Store your files in a file structure that makes sense so you know where things are? I have never used Spotlight to find a file because I put files in sensible places.
I use Spotlight all the time to search for the contents of files. I don't memorize the contents and names of every file on my system, that's what my computer is for.
I want spotlight to open applications and system settings. But full disk indexing makes spotlight basically useless for that, because its index is filled with crap. Instead of opening zed I accidentally open some random header file that’s sitting around in Xcode. It’s worse than useless. And that says nothing of the grotesque amount of cpu time spotlight wants to spend indexing all my files.
A feature I never wanted has ruined a feature I do want. It’s a complete own goal. In frustration I turned spotlight off completely a few months ago. Rip.
I think it's been said in this thread already, but it sounds like what you want is Alfred https://www.alfredapp.com/ it's a great app, use it every few minutes every day.
I am also in OP's boat and, even though these are great suggestions, personally I would like to be able to do a basic thing such as opening an app with a built-in way rather than having to download yet another app to do that. Every major macos update I have to worry about spotlight reindexing things.
What I find really annoying with macos is that with stock/default settings it is the worst UX. You have to download an app to launch apps, an app to move and resize windows, an app to reverse the mouse's wheel direction to be the opposite of the trackpad, an app to manage the menu bar (esp decrease the spacing, so that you can fit items up until the notch). Then, you also need anyway to spend an hour tweaking settings and run custom commands (such as `defaults write -g ApplePressAndHoldEnabled -bool false` so that you can actually type stuff like aaaaaaaaaaaaaaaaaaaaa). These are just needed to make using macos bearable, and do not include any kind of "power user" kind of stuff.
I used to hate macos before getting my own mac, because I had to use some at work in their default settings and it was just a horrible experience.
Spotlight search relevancy is a complete joke. If only they did some embedding based search across the system and paid attention to basic precision recall numbers. This has gone from bad to worse quickly.
this is what grep is for. Why do I need a service constantly indexing my system and wasting resources for the few times a month I might need to run grep <string>?
Ah yes, grep's famous support of PDF and Office documents will surely come in handy. As will its image OCR capabilities.
Oh it doesn't have those things so is a non-option.
Tahoe in general sucks but Spotlight has been a pretty good local search for nearly twenty years. The image OCR added in Sonoma made it even more useful.
Finding files often means more than just looking in folders for adequately named files. Sometimes, it's looking for the contents of files, for things that aren't files (because some kinds of files on macOS aren't actually files, they're folders that are treated specially by the system), and for things like contacts, calendar events, reminders, mail, etc.
I’m not a turn spotlight off guy but it is a bit of a pig in terms of apple’s approaches to system crawling and indexing and how it leaves its metadata detritus all over the disk. I can see the desire to disable it for some.
> Never in the past decade have I thought to myself, "gawrsh, I wonder where this file is on my laptop hard drive."
I do, but 80% of the time I'm able to locate it by opening the directory where I would put it. And 10% it's in the "other" directory. And since I have the shell history, in the remaining case it is still a simple search.
I search for stuff all the time. But full disk search just never seems to solve the problems I have. Whatever keyword I’m looking for will inevitably show up in thousands of unrelated header files, Python files and JavaScript files in various node_modules directories and whatnot. Search in finder (or spotlight) is always way too noisy to actually do what I want it to do. Spending hours of cpu time to build that a useless index is deeply disrespectful.
The typical find oneliner to do a fulltext search invokes sed. sed supports regular expressions, so you can do quite a bit more than just a simple text match. And you can also invoke various filter chains on the results.
Because the index generator is broken and constantly using up CPU and memory to index things you'll never look for? I mean, it shouldn't be that way, but unfortunately is.
I personally disable these kinds of search indexes in favor of find and ag/ripgrep etc. They are very fast on a modern system with SSD.
Not available to regular folks I guess, but use prewritten aliases to simplify.
That might be true in theory, but in practice a find oneliner is still the fastest way to find things. It shouldn't be the case, but a fulltext search is faster than using the OS index, because the former is stable and improved for decades by low level developers, while the later is continuously recreated by people who like Javascript in the UI libraries of the OS.
I was hopeful that they'd finally give us something to make Alfred unnecessary but it's still slow as shit, so I'm still using Alfred.
I essentially use it as an app switcher. Sometimes I'm jumping between 6 different apps across multiple monitors and multiple workspaces on each and it's faster do type the first couple letters of the app I want and hit return than to Cmd+Tab, parse the icons in their unpredictable order (made harder by all icons being squircles now), and tab to which one I want.
But native spotlight is too slow and unpredictable.
And I can tell Windows search to not index some directories. Like node_module with a million small files I don't want to search in anyway except with grep.
