6 ms·

"Imagine a small company trying to write their own web browser from scratch nowadays. It's just not possible! " Imagine the much more common case of a small co

"Imagine a small company trying to write their own web browser from scratch nowadays. It's just not possible! " Imagine the much more common case of a small company (or individual) trying to make a useful app that is available to everyone. If they can make it in a browser, it's possible, if they have to make it for every proprietary platform (iOS, Android, Windows, Mac, Linux...).... often impossible. I'm happy that web browsers can do a lot. Many of the "security nightmare" arguments don't make sense. It wouldn't be a security nightmare if, for instance, Safari could allow a MIDI keyboard to talk to the web browser as it does in Chrome (ignoring Sysex support, which is the security excuse for holding it back.... there is very little need for that in a browser). God forbid I want to do a piano learning app. This is just one of many examples that may not affect you, but affects some subset of apps. (there are tons more, that one happens to affect my app) People are quick to say "I don't need that in a web browser", but you could say that about just about anything, for instance camera support. I do look forward to a day when most apps run in a browser rather than having to be platform specific, or require giant teams of developers to be cross platform.

timw4mail 5y ago
The more features in browser engines, the fewer platforms the browsers work on.
- robbrown451 5y ago
  How so? That doesn't make sense. I mean, if a platform doesn't support something, I wouldn't expect the browser to support it. Even if they decide that it is too hard to support a particular feature on a particular platform, doesn't mean they have to simply abandon the platform. They just don't support the feature.
- Dah00n 5y ago
  The most feature rich browser engine (Chromium) is the one with most adaptation across all the most used platforms so that doesn't make sense unless you count users from something like AmigaOS and TVs and in that way get Firefox out on top by purely counting platforms. Otherwise Chromium on Windows, Mac and Linux beats anyone else by a mile.
  - ntauthority 5y ago
    Even for platforms that Chrome can't easily be ported to, WebKit itself has a very simple and actually reasonable porting API which led to it for example being used on Sony PlayStation platforms, but also being the most important browser engine on the Haiku OS. In fact, Chrome originated as - and a lot of its code structure post-Blink-fork still signifies it as being - a meta-platform of its own with a WebKit port to it.
    - AnIdiotOnTheNet 5y ago
      > but also being the most important browser engine on the Haiku OS May not be the best example since Haiku's browser is often cited by people as a big reason they can't use it as a daily driver.
AnIdiotOnTheNet 5y ago
Warning, unpopular opinion incoming: Cross platform is, frankly, overrated. Depending on the market you're aiming for you have a very limited set of platforms that matter and directly targeting a native look and feel, as well as accounting for hardware and OS constraints, will always produce a better experience. Games? Target Windows and the consoles (web isn't a "real" solution on them anyway!), test on Proton if you care about Linux, and Mac isn't worth your time. Office software? Windows. Any kind of appliance (emulation station, kodi, etc), Linux. Developer tools? Linux and Windows. Bullshit addictive eyeball farming social media platform? Mobile. Artsy stuff? Mac-first and probably also Windows if you feel like it. Even so, people were targeting many more platforms than exist today throughout the 80s and early 90s with native ports for each (or a custom VM target, see: SCUMM, Z-machine, Another World/Out of This World) and they had to deal with assembly and a lot fewer standard APIs and no universal toolkits. And they did it with fewer developers, worse tooling, and no Stack Overflow to copy-paste from. I don't really think having multiple native ports is the herculean task so many developers make it out to be.
- robbrown451 5y ago
  I think your opinion is unpopular for a reason. You haven't given any reason why an app developer would want to rule out a subset of their potential audience. You seem to feel that users either a) have money to spend on redundant devices or b) fall neatly into groups. In my case, I do music apps. They are educational, fun, and inspire creativity. They run best on a computer (screen size, speed being the main reasons) but work on mobile. There is zero reason I can think of that I should pick Mac vs. Windows. If they are running on a phone or tablet, there is zero reason I should pick iOS or Android. So why should I have to pick just one? How is it better that, for technical reasons, I should be subdividing users like this? Your unpopular opinion doesn't really cover this.
  - AnIdiotOnTheNet 5y ago
    I did give a reason, maybe it was too subtle: the majority of your market is on one or maybe two devices and focusing on a better experience on those devices will make your users' lives better as opposed to cobbling together some compromised jank interface in the name of capturing a few outliers.

timw4mail 5y ago

The more features in browser engines, the fewer platforms the browsers work on.

robbrown451 5y ago
How so? That doesn't make sense. I mean, if a platform doesn't support something, I wouldn't expect the browser to support it. Even if they decide that it is too hard to support a particular feature on a particular platform, doesn't mean they have to simply abandon the platform. They just don't support the feature.
Dah00n 5y ago
The most feature rich browser engine (Chromium) is the one with most adaptation across all the most used platforms so that doesn't make sense unless you count users from something like AmigaOS and TVs and in that way get Firefox out on top by purely counting platforms. Otherwise Chromium on Windows, Mac and Linux beats anyone else by a mile.
- ntauthority 5y ago
  Even for platforms that Chrome can't easily be ported to, WebKit itself has a very simple and actually reasonable porting API which led to it for example being used on Sony PlayStation platforms, but also being the most important browser engine on the Haiku OS. In fact, Chrome originated as - and a lot of its code structure post-Blink-fork still signifies it as being - a meta-platform of its own with a WebKit port to it.
  - AnIdiotOnTheNet 5y ago
    > but also being the most important browser engine on the Haiku OS May not be the best example since Haiku's browser is often cited by people as a big reason they can't use it as a daily driver.

AnIdiotOnTheNet 5y ago

Warning, unpopular opinion incoming: Cross platform is, frankly, overrated. Depending on the market you're aiming for you have a very limited set of platforms that matter and directly targeting a native look and feel, as well as accounting for hardware and OS constraints, will always produce a better experience. Games? Target Windows and the consoles (web isn't a "real" solution on them anyway!), test on Proton if you care about Linux, and Mac isn't worth your time. Office software? Windows. Any kind of appliance (emulation station, kodi, etc), Linux. Developer tools? Linux and Windows. Bullshit addictive eyeball farming social media platform? Mobile. Artsy stuff? Mac-first and probably also Windows if you feel like it. Even so, people were targeting many more platforms than exist today throughout the 80s and early 90s with native ports for each (or a custom VM target, see: SCUMM, Z-machine, Another World/Out of This World) and they had to deal with assembly and a lot fewer standard APIs and no universal toolkits. And they did it with fewer developers, worse tooling, and no Stack Overflow to copy-paste from. I don't really think having multiple native ports is the herculean task so many developers make it out to be.

robbrown451 5y ago
I think your opinion is unpopular for a reason. You haven't given any reason why an app developer would want to rule out a subset of their potential audience. You seem to feel that users either a) have money to spend on redundant devices or b) fall neatly into groups. In my case, I do music apps. They are educational, fun, and inspire creativity. They run best on a computer (screen size, speed being the main reasons) but work on mobile. There is zero reason I can think of that I should pick Mac vs. Windows. If they are running on a phone or tablet, there is zero reason I should pick iOS or Android. So why should I have to pick just one? How is it better that, for technical reasons, I should be subdividing users like this? Your unpopular opinion doesn't really cover this.
- AnIdiotOnTheNet 5y ago
  I did give a reason, maybe it was too subtle: the majority of your market is on one or maybe two devices and focusing on a better experience on those devices will make your users' lives better as opposed to cobbling together some compromised jank interface in the name of capturing a few outliers.