I just had need of this, dropped it in - works great. The only slight issue I had is that if there's a tiny amount of overflow, you basically don't get to see the fade because it's designed to disappear as you reach 100% scroll. In practice, that means I've noticed single chars getting lopped off with no visible gradient. I don't think there's much way around this beyond adding a bit of right hand padding to the track
If there’s one thing I’ve learned these past couple of days, it’s to resist the temptation to jab the escape button and start waving my arms! I wonder how much of this cyclical self doubt / self congratulating I go through in my own thoughts without even realising it. If you could verbalise or articulate all the half thoughts, snatches of ideas, feelings and ruminations the human mind goes through on some tasks it might be even more bizarre (or could just be me)
Your post made me laugh because I experienced the same as you but the other way around. I switched from Claude to a multi model harness a couple of days ago and the first model I tried was GLM5.2.
I gave it some simple code porting exercises and watched dumbfounded at the reasoning, which was more like the ravings of a lunatic - but lo and behold, after much confusion and a dizzying number of eureka moments the task was completed very successfully.
I tried Kimi on a similar task, much faster, a little more reassuring somehow in its ramblings, also surprisingly good results.
To be clear, I’m not surprised the results were good because they’re not GPT or Claude, but because the line of reasoning was so bonkers. Coming from Claude, I was just not used to seeing this, but I’ll bet it’s just as nuts with the frontier models and we’re just not allowed to see it (I’m about to read the links you shared).
Agree wholeheartedly that transparency is of grave importance.
Strange, I managed to sign up with an email alias and pretty much always have my proton vpn turned on, never had any issues with Claude code or web - maybe I’m about to? :D
Been thinking of switching to a multi model harness anyway.
That the frontier models and subs are all big tech is probably what bothers me most about “AI” right now, but I’m bullish on advancements in the capabilities of local models. I suspect and hope that, in time, the field will level and we will have very capable local, offline models and the landscape will be much as it is now with subscription compute in the cloud for enterprise and self host / local first for indies / hackers etc.
I used Kagi for a bit but began to question what I was paying for, there are plenty of other search providers basically doing the same thing, DDG, Ecosia, Brave search etc. I mostly just use DDG now because search as a paid offering that scrapes results from other indexes doesn’t seem like good value
I love this! Great work, took me 13 mins to solve my first just now with no hints, that’s perfectly aligned with the time it takes to drink a coffee in bed ;)
I stress over this with my own website-for-work. If I make the developer’s version of my site, who am I talking to? Other devs. If I make the version that appeals to agencies and casual users, there’s a constant voice in my head trying to drag me back to something simpler, lighter, judging me for that threejs hero section. As with all things, I guess it’s a matter of finding the right balance. Web development sure is in a very strange place and transitioning hard right now - off topic but I’m seeing more and more people looking for work and fewer and fewer job postings, especially for freelancers like myself. But maybe I’m not advertising AI bot integrations hard enough.
Great article, the last couple of paragraphs made me laugh! I love the part about things not masquerading as something else and being honest about what they are.
I was wondering if PenPot (https://penpot.app) might be sitting pretty in this new agentic era, considering that they took the direction of designs being actual markup, unlike the canvas approach in fig - if that’s even something that interests them.
I have a similar story but with neovim (and for the same reason as the author; growing unease with big tech). Tried and failed to make the switch a few times but made a concerted effort to stick with it throughout a specific project and now it’s second nature. I found it useful to research idiomatic (n)vim ways of doing things whenever I’d get frustrated or feel I’d be doing something more quickly in VSCode and commit them to memory by using new commands a few times over. Right now I’d say search/replace is the only thing that’s still not as ergonomic for me in vim as it is in Code. What I do is visually highlight my search phrase, hit asterisk then :%s//replaced - I learned that you can omit the search pattern using this technique.
Anyway nvim and helix are both amazing and terminal editors are both cool and sexy, so why wouldn’t you? ;)
I've written a fair bit of React and dipped in and out of Vue, Solid and Svelte over the past few years - I like to check in with different frameworks periodically, call it curiosity or FOMO. The general sentiment of the article is that React is old and slow compared to Solid/Svelte etc. While that's true, how many apps really need to squeeze that extra performance out of their underlying framework? Not many, I'd guess. For example, the ubiquitous krausest benchmark tests [0] operate on thousands of rows and the margins in results are shrinking - just today I noticed that the latest alpha from Vue has made huge strides forward.
React cannot iterate as quickly as other, smaller frameworks because of its size, but I guess that could also be seen as a positive thing. Even so, things like the React compiler are clawing back performance, taking cues from Solid, Svelte et al and these frameworks become more alike all the time.
For me the choice comes down to how I reason about the code I write. As others have pointed out, React feels closer to metal than Svelte - I find it easier to reason about because there's less magic going on behind the scenes. I really want to like Svelte, but I just can't click with it at all and I find the documentation lacking in deep, 'under the hood' detail.
On the flipside, I find Solid's docs to be superb - in depth articles on their reasoning, differences to React, etc [1][2]
On the whole, though, I find all these frameworks to be pretty good and what you can build is unlikely to be hamstrung by your choice in any way; though of course, React has a huge community behind it that you can't ignore. For hobby projects, try them all out - I had not worked with Vue 3 at all until recently, I just picked it up to try making a drum machine with the lovely Elementary [3] DSP library and I am really enjoying it! I hope we continue to see lots of development of all these frameworks and new ones pop up, because it's very clear to see how they all feed off one another, and that's good for everyone.
Shout out to Alpine.js [4] which flunks those benchmarks every time but remains my go-to for sprinkling reactivity in 'regular' websites.