Hacker News Reader

Comment by maxpert | Hacker News Reader

Reminds me of of good old PHP shared hosting days. It was so much fun to just write my code and then just drop the folder over Filezilla and everything just works!

Bug no problem! Live editing!

View on HN

As everything has become so complex people decided that the solution is to build containers that are even more complicated but you can move them around.

It starts to feel like there’s an appetite to try another way where you actually move around the stuff you created and not entire containers of your setup.

And the cycle begins again.

I'd assume that the end result of the drop is running in a container.

It seems to run only static content, it's going to be served by Netlify and only browsers will run the code. If that's really it, there's no need for containers here.

My thoughts exactly. This could be replicated by just taking the html, piping it to s3 bucket then create a dns record that points to it. No build nothing. Just html -> "server"

I have fond memories of the FTP side bar in Macromedia Dreamweaver. Press the push button and website gets instantly updated.

Last company I worked for pushed me on multiple occasions to install an FTP extension for VSCode that would upload files on save. Glad I got out, everything was a nightmare there.

I used to work like this in PHP at my first software job circa 2013. Wrote my own deployment tooling with all the warts you'd expect. Working this way is fine for simple setups but I'm glad I've moved onto saner workflows for my larger projects!

Oups, my connection string is visible by everyone

I know people that still work like this lol. Fixing stuff right in production.

It’s not that I want to, it’s that I’m forced to.

That's actually normal now. We've come full circle, but it's called "observability" - since you can't integration test the 100 microservices that you have (for no reason), just throw it into prod and see of anything turns red.

And "feature flags"...

I'd suspect it should be possible to test those 100 microservices. If you can do something, then you can do it twice, especially if your automation is good. If you can't, then maybe some of the systems came together in a non-reproducible or undocumented way and it's time to clean up a little. If you know your traffic you can build load tests that simulate production use.

Ahh feature flags, the new #ifdef.

Running tests in production doesnt mean you didn't run an array of other tests in pre production envs

Also, 'live editing' != testing in production.

You could find a bug in pre-prod and live edit it in prod

It's obviously not really suitable for businesses anymore, but if you don't have a personal site that you can do this on, I feel bad for you.

I am so glad I am not the only one. May be I will admit I am stupid. Because every time I think back everything was simple. It wasn't easy, some things was slightly tedious, but it was simple.

Now most of the time I dont have a clue why anything works. It is blackbox feeding into another blackbox which relies on another blackbox somehow we all agree that is how things should be done so we stick with it.

Reply on news.ycombinator.com