I used the Dark Sky app since it came out in 2012. I used to consistently get notifications about precipitation with Dark Sky, that were consistently accurate. The Apple weather application seems less motivated to send me notifications with any sort of regularity. I almost never get them now, even with fresh iOS installs. I haven't moved either.
And just anecdotally, Dark Sky was a delight to use. Apple Maps makes it a chore to extract the same utility from their app.
OK, notifications aren't the same as accuracy though. Neither is delight. I'm just talking about the supposed drop in accuracy that I just don't see.
(And Apple notifications are a mess generally. I constantly have notifications for something yesterday only show up today. I'm not sure that has anything to do with Weather specifically, their whole notification priority system is borked.)
I think the point was (and i'm certain my experience is) that with Apple Weather you just don't get the notifications at all (or rarely do) so it's very hard to get a feeling as to how accurate they are.
Yes, this is what I meant. I don't know what the notification thresholds are, so I use the fact that I'm getting a notification at all as a proxy for accuracy.
Got it. Yeah, where I am it's usually pretty obvious that it's probably going to rain in the next hour or two, so I look at the chart to see exactly when. I don't rely on notifications. So for me the accuracy seems the same. But if you're basing it on notifications then I could totally see why you could have a different impression.
I think -- and I might be wrong, since this is from over a decade ago -- that when I first used Dark Sky, I ended up disabling notifications because it would constantly warn me of precipitation, but then when I checked the graph there was none because the model had since updated, and I wound up turning them off. So notification thresholds are probably something hard to get right, and what is appropriate for one geographic area might not be optimal for another.
Back then (2010-2013ish) I was driving a motorcycle primarily so I was hyper aware of the immediate weather and Dark Sky was like magic in that use case.
Yes. tmux is essential. It's great to be able to monitor a session from desktop, or to pick up an existing conversation i'm having on the computer on my phone. In my shell, I have gemini flash wrapper that I use to manage my sessions so I can quickly connect to an existing one, or create a new one with natural language.
> He doesn't seem to be credited on this page, but I believe Pieter Levels (@levelsio) actually popularized this scheme. The author documents a nearly identical scheme.
I've been doing this (tailscale + termius + tmux + ssh) for at least a year and a half. First with Aider in this exact setup, and now with Claude Code and Codex.
I agree. Jobs being around for the birth of the GUI probably played a big role in that. Pushing past text-based terminals to graphical interfaces meant having to spend every moment thinking about purposeful interaction and design.
Maybe we could use Vison Pro to recreate a visit to Xerox Parc in 1979 to inspire the current designers to use UI patterns that gave been working well for fifty years.
Apple has been rudderless on the interaction design front for over a decade now. The windowing mess is evidence of it. We now have the cmd+tab (app switcher), Spaces, Mission Control, (full screen) split screen, Stage Manager, and now tiled window control. None of those interaction metaphors have been expanded upon since their initial launch.
I'm a "mac guy". I understood why Apple initially eschewed windows style alt-tab, given the emphasis on app-centricism. But now, they've created a thousand different ways to switch windows without giving us a proper window switcher. There are apps that bring alt-tab to Mac, but they are all bad because Apple doesn't give developers access to the low-level APIs to create performant and fully featured window management.
Before, Apple had an endless well of great ideas to tap. That's how we got the term "Sherlocked". However, now that they've locked down macOS so much, they've suffocated themselves of new ideas.
I strongly believe we need to return to the drawing board and design a Window Environment from scratch because we've gone so far down the "everything's a mess" route that it's impossible to work our way out again.
Here's an example: panes, tabs, windows, apps, spaces. These things all fight against each other, have their own little silos, get treated differently by every single app, etc, etc.
> Here's an example: panes, tabs, windows, apps, spaces. These things all fight against each other, have their own little silos, get treated differently by every single app, etc, etc.
Disagree. The fact that these are independent is a huge organizational boon for those who know how to apply some combination of them.
Example: I set up spaces (virtual desktops) to have themes: one of general web, one for iOS dev, one for Android dev, etc. within that, it’s useful to be able to further organize with windows and tabs — windows can represent projects for example, and tabs can represent files. These work together to prevent a Cambrian explosion of windows that would be impossible to manage no matter how your environment works.
It’s useful for apps to be distinct from windows, too. This allows things like moving all windows belonging to a particular app between screens, or minimizing/maximizing them all, hiding them all, etc. Panes in most circumstances are an entirely different beast than tabs and windows… it wouldn’t be useful to turn an inspector palette into a tab.
