I found it strange that StarTalk hadn't been mentioned yet. Neil deGrasse Tyson is my favourite science communicator.

An who can forget Lord Nice!

Can't wait for my son to try this. He has general coordination issues due to a brainstem injury but the eyes are probably the part of the body he can better control. I'm not a fan of Apple's software and didn't have a great experience with the Vision Pro but I am excited to try it out.

Also not available in the UK for some reason.

But available on live TV in the UK (BBC1).

HN crowd doesn't know what linear TV is any more.

Live on iPlayer too (as live as iPlayer gets - 30 seconds or so behind DTT)

BBC gotta get that TV licence fee!

I get that "Web 3.0" is catchy as "Web 2.0" was back in the day but I wish people would stop trying to version "the web". It doesn't make any sense.

Consider it like era's/decades. It doesn't have to be perfectly accurate but it gives you a rough idea of a point in time. When I think 'web 2' I think of Digg, Facebook, social media. At the time it was more of a buzzword but in retrospect it defines that era. It doesn't have anything to do with 'versioning' the web.

100%.

The web evolved constantly from birth and became more interactive over time.

I always hated that silly “web 2.0” branding.

There's still a chance as some media are painting it as a good security measure.

Maybe, until you need to send those docs somewhere.

Beware:

https://www.manchestereveningnews.co.uk/news/greater-manches...

I agree. It will take some time until we have a Rust framework that is as robust and feature-rich as Django or Rails.

Rails was released 16 years ago (according to wikipedia). Rocket just got the first release that builds on stable rust. Thus: Give it 15 years, we'll be there!

I used Django from pretty much it's first public release and it was broadly feature complete back then.

The ecosystem between then and now is night and day, though, and that's an extremely important part of a framework.

In case of Django, I actually don’t feel much difference in this regard—it’s always been very much batteries included. It is natural for a project to have no dependencies but Django (and maybe psycopg2 bindings), business domain-specific dependencies excluded. Happily used it since before the first major release.

It had been in production use for a while prior to its public release, yeah?

IIRC, it was (like Rails) extracted from an existing web application.

Yes. I believe it was spun out of an internal project at Lawrence Journal-World.

> Rocket just got the first release that builds on stable rust.

As far as I can tell, Rocket is still on 0.4.5.

Your comment and the comment you are replying to are both correct.

Rocket's current released version is 0.4.5, and that version builds with a stable Rust toolchain.

It looks to me like 0.4.5 still requires nightly:

  error: failed to run custom build command for `rocket_codegen v0.4.5`
  
  Caused by:
    process didn't exit successfully: `/home/vlad/code/test-rocket/target/debug/build/rocket_codegen-f5b5d853ac3ddd89/build-script-build` (exit code: 101)
    --- stderr
    Error: Rocket (codegen) requires a 'dev' or 'nightly' version of rustc.
    Installed version: 1.47.0 (2020-10-07)
    Minimum required:  1.33.0-nightly (2019-01-13)

Edit: formatting

Indeed. Apparently the support for stable Rust was merged onto master months ago, but there hasn't been a release since then. 0.5 should be the first version that builds on stable Rust according to this thread: https://github.com/SergioBenitez/Rocket/issues/19

You can use 0.5.0-dev if you want stable rust.

I want to caveat this by first saying I’m a fan of Rust in general, but does one not see the irony in the article’s claim to the effect “yes, it’s ready as a Rails replacement”, yet requires you to use an unreleased dev version of a library to even build from a stable release of the language?

Yeah, and I don't know how anyone could claim its a Rails replacement. Diesel, warp, rocket et al are great but the DX and feature completeness aren't there yet.

That said, with Cargo using the unreleased version is as simple as changing `rocket = "0.4.5"` to `rocket = { git = "https://github.com/SergioBenitez/Rocket" }` so it's not like you even have to clone it manually. You can even specify branches if you urgently need a PR and patch entire workspaces with forked versions and so on.

Don't bother waiting. By that time there will be another set of frameworks.

I kinda disagree, I think people are more pessimistic about fracturing than is warranted. For greenfield work, it seems like people have mostly converged on React and Vue, which are conceptually fairly similar. State libs have changed, but again the concepts translate pretty well. Similarly, webpack is the lingua franca. Years ago, we had a new build tool every week, and multiple competing frameworks (backbone, angular, and ember) with very different approaches.

I don't think any One True Way will emerge. Svelte is coming out to challenge React and Vue, Alpine+Laravel LiveWire/Phoenix LiveView challenges the concept of needing a front end framework, WASM is challenging needing Javascript at all, and Snowpack/Parcel/Rollup are coming for Webpack's crown.

You have the option of not using these sort of tools and styling the code yourself.

You might find this project more appealing: https://news.ycombinator.com/item?id=22573743

Thanks! Still learning about it, but I think I'll join.