Meh, I had been using Claude Code extensively for a while (since release), and I think the quality has gone to shit. I have no data to back up this claim, so it might be placebo.
GLM 5.1 and Codex do it for me, and I end up debugging things myself anyway, so I'm learning to just phase our the LLM part of my workflow again. Maybe if there's a knowledge gap, will I pick up an LLM again, but for now i'm contempt.
I love to explore, I don't love to work, and I hate to work for someone dumber than me.
The work experience I've had was that CEO's/Founders are genuine retarded when it comes to knowing things about their own company, let alone the industry.
When there's a gap, they hire someone to do the work for them.
This doesn't apply to all leaders, but definitely to most.
Also a lot of recent features are AI related and rely on talking to Adobe servers, which would require a valid subscription. They're probably betting the AI features are valuable enough that local only pirated copies aren't a threat long term.
I love this pattern, but I always felt like squeezing idiom B into idiom A when doing this in e.g. TS. I never tried effect-ts, so maybe I'm missing out. Anyone have experience in this?
Now is also a good time because they have a discount offer this month for using GLM-5-Turbo. During off-peak hours, only 1x multiplier will be deducted (otherwise it is 2x). I’m am on the Lite plan and have never maxed my usage quota (their Christmas deal offered 3 month for 7$).
I have been very disappointed in the Lite plan over the last few months. It started great, but they are obviously quantizing and cutting costs on the low end plans. The agents go into bad loops and contradict themselves, inject chinese characters, etc. There is obvious compression happening which makes it unreliable and unsuitable for serious work.
No, if a website is that obnoxious I just close the tab. It is not worth yet another mental drain on my limited attention span to read slop.
Just give me the boring single .htm page with your thoughts or a Wordpress site with minimal plugins. I'd hate to think the strain the author puts on people with accessible needs making this.
Many posts on this thread complain about how hard and unpleasant to read the article is. How is that not a valid complaint? And calling their "mindset" stupid... wow.
Embarrassingly badly generated article, with no real takeaway other than "I let an LLM dig into the code, here's what words it chose to describe EmDash".
> Joost put it well:
> It’s not a CMS with AI features bolted on. It’s a CMS where AI agents are first-class builders.
Joost asked ChatGPT what he should say about the CMS, and you felt like it was a good quote.
> Why I won’t use it
> I migrated to Astro partly to get away from the CMS.
Well then you never needed a CMS in the first place? I also don't need use a CMS for my site, but I still maintain a CMS for customers because they do need it.
> Does it solve the right problem?
This is the only thing I cared about from this article, and the answer is [bag of words]. Are people really this desperate to put their names on new tech? Is it an "I want to be included!" mindset that gets people to prompt an hour of their life away?
> Astro itself wasn’t an obvious success from day one.
Astro is just the framework they built on, what does this sentence have to do with EmDash? I'm so confused about what this article is trying to tell me.
Also, how come you did not write anything about what it was like when WordPress had just released? I'm sure there are enough people who can help out with that. Did it have competitors? I wouldn't know, I was eating sand when it came out.
To say Astro wasn't a success from day one is a truism. No JavaScript frameworks have been an obvious success from day one. How could they be? Even very well-designed and innovative frameworks and libraries struggle to gain adoption in such a crowded space where tooling as significant as a framework has major inertia. It really is a bunch of words.
GLM 5.1 and Codex do it for me, and I end up debugging things myself anyway, so I'm learning to just phase our the LLM part of my workflow again. Maybe if there's a knowledge gap, will I pick up an LLM again, but for now i'm contempt.
reply