They work together as long as one’s mental model isn’t overly simplistic.
I didn't really explain my objections very well in such a throwaway statement. My issue isn't that those things exist; pretty much the opposite. I think it's good that they exist and they almost certainly always should. My problem is when app A implements tabs differently from app B, app C uses the same shortcut for switching windows that app D uses for switching tabs, etc. It just all seems so disjointed.
What I want is for an OS that treats things like tabs as first-class citizens, not a byproduct that each app implements in its own way.
macOS actually has native tab support built in. It supports everything you’d expect from a tab system, including merging windows into tabs and splitting them back out. All third party devs have to do is opt in and tell the OS which windows are intended to participate. It’s been a feature for a long time now.
Problem is that most third party apps don’t opt in and instead reimplement tabs themselves. It’s mostly native Mac apps made by small boutique developers that use native tabs.
I only very recently changed my System Settings > Desktop & Dock > Windows > Prefer tabs when opening documents to "Always". I'm pretty sure the default is "In Full Screen".
Now, something like TextEdit creates new files in new tabs rather than new windows. It's great! But by default, everything on macOS seems to use windows rather than tabs. I don't even think most people know about the "Prefer tabs" option at all.
I say this only half in jest, as it's what I'm using on my personal desktop. It's far from perfect though, for some reason the taskbar/launchpad just feels awkwardly shaped, some of the keyboard based navigation gets a little wonky sometimes. And I feel that there's a few areas where the split brain between docking regions and windows isn't quite as polished as it needs to be.
On the plus side, it performs better than Gnome, and is slightly more consistent than a lot of Linux desktop options. In the end, I do think that OSes need to expose a bit more details in a somewhat consistent way as cross-platform apps and even electron-like options are boing to become more common place as people jump between and need to support Window, Linux and MacOS targets, with the variances they all have.
Personally, there are a lot of things I like and dislike between all of them.. I spent a large amount of time tweaking Budgie to be almost exactly what I want at one point, and an update borked my setup... I just switched to Pop/Cosmic and dealt with it from there.
Agreed. Lots of conflicting principles. Like, it doesnt make sense to increase the padding across the UI and at the same time, push quarter-screen layouts. I'm almost begging Apple to create something that I don't like, but in a way that's thoughtful and committed to that bit.
I agree. If you (or anyone else who shares your sentiment) hasn't tried it before, you might enjoy "distro hopping" across different Linux distros, using something like Ventoy to make it easy to try out different desktop environments.
There's a veritable zoo to try out. KDE, COSMIC, XFCE, GNOME and its many derivatives (like Pantheon or the previous COSMIC), Unity, etc. Also interesting are the old XMotif based ones like CDE (which has the delightful "chiseled marble" look of 90s-era DEs).
Windows also has its own share of alternative shells, but AFAIK only Cairo DE is actively supported and developed.
I feel like there are a million reasons why this is the wrong choice. There are probably 10 xkcd comics alone that explain why this is a bad idea.
"Let's throw every away and start from scratch" is a tempting idea, but it rarely works. Even taking your example of panes, windows, tabs, apps, and spaces, each of those have a separate and identifiable use case that, IMO, is valid. At least in my mind, I have a mental model around where panes, windows, tabs, and apps are appropriate, and I personally rarely use spaces (though I certainly understand people who like them), and they've never bother me because I can safely ignore them.
And when you look at the issues identified in the article, they all seem very fixable to me. Fixable starting with Apple getting new design leadership, and given the guy responsible for Liquid Glass jumped to Meta, sounds like it was a good thing for Apple.
I misworded that. See my other comment; tldr, I want windows, tabs, etc. to be properly implemented by the OS, to a standard, rather than in many different conflicting ways.
Cmd+~ switches windows of a given app in case you didn't know (not disagreeing with you but it is one shortcut I find super useful and it helps switch windows).
This is why I regard ChromeOS with fear. Because it really does feel like everything is just converging on the browser as the OS, and a browser is not a goddamn OS.
I have a non-ANSI keyboard so tilde is in a super weird place for that (next to left shift). I swapped the shortcut to Option+Tab, makes much more intuitive sense.
Alt tab is okay but it has performance issues (sometimes window switching is instant, sometimes there's a slight but perceptible delay). The devs have acknowledged it and said that it cannot be fixed.
Other pet peeves include the smallest window thumbnails are enormous and enabling mouse hover to switch windows causes me to switch to the wrong window, and thumbnails are often stale.
