem that QT, GTK or all those other UI frameworks didn’t already solve 20 years ago. But for QT you need at least a few developers with passing knowledge of something other than js and css. And those guys are expensive.
OR, it is a huge conspiracy by Micron et al to increase demand for memory modules.
If you’re talking about an app that exist solely as Electron, then you might be right. But the primary benefit of Electron is that you can distribute your already existing webapp as a downloadable app, which reduces the amount of maintenance significantly.
Also, when it comes to UI diversity and customization, nothing beats HTML+CSS.
And as you mentioned, there’s a looot of webdevs. Electron empowers those people to easily create applications. Which they did, they created many useful apps. An application that isn’t perfect resource usage-wise is often much better than no application at all.
Think of Minecraft. Java is arguably the worst language to use for a chunk-based 3D game. But it’s still better than no Minecraft at all.
Writing the same app in Javascript for a DOM browser is quicker and easier than writing it in a native GUI toolkit, as you don’t need to think about life cycles or memory management as much. Of course, nothing comes for free, so the cost is borne by the userbase needing more RAM and faster CPUs to get the same work done. Which is philosophically in the same tradition of offloading negative externalities as dumping toxic waste in rivers.
I mean “didn’t care enough” == pay an entire separate dev team to develop for potentially each OS (at least windows probably) in languages unfamiliar to the web team. Then separately maintain that desktop app in perpetuity while the backend changes and drifts. This all for a vanishingly small subset of the userbase that could have used the website with identical functionality. It’s a similar issue with mobile apps.
I fully see the appeal of electron apps even if it’s not perfectly native feeling. I mean vscode is pretty universally liked n that’s electron. Also, it being a web app is a huge boon for anyone needing a code editor in browser.
Perhaps frameworks like Tauri will supplant electron soon enough for better performance and native feel.
The good electron apps are the ones you don’t realise are using electron.