Follow

We don't "browse" the web anymore. We don't view "web sites" either. Let's face it, we run applications in what is essentially a bloated runtime with enough features to be its own OS.

@ryan659 Indeed. Main usage of the Linux kernel is to boot a browser which runs JavaScript. Funny if it wasn't reality.

@ryan659 Yes. That's for sure. Web browsers have grown into a rich client platform. There's a bunch of reasons for that, some better than others. Choice of technology aside, I still wonder whether this is all bad.

@z428 Web browsers have been on (then-) current platforms since the web has existed. But in that time the performance demands for web browsing have increased significantly. Does someone really need to write an application in the web, likely only optimised for one web browser (Chrome..) with relatively high resource requirements compared to a native application which doesn't need anywhere near as many features as a web browser provide?

@ryan659 Personal experience: People don't *want* to write applications on the web. They do so because in many ways it seems the easiest, sometimes the only feasible way if you need applications to be available from various (personal, corporate) networks, on devices of various kinds, on different operating systems in different versions. The web essentially has filled the gap left by the failure of platforms and approaches such as Java Web Start, rigid corporate firewalls (leaving HTTPS as in ...

@ryan659 ... many cases the only really working option for access to the outside world), little to no real "cross-platform" development tooling (try building and maintaining a rich, desktop-integrated, up-to-date application for MacOS, Linux, Windows with something that isn't web/electron these days...) and a few others, not even talking about mobile. To me, browsers being used as application platforms actually just shows one thing: There would have been a dire need for an easy-on, ...

@ryan659 ... platform-agnostic, rich client environment that also includes aspects such as networking and deployment to various systems, best of all in a "no-install" manner. It's a pity that apparently the web browser, even in 2020, is the only really way to do so. 😟

@ryan659 (Bottom line: Why didn't we manage to come up with something that is same as straightforward and easy to handle for a developer but less resource-hungry than current browser implementations...?)

@z428 The problem is how all-encompassing browsers have become. Do we really need 3D graphics, gamepad support in a browser? Push notifications? WebUSB?

@ryan659 Well. Everyone who did that before knows how messy things like CSS, JavaScript, ... are. Did we ever ask *why* people repeatedly and at large choose this platform to build desktop applications, games, ...? 😉

@z428 Just because they choose to use it does not mean it should be morphed into something which allows that. But, that is my opinion.

@ryan659 I agree with you. But the point is: What else should they choose, given tight budgets (especially in terms of time), limited amount of devs at hand and a wide range of platforms and devices to support...? Would they have any *better* choice at the moment? I doubt so; everything else is way more tied to a particular platform and leaves some aspects well-addressed in a web browser (notifications / communication ...) unaddressed. 😐

@z428 This is true, I suppose. It does mean we need to find something that isn't trying to morph what was originally a document format (HTML is just markup, ultimately) into something it just isn't. Maybe some day it will appear, but for now it just seems that turning one thing into something else isn't the best way forward.

@ryan659 I agree, yet it seems to boil down to three basic options: (a) We accept web browsers to be used as cross-platform, cross-device, cross-network rich client platform and try to make it as good and efficient as it can be. (b) We despise web browsers being used that way and come up with a much better solution that manages to attract a really "critical" mass of developers and users really quickly. Or (c) we keep living in the world how it "used to be" and accept that a vast ...

@ryan659 ... majority of people (devs and users alike) couldn't care less. (b) would be my favourite approach, unfortunately, here, even Google as large as it is has pretty much failed (Dart). FLOSS community will be even less able to do so. Which essentially leaves us with (a) and (c) to choose... 😉 What would be your suggestion?

@z428 I do agree with you. As much as I'd prefer it be otherwise, but at present a better solution does not exist.

@ryan659 Yes. And sometimes it's a bit confusing/frustrating: I see a lot of effort being put into making browsers more "lightweight" to adhere to that "old" understanding of the web being hyperlinked documents, but I see very few (no?) effort put into trying to provide an alternative to the browser for what it's mostly used right now - a rich client application development platform. 😟

Sign in to participate in the conversation
Fosstodon

Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.