We used Greptile where I work and it was so bad we decided to switch to Claude. And even Claude isn’t nearly as good at reviewing as an experienced programmer with domain knowledge.
My experience is that Claude or others are good at pointing out things I will want to look at and then I can go review more thoroughly. So it's helped to some degree.
But like everything else with it, it tries to do too much.
What I want is a review "wizard" agent -- something that identifies the pieces I should look at, and takes me through them diff by diff asking me to read them, while offering its commentary ("this appears to be XX....") and letting me make my own.
Fair point, the UI was partly AI-assisted, and the design is not our strongest area at the moment but something we are actively improving.
Appreciate the feedback.
I appreciate the prompt engineering effort, but it's a lot of claims and not much substance. Also, I wouldn't call Japanese used in anime "real Japanese", especially from anime like Attack on Titan. And how do you get around the copyright issue? Does this fall under fair use?
I like how the conclusion of the article paints a very positive picture for the language. It’s as if the entire negative tone of the article was a poor attempt to sound insightful for clicks.
This article is full of statements that are not backed up by any material proof. The author claims that Swift features no longer compose without giving a single example. I write Swift SDKs for a living and I have never thought that any of the recent features felt incompatible, incongruous, or redundant. Sure, each has its strengths and weaknesses, but that results in an expressive language where every problem has several solutions with different tradeoffs.
When it comes to governance, it’s not without wrinkles, but it’s not as bad as the author makes it sound. If anything, things got a bit less “dictatorial” since Chris left. Different core team members pull in slightly different directions which made the language more balanced in my view.