Perhaps if macs let you configure Spotlight to ignore some directories you could tell it to ignore the entire disk? Which would disable it in practical terms.
Not a fan of Spotlight here too. But, you certainly can instruct Spotlight to ignore some directory or drive through System Preferences > Spotlight and selecting the Privacy tab in it which allows you to add the directory or drive that you want Spotlight to not index.
Challenge with trying to use Raycast more broadly in lieu of Spotlight for systemwide search is Raycast appears to be built on top of the spotlight indexes (mds mdworker)
What happens if the bubble bursts - can we still use all the powerful models to create all this code? Aren't all the agents effectively using venture capital today? Is this sustainable?
If I can run an agent on my machine, with no remote backend required, the problem is solved. But right now, aren't all developers throwing themselves into agentic software development betting that these services will always be available to them at a relatively low cost?
If the bubble bursts we club together to buy one of those big GPU servers (now available at rock bottom prices thanks to the bubble bursting) and run a shared instance of GLM-4.7 (the current best-at-coding Chinese open weight model) on it.
Wow what an interesting animal, haven't heard about it before.
> the chips are stacked against them.
Wikipedia says: "This reproductive method enables the asexual desert grassland whiptail lizard to have a genetic diversity previously thought to have been unique to sexually reproductive species."
Sorry for the late response. I wanted to find time to do more research about this before responding but I'm gunnuh accept that it's not happening on a useful timeframe.
I'm not sure which Wikipedia article that is or if there has been revision between when you checked, I didn't find that in the Wikipedia articles for Telidae or Parthenogenesis. The Parthenogenesis article notes that it's controversial whether this is a threat to their ability to adapt.
So I may have been wrong with that example and I thank you for correcting me. I stand by the statement that evolution can get stuck in local minima but I may have been wrong about the Telidae.
Funny to see the last screenshot of OS X from 2014 in the article. I would love to use a system with such a high contrast and information density. But I also remember very well how many users were upset with the most recent design changes at that time: The all caps section titles in the sidebar, and the gray icons that were previously colored.
Tiger with 10.5/10.6-style 2D grid virtual desktops and Mavericks traffic light buttons, or alternatively Mavericks with aqua scrollbars and 10.5/10.6-style 2D grid virtual desktops is very close to my ideal desktop environment.
It was really nice. With the current linear design I organize desktops by theme (e.g. one for dev, one for research, etc) and with the 10.5/10.6 design I'd use vertical desktops for subcategories — so following the same example, on a single screen setup I might have desktops arranged something like:
│ Rails Docs/Search │ Backend Dev │ Music │
├───────────────────────┼────────────────┼────────────┤
│ UIKit Docs/Search │ iOS Dev │ Chat │
├───────────────────────┼────────────────┼────────────┤
│ MDN/Web Dev Search │ Web Dev │ Email/News │
With this, I quickly develop muscle/spatial memory for where each category "lives" and can navigate there in a flash. It also substantially reduces the need for individual programs like browsers to bear organizational load, so for example suddenly "just" single-tier vertical tabs become sufficient, making browser workspaces and tree style tabs much less necessary.
The added dimension really made things intuitive so you weren't left guessing which vertical had to do with what. One space would give you enough context to know what might be next navigating up/down. It's really a shame how disruptive the change to Mission Control was by removing them.
I tend to organize my spaces by projects and then a dumping ground for "everything else" like general browsing and music.
For projects, unique windows are typically: IDE, Browser(s)
For apps I commonly use across spaces, I assign them to "All Desktops" so they follow me, like iTerm2 and Heynote for keeping notes / task lists even if they cover multiple projects.
Wow. I'm still on 18.7.1, saw the update to 18.7.2 yesterday (100% sure on this), but didn't want to install it at that moment as I needed the phone, and deferred the update to today.
Now I don't see any iOS 18 updates at all, only the iOS 26 prompts. What a dick move, Apple. Especially if this is a) a security update, and b) iOS 26 is known to run poorly on older phones like mine.
A comment on the first pull request provides some context:
> The stream of PRs is coming from requests from the maintainers of the repo. We're experimenting to understand the limits of what the tools can do today and preparing for what they'll be able to do tomorrow. Anything that gets merged is the responsibility of the maintainers, as is the case for any PR submitted by anyone to this open source and welcoming repo. Nothing gets merged without it meeting all the same quality bars and with us signing up for all the same maintenance requirements.
The author of that comment, an employee of Microsoft, goes on to say:
> It is my opinion that anyone not at least thinking about benefiting from such tools will be left behind.