I think MacOS would be much more unusable without alt-tab, like this app is a critical piece of software that has to be installed on any MacOS machine I'm using. But it does sometimes miss a window, like it doesn't show up in the window list in alt-tab even though it is open. And then I have to use the three finger swipe up thing to find it. Not a huge deal but occasionally annoying, and I assume it is because they are limited in the access they have to this information, which Apple could make available if they wanted, but have decided not to.
I am nostalgic about their operational principles, but I was already willing to give these up for the convenience of an LCD panel circa 2003. We had a lot of LAN parties to attend back in those days.
I think it was a time before the vintage feels for CRTs really. Sure, there were greybeards naysaying the coming of the LCD due to motion artefacts and smearing but the rest of users just wanted something high-res and flicker-free.
It’s less the feels (though the glow from a CRT in a dark room is entirely different from an LED) and more that the games especially were designed for the bleed and flow that CRTs have.
It’s similar to how subpixel antialiasing really depends on the screen design and what order the colors are in.
The pixelated 8bit aesthetic is more reminiscent of early emulators in LCD than how it actually was “on hardware”.
Yeah—this was my rationale. My first game was Chrono Trigger, which the graphics are stunning for, so I was happy with chunky pixels. I started playing Earthbound though, and the graphics are much more underwhelming. The CRT shader has made it look a lot better though.
When LCDs arrived it was not only greybeards not liking them. I certainly wasn't old enough for that title - but for a while, CRT screens were just so much better! Those early panels (at least the ones I saw) were horrible, sooo slow, totally muted colors and minimal viewing angles.
Agreed. Gemini 3 is still pretty bad at agentic coding.
Just yesterday, in Antigravity, while applying changes, it deleted 500 lines of code and replaced it with a `<rest of code goes here>`. Unacceptable behavior in 2025, lol.
It has become very quickly unfashionable for people to say they like the Codex CLI. I still enjoy working with it and my only complaint is that its speed makes it unideal for pair coding.
On top of that, the Codex CLI team is responsive on github and it's clear that user complaints make their way to the team responsible for fine tuning these models.
I run bake offs on between all three models and GPT 5.2 generally has a higher success rate of implementing features, followed closely by Opus 4.5 and then Gemini 3, which has troubles with agentic coding. I'm interested to see how 5.2-codex behaves. I haven't been a fan of the codex models in general.
When Claude screws up a task I use Codex and vice versa. It helps a lot when I'm working on libraries that I've never touched before, especially iOS related.
(Also, I can't imagine who is blessed with so much spare tome that they would look down on an assistant that does decent work)
> When Claude screws up a task I use Codex and vice versa
Yeah, it feels really strange sometimes. Bumping up against something that Codex seemingly can't work out, and you give it to Claude and suddenly it's easy. And you continue with Claude and eventually it gets stuck on something, and you try Codex which gets it immediately. My guess would be that the training data differs just enough for it to have an impact.
I think Claude is more practically minded. I find that OAI models in general default to the most technically correct, expensive (in terms of LoC implementation cost, possible future maintenance burden, etc) solution. Whereas Claude will take a look at the codebase and say "Looks like a webshit React app, why don't you just do XYZ which gets you 90% of the way there in 3 lines".
But if you want that last 10%, codex is vital.
Edit: Literally after I typed this just had this happen. Codex 5.2 reports a P1 bug in a PR. I look closely, I'm not actually sure it's a "bug". I take it to Claude. Claude agrees it's more of a product behavioral opinion on whether or not to persist garbage data, and offer it's own product opinion that I probably want to keep it the way it is. Codex 5.2 meanwhile stubbornly accepts the view it's a product decision but won't seem to offer it's own opinion!
Correct, this has been true for all GPT-5 series. They produce much more "enterprise" code by default, sticking to "best practices", so people who need such code will much prefer them. Claude models tend to adapt more to the existing level of the codebase, defaulting to more lightweight solutions. Gemini 3 hasn't been out long enough yet to gauge, but so far seems somewhere in between.
>> My guess would be that the training data differs just enough for it to have an impact.
It's because performance degrades over longer conversations, which decreases the chance that the same conversation will result in a solution, and increases the chance that a new one will. I suspect you would get the same result even if you didn't switch to a different model.
So not really, certainly models degrade by some degree on context retrieval. However, in Cursor you can just change the model used for the exchange, it still has the same long context. You'll see the different model strengths and weaknesses contrasted.
