I find it challenging to take this guy seriously. In this particular video, he first claims to have extensive experience writing Ruby code, yet he seems to struggle with understanding a basic Rails CRUD example. I even checked his GitHub profile for past noteworthy contributions but was again disappointed. Is my critique too harsh, or is he just not accurately representing his expertise? Also did you know he used to work at twitch? I am sure but he may have mentioned it before.
Most serious JS engineers I've spoken to regard Theo/t3.gg as a social media celebrity. His priority is creating a brand and selling products, not really giving valuable advice. If you watch enough of his videos and look at this Twitter, this becomes evident quickly.
I also don’t release very much of my code to the public.
Just looking at the merits though:
The article author is clearly a medium article programmer that bounces from one thing to the next to the next to the next depending on which medium article he happened to read most recently.
Theo demonstrates that I have to take what he says with a grain of salt, as he is unquestionably wrong to suggest that sql in the presentation is fine for anything but quick and dirties (I’d suggest that even quick and dirties is a dangerous proposition for sql in the presentation. Use Jupyter or something similar for quick and dirties).
Neither of these two are people I would ever find myself listening to.
I agree that the quantity of publicly available code isn't the most reliable indicator of someone's seniority.
My issue with this individual arises from the discrepancy between his public claims of significant expertise in the content he produces. He positions himself as a highly experienced developer, attracting a large following of junior developers who take his advice at face value.
I am trying to collect data points supporting his claims of seniority. For instance, his website prominently features a statement that he is the creator of the T3 Stack. However, a review of the contributor statistics for the T3 Stack https://github.com/t3-oss/create-t3-app/graphs/contributors reveals minimal contributions from him, which raises questions about the validity of his claims.
He seems a bit biased maybe?
Its fair that a rails controller does not exactly screams what its doing (implicit view rendering, routing etc).
But then, for someone not deep into react, a magic string like "use server" also does not clarify anything.
Also, he argues that just because you can it does not mean you should scatter sql queries around your views. But again, just because you can do a bunch of implicit stuff on a Rails controller it does not mean you should.
Yeap. I had the same feeling. If he never mentioned that he worked with Rails before I would understand not being familiar with the Rails concept of the "New" and "Edit" actions, but he mentions a lot of time throughout the video that he worked with multiple Rails backends and he doesn't even have familiarity with the most simple concept of Rails Controllers and Routes?
I felt that he has a bit of an agenda against DHH and I guess that's why he thrashes Rails as much as he did on that video.
Yeah this video is just cheesy strawman arguments. Make a claim that's not true then argue passionately about why it's a bad idea.
Like droning on about scaffolding without acknowledging that it's just get you going training tool that you're expected to grow out of almost immediately.
Or bemoaning that Rails is "framework" and so you're forced to use all of it. It's just reasonable defaults you can opt out of really easily.
e.g. swap out ActiveRecord for Sequel or replace ActionController with GrapeAPI.
Don't like backend rendering? Just remove it. You can pick only the bits of Rails you want via Railties.
Even going on about how you're forced to put business logic in models is pretty silly. It's right there in the Rails Guides how to keep models only for data access and move business logic into business logic into pure service classes. Don't like that? Use modules or engines or Trailblazer or any number of alternatives.
