The Neverending Story(garrettdimon.com) |
The Neverending Story(garrettdimon.com) |
- Web standards didn't move fast enough for developers and users. People _wanted_ things like games, streaming videos and online shopping, which were impossible to do properly with web standard technologies for much of the 90s and 2000s. This wasn't driven by "big corporations" and Macromedia was never a big corporation anyway.
- How have React, or Angular for that matter, "failed"? How are they things not to bet a career on? React remains the default choice for most web app development as it has been for over a decade now. And I still see plenty of Angular jobs, far more than are looking for web components, the supposed standardised alternative.
- Web standards are _not_ decentralised in any way, shape or form. They are organised through a handful of centralised organisations, mainly the W3C, which are controlled by their major funders. Who are, surprise surprise, large corporations. Yes, the are somewhat open to outside participation, but it still is browser vendor employees who do most of the key work and make the key decisions, with Google/Chrome being the dominant player.
- Many, if not most, web standard technologies are invented by small numbers of engineers working for large corporations with no different incentives than those working on the technologies the author disparages. Canvas and CSS animations were created by Apple. Web components were created by Google. They were then adopted and standardised by other corporate employees.
- What "churn" have React and Angular really suffered? React was released in 2013 and has seen only one really major change — Hooks — in that time. Otherwise it has been a remarkably stable base to build software on. Most of the "churn" has actually occurred at the _community_ level, through the rise and fall of various open-source tools and libraries.
- The progress of web standards has not been the consistently slow but irresistible tide the author claims. Progress in the late 90s and 2000s was essentially nonexistent, being held back by Microsoft/IE's lack of interest and then the huge mass of users stuck on outdated browser versions. Then, once evergreen browsers reached a critical mass, progress sped up enormously, and in the modern era it easily outpaces the supposed churn of frameworks like React and Angular.
That hit home :-) Great quote
Flash was surreal in how good it was: a fantastic editor, language, and format for building and distributing hermetically encapsulated containers of multimedia greatness.
We need another Flash. It was S-tier and too good for us.
By contrast, Javascript runs without any action from the user. Consequently, web designers have grown to rely on it always being there, and now many sites rely on it to do things that can be done in pure HTML.
And oh by the way, they also violate my privacy something fierce.
None of that emerged again when flash got killed and the reason is technological - comparatively current technologies sux for those people.
Even more, it was great at turning people who weren’t Flash developers—or even developers at all— into Flash developers, by giving them an easy to use toolset that is still unmatched today.
2. This point is extremely funny:
"Every single programming language and framework in the world is the result of someone or some corporation believing they can do better."
... as we're looking at web which is literally a badly-designed and ill-fitting mish-mash of things someone (and in recent years increasingly one single corporation) decided they can do better
It was so easy and accessible, and there is no modern comparison.
If flash was no more capable than what a browser can do with web standards, its a tooling problem for the people who "skill is primary art being able to create cool and great things".
Then the question becomes, why hasn't anyone created tooling on par with what existed for "flash developers" that instead of generating actionscript to be executed by the flash plugin, generated javascript, webassembly, interaction with canvas elements and webgpu. Is it simply because no one sees real business opportunities there? (/ if there is a real growable business opportunities, then why is no one pursuing it?)
If that's the case, then yes, we lost something with the move away from flash (i.e. the inability to use all investment in said legacy tooling that worked, but can no longer generate usable things), but one can also view that as the reason for wanting web standards.
When one is using tools that generate things to web standards, it doesn't matter if those tools become "legacy", if they still work for you, there's a stronger likelihood that they will continue to generate an output into the future that is usable as they are not dependent on a proprietary system to run them, even if the "business opportunity" for the developer of the tools evaporates and they no longer desire to invest in them.
> why hasn't anyone created tooling on par with what existed for "flash developers" that instead of generating actionscript to be executed by the flash plugin, generated javascript, webassembly, interaction with canvas elements and webgpu.
First, I think there is aspect of "javascript, webassembly, interaction with canvas elements and webgpu" being quite hard to generate in a way that would achieve the same as flash did. As API, they comparatively sux although they are improving.
What you suggest is much harder then it looks like and involved technologies are not making it easier.