They just have different strengths and weaknesses.
if claude is stuck on a thing but we’ve made progress (even if that progress is process of elimination) and it’s 120k tokens deep, often when i have claude distill our learnings into a file.. and /clear to start again with said file, i’ll get quicker success
which is analogous to taking your problem to another model and ideally feeding it some sorta lesson
i guess this is a specific example but one i play out a lot. starting fresh with the same problem is unusual for me. usually has a lesson im feeding it from the start
I care very little about fashion, whether in clothes or in computers. I've always liked Anthropic products a bit more but Codex is excellent, if that's your jam more power to you.
- Planning mode. Codex is extremely frustrating. You have to constantly tell it not to edit when you talk to it, and even then it will sometimes just start working.
- Better terminal rendering (Codex seems to go for a "clean" look at the cost of clearly distinguished output)
the faddish nature of these tools fits the narrative of the METR findings that the tools slow you down while making you feel faster.
since nobody (other than that paper) has been trying to measure output, everything is based on feelings and fashion, like you say.
I'm still raw dogging my code. I'll start using these tools when someone can measure the increase in output. Leadership at work is beginning to claim they can, so maybe the writing is on the wall for me. They haven't shown their methodology for what they are measuring, just telling everyone they "can tell"
But until then, I can spot too many psychological biases inherent in their use to trust my own judgement, especially when the only real study done so far on this subject shows that our intuition lies about this.
And in the meantime, I've already lost time investigating reasonable looking open source projects that turned out to be 1) vibe coded and 2) fully non functional even in the most trivial use. I'm so sick of it. I need a new career
what.cd was the world's greatest music discovery mechanism. You could always ask for recommendations in the forums or in the comment thread of the albums pages. The community always delivered. I miss that type of camaraderie. I also spent more on music as a member of that community than since it has been disbanded.
What.cd was the Library of Alexandria for recorded music, the depth of what was collated and properly labelled there was far beyond anything that has ever existed on any other service, paid for or not. Every permutation of every release, endless live recordings, often multiple of the same event, absolutely incredible.
Private trackers as I understand it, are still a thing in the mid 2020s. Did a replacement that matches (or surpasses) What.cd not pop up in the meantime?
I'm just wondering how a strong community like that was struck a deathblow. It's not like all of its content disappeared.
Orpheus and redacted (previously passtheheadphones) both appeared shortly after what.cd’s demise. I believe they both now have more total torrents than what.cd, however the depth is still not what what’s was 9 years on (I know this because some of my uploads from what are still missing, partially because I no longer have the source material). And, the “cultivation” (ensuring no duplicates, recommendations for releases, general community, etc) is nowhere near what’s.
I would say all other media (or at least, the media I care about - film, tv, books) has what.cd equivalents, sometimes multiple. I think Spotify and AM killed 95%+ of “true” private tracker interest for music, especially with lossless and surround releases being available. The diehard core are still there (names from 15 years ago are still active) but it’s really not the same.
Orpheus and Redacted existed but it's kind of hard to beat the convenience of streaming for the low price in 2025.
Granted you can set up automated *arr systems with PLEXAMP to get a pretty seamless "personal Spotify" setup IME getting true usefulness out of trackers of What's quality always required spending real money - to obtain rare records/CDs on marketplaces - or at least large amounts of time if you went the "rent CDs from the library" route. I personally haven't ran into much RYM releases lacking on Apple Music and what is lacking I can find on Bandcamp or YouTube.
OiNK before that, too. Once waffles and what disappeared then I was never 'able' to get on to one of the newer ones… the whole process is some real archaic thing. Used to have a great 'profile' on those others, but yeah.
While I agree that `--dangerously-skip-permissions` is (obviously) dangerous, it shouldn't be considered completely inaccessible to users. A few safeguards can sand off most of the rough edges.
What I've done is write a PreToolUse hook to block all `rm -rf` commands. I've also seen others use shell functions to intercept `rm` commands and have it either return a warning or remap it to `trash`, which allows you to recover the files.
> There’s an odd tendency in modern software development; we’ve collectively decided that naming things after random nouns, mythological creatures, or random favorite fictional characters is somehow acceptable professional practice. This would be career suicide in virtually any other technical field.
I worked in finance – we gave our models names that endeared us to them. My favorite is a coworker naming his model "beefhammer".
And just anecdotally, Dark Sky was a delight to use. Apple Maps makes it a chore to extract the same utility from their app.
reply