The read here is: Microsoft is so abuzz with excitement/panic about AI taking all software engineering jobs that Microsoft employees are jumping on board with Microsoft's AI push out of a fear of "being left behind". That's not the confidence inspiring the statement they intended it to be, it's the opposite, it underscores that this isn't the .net team "experimenting to understand the limits of what the tools" but rather the .net team trying to keep their jobs.
The "left behind" mantra that I've been hearing for a while now is the strange one to me.
Like, I need to start smashing my face into a keyboard for 10000 hours or else I won't be able to use LLM tools effectively.
If LLM is this tool that is more intuitive than normal programming and adds all this productivity, then surely I can just wait for a bunch of others to wear themselves out smashing the faces on a keyboard for 10000 hours and then skim the cream off of the top, no worse for wear.
On the other hand, if using LLMs is a neverending nightmare of chaos and misery that's 10x harder than programming (but with the benefit that I don't actually have to learn something that might accidentally be useful), then yeah I guess I can see why I would need to get in my hours to use it. But maybe I could just not use it.
"Left behind" really only makes sense to me if my KPIs have been linked with LLM flavor aid style participation.
Ultimately, though, physics doesn't care about social conformity and last I checked the machine is running on physics.
There's a third way things might go: on the way to "superpower for everyone", we go through an extended phase where AI is only a superpower in skilled hands. The job market bifurcates around this. People who make strong use of it get first pick of the good jobs. People not making effective use of AI get whatever's left.
Kinda like how word processing used to be an important career skill people put on their resumes. Assuming AI becomes as that commonplace and accessible, will it happen fast enough that devs who want good jobs can afford to just wait that out?
I'm willing to accept this as a possibility but the case analysis still doesn't make much sense to me.
If LLM usage is easy then I can't be left behind because it's easy. I'll pick it up in a weekend.
If LLM usage is hard AND I can otherwise do the hard things that LLMs are doing then I can't be left behind if I just do the hard things.
Still the only way I can be left behind is if LLM usage is nonsense or the same as just doing it yourself AND the important thing is telling managers that you've been using it for a long time.
Is the superpower bamboozling management with story time?
The obvious case in which you would be "left behind" is the one in which LLM usage is hard, and you cannot otherwise do the hard things that LLMs are doing (or you can do them, but much slower and/or to a lower standard of quality.)
Sure. Although all of the hard things that I need to do I have a history of doing fast and to high standards.
Unless we're talking about hard things that I have up til now not been able to do. But do LLMs help with that in general?
This scenario breaks out of the hypothetical and the assertive and into the realm of the testable.
Provide for me the person who can use LLMs in a way that is hard but they are good at in order to do things which are hard but which they are currently bad at.
> Provide for me the person who can use LLMs in a way that is hard but they are good at in order to do things which are hard but which they are currently bad at.
A PM using LLM to develop software product without DEV?
This sounds like a hypothetical person, unless you have a particular PM who is willing to donate some time and their LLM account to run the experiment for us.
Also Im picking the problem. I have a few in mind but I would want to get the background of the person running the experiment first to ensure that the problem is something that we can expect to be hard for the person.
If you're not using it where it's useful to you, then I still wouldn't say you're getting left behind, but you're making your job harder than it has to be. Anecdotally I've found it useful mostly for writing unit tests and sometimes debugging (can be as effective as a rubber duck).
It's like the 2025 version not not using an IDE.
It's a powerful tool. You still need to know when to and when not to use it.
> It's like the 2025 version not not using an IDE.
That's right on the mark. It will save you a little bit of work on tasks that aren't the bottleneck on your productivity, and disrupt some random tasks that may or may not be important.
It's makes so little difference that plenty of people in 2025 don't use an IDE, and looking at their performance from the outside one just can't tell.
Except that LLMs have less potential to improve your tasks and more potential to be disruptive.
You're right on the money. I've been amongst the most productive developers in every place I've worked at for the past 10 years while not using an IDE. AI is not even close to as revolutionary as it's being sold. Unfortunately, as always, the ones buying this crap are not the ones that actually do the work.
Even for writing tests, you have to proof-read every single line and triple check they didn't write a broken test. It's absolutely exhausting.
I've encountered LLM generated comments that don't even reflect what the code is doing, or, worse, subtly describe the code inaccurately. The most insidious disenchanting code I've ever seen has been exactly of this sort, and it's getting produced by the boatload daily now.
I really don't understand what is going on. I try to, I read the papers, the threads, I think about it. But I can't figure this out.
How can it be that people expect that pumping more energy into closed systems could do anything else than raise entropy? Because that's what it is. You attach GPU farms to your code base and make them pump code into it? You're pumping energy into a closed system. The result cannot be other than greater entropy.
