The roadmap for web tech is in limbo, riddled with over-bloated,
semi-open source, toolings of toolings, enterprise cloud service traps,
and all encompassing front-end first and last libraries that are always
in canary experimenting with other people's live apps and the developers
that use them causing regular app rewrites. Things like MVC, design
patterns, separation of responsibilities, abstraction, resuability, code
once deploy everywhere are non-existent. We choose stacks purely based
on popularity, and always and code the same sign-in form page from the
ground up for lack of a better option. The learning curve for web
development has gone full pill. It takes 4 times longer to do anything
because there a lot of ops. It started with just server ops, then added
dev ops, now it's ops of ops that manage tools of tools. There are so much
tools of tools that there are now automation of tools, which in fact is
another tool in itself. And 22 years later, we still get auto-scaling
on AWS wrong. Cloud services are built on top of other cloud services
or combine two or more services together in order to charge more
for the use of that service. Then services are created on top of these
cloud service derrivatives by 3rd parties charge even more. These
services do provide "open source" libraries that work with their service
but, requires a download of private code in order to use them. They built
their libraries using other open source projects, but noone can build on
top of their libraries because it's built around a lot of secrets. Web
development is exponentially harder today and bares the huge risk of
learning time and death months later. A library designed to temporarily
shim web components in 2013 beat native web components today due to
better marketing overtime. Then that same library that invented front-end
first, and destroyed a generation from learning MVC and good design
patterns decide to force their community to write components on the
server and claim it's stable, but you still can't do a simple server
redirect to a dynamic URL after a form submission. With this new
direction of putting components on the server, a framework built on top
of this library decides to add another highly opinionated routing system
into their core instead of just creating another framework to save
marketing costs, and now causes a simple hello world to require
200 MB of packages minimum and takes no responsibility on the surcharges
that get billed when a serverless function exceeds the free tier. It
all went downhill when Oracle changed the MySQL license.
There is absolutely nothing wrong.