What’s interesting is that it’s the averaging of many writers with voices which do sound human and enthusiastic and quirky sound like no personality at all.
There is an absolute obsession of using three variations for the message, the wording, and the life behind it. And the nonstop em dash! My favorite side effect of LLM slop is that it exposes people who have no respect for their readers, don't proof read their own works, and the life behind it.
Shouldn't the intro paragraph carefully, clearly explain the benefits?
This is the intro paragraph "Search is a browser for the Mac with nothing in the way. About ten megabytes, the engine already in macOS, and a window that opens in a third of a second. Free, for any Mac on macOS 14 or later."
I just think "with nothing in the way", is really lacking as the value proposition.
For me, simplicity is the value proposition. All I really want from my browser is for it to be lightweight, open source, good looking, and have basic organization (groups, bookmarks etc.). This checks all the boxes without any cruft.
The quality of the ad blocker and extension support are the main question marks for me, but they've been good enough so far.
Been using it with vimium for about 2 months. Other than not always capturing the focus on vimium fields it does what it promises on its tin: a minimal (as in resources) browser, a renderer of pages, that brought many of arc defaults
I love search’s aesthetic, its integration with iCloud Keychain, its minimalism… but dude, spaces was really unintuitive it made me uninstall really quickly :(
In my opinion spaces and profiles should be tied together, and they should be dots along the bottom. I’m also not sure why we need bookmarks in these side bar folder style browsers.
I hate the effect of design minimalism bleeding into functional minimalism. I also hate the low information density and undiscoverability that also often accompany it.
Say it with me: Form follows function. Form follows function. Form follows function. Form follows function!
We'd hate if a car's climate controls were behind an aesthetically beautiful door in the console
Really wish they'd brought over the "command pallette" interface from Arc for searching, browser actions, and switching tabs. IMO, this is even more important for navigation across fullscreen tabs than Arc's sidebar is.
Imagine thinking "I wish I could use my chrome extensions in safari" and instead of asking your agent to port the extensions over, you conclude that you should vibe code a new webkit browser with chrome extension support
hi, I’m Drice, I make Search. some context: I started it when Arc stopped moving, wanting Arc’s way of working (spaces, pins, split view, peek) in something native and small. it uses the WebKit already in macOS, so the app is ~10 mb and gets WebKit’s security fixes with macOS updates. Chrome extensions run through Apple’s WKWebExtension, with a compatibility layer for the APIs it lacks; most popular ones work, not all yet. it’s MIT, code at github.com/driceroland/Search. happy to answer anything, and bug reports are very welcome!
Yet another WebKit-based browser. Why? We already have Orion [0] that does exactly what this one does.
Also why WebKit? It has monthly memory corruption vulnerabilities being discovered.
Why not just improve Ladybird? [0] instead of creating a worse version of existing browsers that still suffer from weekly memory corruption vulnerabilities.
Many years ago when Brave had just came out and it was an Electron browser---and Electron had like 10 use cases world wide---I was so frustrated with browsers that I thought man I can fork this and just fix it for myself. It wasn't about a revolutionary thing like Ladybird or if Brave was good foundation; it was only the lowest hanging fruit for me, the only possible one back then. I couldn't even fathom compiling Chromium, not to mention changing it.
WebKit is designed to be easily embedded, is already compiled in the mac and has a very straight forward API and is a part of Apple world, so you just make the browser shell and what you want. That is probably the rationale behind 99% of WebKit browsers. Search was 6MB worth of download, you can't have that with Chromium, and I don't think if people have the patience to use a half done browser engine so that it can be finished---although I personally am counting down for the day I switch to ladybird.
Uhh.. I would argue that we don't have /enough/ Webkit based browsers. There are currently no Webkit browsers for Windows unless you want to run Something like Gnome web/Epiphany in WSL.
Kagi Orion looked promising, but they quit developing their windows version recently.
fair question. orion is great, and closed source. Search is MIT-licensed, ~10 mb, and built around arc’s way of working (spaces, pins, split view, peek, arc import), plus chrome extensions. on webkit: it’s the engine macos ships and patches with each security update, so Search gets those fixes without us maintaining an engine. that’s a big part of why we picked it.
“ Two ways to hold it. Titles across the top, or down the side. The grey slides to the tab you pick either way, and a pinned site is one letter.”
I really can’t stand LLM language.
Yes, it is riddles all the time. And /unslop doesn’t work too much.
Everyone who uses LLMs to write ends up sounding like a coke-addled gnome wizard.
(Particularly if they use Claude. Even getting Gemma to write the text would be an improvement)
What’s interesting is that it’s the averaging of many writers with voices which do sound human and enthusiastic and quirky sound like no personality at all.
There is an absolute obsession of using three variations for the message, the wording, and the life behind it. And the nonstop em dash! My favorite side effect of LLM slop is that it exposes people who have no respect for their readers, don't proof read their own works, and the life behind it.
Shouldn't the intro paragraph carefully, clearly explain the benefits? This is the intro paragraph "Search is a browser for the Mac with nothing in the way. About ten megabytes, the engine already in macOS, and a window that opens in a third of a second. Free, for any Mac on macOS 14 or later."