Hum... In theory the closed system includes a database with most of the humanity's written works, and the people that know how the thing works expect it to push some information from the database into the code. (Even though, I will argue that the people that know how the thing works barely use it.)
The reason LLMs fail so often are not related to the fundamental of "garbage in, garbage out".
Yea, "using an IDE" is a very good analogy. IDEs are not silver bullets, although they no doubt help some engineers. There are plenty of developers, on the other hand, who are amazingly productive without using IDEs.
I feel like most people that swear by their AI are also the ones using text editors instead of full IDEs with actually working refactoring, relevant auto complete or never write tests
Tests are one of the areas where it performs least well. I can ask an LLM to summarize the functionality of code and be happy with the answer, but the tests it writes are the most facile unit tests, just the null hypothesis tests and the like. "Here's a test that the constructor works." Cool.
They are the exact same unit tests I never needed help to write, and the exact same unit tests that I can just blindly keep hitting tab to write with Intellij's NON-AI autocomplete.
This is Stephen Toub, who is the lead of many important .NET projects. I don't think he is worried about losing job anytime soon.
I think, we should not read too much into it. He is honestly exploring how much this tool can help him to resolve trivial issues. Maybe he was asked to do so by some of his bosses, but unlikely to fear the tool replacing him in the near future.
They don’t have any problem firing experienced devs for no reason. Including on the .NET team (most of the .NET Android dev team was laid off recently).
I can definitely believe that companies will start (or have already started) using "Enthusiasm about AI" as justification for a hire/promote/reprimand/fire decision. Adherence to the Church Of AI has become this weird purity test throughout the software industry!
I love the fact that they seem to be asking it to do simple things because ”AI can do the simple boring things for us so we can focus on the important problems” and then it floods them with so many meaningless mumbo jumbo that they could have probably done the simple thing in a fraction of the time they take to keep correcting it continuously.
It is called experimentation. That is how people evaluate new technology. By trying to do small things with it first. And if it doesn't work well - retrying later, once bigger issues are fixed.
Hot take: CPython is not an important project for Microsoft, and it is not lead by them. The faster CPython project had questionable acheivement on top of that.
Half of Microsoft (especially server-side) still runs on dotnet. And there are no real contributors outside of microsoft. So it is a vital project.
They also laid off one of the veteran TypeScript developers. TypeScript is definitely an important project for Microsoft, and a lot of code there is written in it.
Anyone not showing open AI enthusiasm at that level will absolutely be fired. Anyone speaking for MS will have to be openly enthusiastic or silent on the topic by now.
> Microsoft employees are jumping on board with Microsoft's AI push out of a fear of "being left behind"
If they weren't experimenting with AI and coding and took a more conservative approach, while other companies like Anthropic was running similar experiments, I'm sure HN would also be critiquing them for not keeping up as a stodgy big corporation.
As long as they are willing to take risks by trying and failing on their own repos, it's fine in my books. Even though I'd never let that stuff touch a professional github repo personally.
exactly ignoring new technologies can be a death sentence for a company even one as large as Microsoft. even if this technology doesn't pay off its still a good idea to at least look into potential uses.
Only in very specific circumstances where there are clear moats to be built (mobile was one of these that Microsoft missed, but that's a PLATFORM in a way no AI product at the moment comes close to). As far as I can tell, there is no evidence of such a thing with the current applications of AI and I am unconvinced that there ever will be. It's just going to ride on top of previous platforms. So you may need some sort of service for customers that are interested, but having the absolute best AI story just isn't something customers are going to care about at the end of the day if it means they would have to say migrate clouds.
At the moment, I'd arguing doing much more than what say Apple is doing would be what is potentially catastrophic. Not doing anything would be minimally risky, and doing just a little bit would be the no risk play. I think Microsoft is making this mistake in a big way and will continue to lose market share over it and burn cash, albeit slowly since they are already giants. The point is, it's a giant that has momentum going in the opposite direction than what they want, and they are incapable of fixing the things causing it to go in that direction because their leadership has become delusional.
i dont think hey are mutually exclusive. jumping on board seems like the smart move if you're worried about losing your career. you also get to confirm your suspicions.
This is important context given that it would be absurd for the managers to have already drawn a definitive conclusion about the models’ capabilities. An explicit understanding that the purpose of the exercise is to get a better idea of the current strengths and weaknesses of the models in a “real world” context makes this actually very reasonable.
So why in public, and why in the most ham-fisted way, and why on important infrastructure, and why in such a terrible integration that it can't even verify that things compile before opening a PR!
In my org, we would have had to bypass precommit hooks to do this!
I'm often annoyed how slow/unreliable Spotlight is, especially in Mail, but what's the alternative here?
reply