[blorenz]

You have the benefit of hindsight at this time. You can draw parallel to history of flight and all the crazy contraptions that people attempted. Great technology can emerge from the combination of numerous shit shows. The whole is greater than the sum of the parts.

[Twirrim]

> You have the benefit of hindsight at this time.

People have been pointing out it's a shit show with no end in sight for the entire duration of the phase. Pointing out the performance impact and cost to end users, how diabolical it is for those on lower latency or poorer network connectivity (i.e. most of the world), and so on.

Same thing as always happens with these pendulum swings, newer engineers come in convinced everyone before them is an idiot, are capable of building their new thing and hyping it up such that other newer engineers are sold on it while the "old guard" effectively says "please listen to me, there's good reasons why we don't do it this way" and get ignored. Worse, they'll get told they're wrong, only to be proven right all along.

I'm not denying there are obstructionist greybeard types that just refuse to acknowledge merits in new approaches, but any and all critique is written off as being cut from the same cloth.

It's perfectly possible to iterate on new ideas and approaches while not throwing away what we've spent decades learning ('Those who do not learn history are doomed to repeat it'), but tech just seems especially determined not to grow up.

[freeopinion]

I guess I've become a grey beard. I've done the whole journey from CGI everything to a bit of js to SPA. As much as I'd really like to be nostalgic about the good old days, there are reasons everything got pushed into the client. One of those reasons is maintaining state.

"HTML over the wire" isn't really a return to the good ol' days. It's still the client maintaining state and using tons of js to move data back and forth without page reloads. It just changes the nature of 1/2 the data and moves the burden of templating back to the server.

It is amusing that they make a claim that reads a lot like "eliminate 80% of your Javascript and replace it with Stimulus". What is Stimulus? A Javascript framework.

[combatentropy]

They mean JavaScript that you write.

[epistasis]

I'm not a front end engineer, but it always seemed crazy to me. I remember testing out the Google Web Toolkit when it came out more than a decade ago, and the craziest thing about it to me wasn't the Java --> JavaScript compilation, it was that the server just dumped an empty page and filled everything in with JavaScript on the client.

Then, remember the awful awful #! URLs? Atrocious, and seemed like obviously a terrible idea from the start, yet they spread, and have mostly died, thankfully. But even with the lessons from these bad tech designs, new frameworks come out that repeat mistakes, yet get incredible hype.

[filmor]

Hashbang URLs are gone because of the PushState API, not because people have given up on the idea.

[epistasis]

Hashbangs preclude even the option of the server even seeing the state that the client wants from the first request, necessitating severa hops. PushState, though it allows URL transitions without a full reload, is an entirely better and different idea.

[mruniverse]

Around the time that GWT came out, offshoring was a big thing. And most of the contractors only knew Java. Also Java was the trusted language and javascript was not.

[bryanrasmussen]

The only big GWT project I've ever been on was a governmental project that I won't go into (because it's Danish and I would have to describe a bunch of stuff that everyone in Denmark knows and nobody outside would probably care about), but the company providing it was porting their Java version to JavaScript and had a significantly large codebase to leverage.

[Twirrim]

AWS used to rely on JWT for their consoles (haven't for a few years now, most folks migrated away some 5 years ago)

It's why they used to be horrendously bloated with large javascript bundles that took so long to process on the client side.

Roughly speaking the idea was "We don't have any Javascript developers, but we do have Java developers. JWT allows us to bridge that divide". Neat in theory, and an understandable decision, but diabolical in practice!

[solarengineer]

I believe you mean “GWT” and not “JWT”.

[Twirrim]

Yup you're right, that was quite a brainfart :)

[Cederfjard]

Very true. I think that mostly, web dev mainstream has taken a rational path. It’s with the benefit of hindsight as you say, or the yoke of unusual requirements, that people now say ”we did it all wrong”.

[robmurrer]

except we could fly before ;D