I just think "with nothing in the way", is really lacking as the value proposition.
For me, simplicity is the value proposition. All I really want from my browser is for it to be lightweight, open source, good looking, and have basic organization (groups, bookmarks etc.). This checks all the boxes without any cruft.
The quality of the ad blocker and extension support are the main question marks for me, but they've been good enough so far.
It should, but LLMs do not really understand value judgements about text like that.
Discussion on 24.sep.2026 https://news.ycombinator.com/item?id=49828120 32 comments
Also related recently: "Show HN: Zigzag, a native Mac browser for people leaving Arc " https://news.ycombinator.com/item?id=50496787 74 comments
Been using it with vimium for about 2 months. Other than not always capturing the focus on vimium fields it does what it promises on its tin: a minimal (as in resources) browser, a renderer of pages, that brought many of arc defaults
I love search’s aesthetic, its integration with iCloud Keychain, its minimalism… but dude, spaces was really unintuitive it made me uninstall really quickly :(
In my opinion spaces and profiles should be tied together, and they should be dots along the bottom. I’m also not sure why we need bookmarks in these side bar folder style browsers.
I hate the effect of design minimalism bleeding into functional minimalism. I also hate the low information density and undiscoverability that also often accompany it.
Say it with me: Form follows function. Form follows function. Form follows function. Form follows function!
We'd hate if a car's climate controls were behind an aesthetically beautiful door in the console
Really wish they'd brought over the "command pallette" interface from Arc for searching, browser actions, and switching tabs. IMO, this is even more important for navigation across fullscreen tabs than Arc's sidebar is.
there is command + K for a tab switcher, but adding other capabilities to that would be definitely be cool.
It is so minimal, it somehow feels not finished. A very strange feeling to have. I liked it, but there is much room for improvement.
Look clean, love it! Any plan for a Linux port ?
Such a nice respite from the Chrome bloat.
Imagine thinking "I wish I could use my chrome extensions in safari" and instead of asking your agent to port the extensions over, you conclude that you should vibe code a new webkit browser with chrome extension support
No AI integration? What is this, a browser for cavemen?
Open source, macOS only.
Made by design studio Office Commun
Looks good, but IMO it's too minimalistic to really be usable.
Absolutely hate merging URL and tab, for instance. I like to see what URL I'm on.
Switched to this for a week, the cold start time is real and the memory footprint stays low even after a few days with 20+ tabs
I'm posting this on it. It's actually quite a nice browser.
hi, I’m Drice, I make Search. some context: I started it when Arc stopped moving, wanting Arc’s way of working (spaces, pins, split view, peek) in something native and small. it uses the WebKit already in macOS, so the app is ~10 mb and gets WebKit’s security fixes with macOS updates. Chrome extensions run through Apple’s WKWebExtension, with a compatibility layer for the APIs it lacks; most popular ones work, not all yet. it’s MIT, code at github.com/driceroland/Search. happy to answer anything, and bug reports are very welcome!
Hi Drice Search. I can't install on my Macbook Pro Intel chip using macOS Sonoma 14.6.1.
Also, did you give Arc's baby, Dia, a try? I learned about it when they reached out to not hire me.
Yet another WebKit-based browser. Why? We already have Orion [0] that does exactly what this one does.
Also why WebKit? It has monthly memory corruption vulnerabilities being discovered.
Why not just improve Ladybird? [0] instead of creating a worse version of existing browsers that still suffer from weekly memory corruption vulnerabilities.
[0] https://orionbrowser.com/
[1] https://ladybird.org/
Many years ago when Brave had just came out and it was an Electron browser---and Electron had like 10 use cases world wide---I was so frustrated with browsers that I thought man I can fork this and just fix it for myself. It wasn't about a revolutionary thing like Ladybird or if Brave was good foundation; it was only the lowest hanging fruit for me, the only possible one back then. I couldn't even fathom compiling Chromium, not to mention changing it.
WebKit is designed to be easily embedded, is already compiled in the mac and has a very straight forward API and is a part of Apple world, so you just make the browser shell and what you want. That is probably the rationale behind 99% of WebKit browsers. Search was 6MB worth of download, you can't have that with Chromium, and I don't think if people have the patience to use a half done browser engine so that it can be finished---although I personally am counting down for the day I switch to ladybird.
I do like Orion as well, but it's substantially larger (580 MB vs 11MB for me) and isn't open source.
Anecdotally, I also like Search's look better and have found it more stable (though it's been a while since I really tried using Orion to be fair).
Orion is far more stable now in my experience. Worth looking at again.
It consumes less memory than Blink on macOS.
Uhh.. I would argue that we don't have /enough/ Webkit based browsers. There are currently no Webkit browsers for Windows unless you want to run Something like Gnome web/Epiphany in WSL.
Kagi Orion looked promising, but they quit developing their windows version recently.
fair question. orion is great, and closed source. Search is MIT-licensed, ~10 mb, and built around arc’s way of working (spaces, pins, split view, peek, arc import), plus chrome extensions. on webkit: it’s the engine macos ships and patches with each security update, so Search gets those fixes without us maintaining an engine. that’s a big part of why we picked it.