> Unless previously approved, Anthropic does not allow third party developers to offer claude.ai login or rate limits for their products, including agents built on the Claude Agent SDK. Please use the API key authentication methods described in this document instead.
Which I have interpreted means that you can’t use your Claude code subscription with the agent SDK, only API tokens.
I really wish Anthropic would make it clear (and allow us to use our subscriptions with other tools).
> The fact that Codexbar (also by Steipete) takes 7GB of RAM on macOS shows just how little attention to performance/design he pays to his apps.
It's been running for weeks on my laptop and it's using 210MB of ram currently. Now, the quality _is_ not great and I get prompted at least once a day to enter my keychain access so I'm going to uninstall it (I've just been procrastinating).
Last I checked it spawns claude subprocesses that quickly eat up your RAM and CPU cycles. When I realized the UI redraws are blocking (!) I noped out of it.
I would be genuinely, truly surprised if even 10% were. I think the people on HN who say this are wildly disconnected from the security posture of the average not-HN user.
Clients and customers will not stand for this. I don’t agree but I’ve seen it enough times now it doesn’t surprise me. They want an app, doesn’t matter if you have an identical web-based version that does the exact same thing, they want an app.
I write cross platform apps using Vue/Quasar (previous Angular/Ionic, and before that Titanium), I have put up a web-based version of their app (as a fallback and as an early MVP) and it’s like pulling teeth to get anyone to even play with it. Then you put an app up on TestFlight and suddenly they are using it.
And that’s just trying to get the to use the web while I’m still setting up crap for a “native” app. The idea of not having an app is a non-starter.
Again, I don’t agree with them, I’m just telling you what it’s like out there if you are developing software for other people. An app brings “prestige”, they want be able to say “we have an app”. And no, saving a webpage to the home screen is not a viable alternative (trust me, I’ve tried). Clients and customers reject that and there are extra limitations with that approach (or there were last time I tried, around using the camera feed, things that work fine in mobile Safari).
OK, but you sound like you consult for businesses. Who may not be rational. That doesn't necessarily mean end users are demanding an app. They could be, but your anecdote here doesn't support that conclusion. Patreon's clients are end users.
Sometimes yes but this not apple-to-apples unfortunately. Shipping and support are two places where Amazon _currently_ reigns supreme. Getting an item fast and returning an item easily are often lacking in alternative marketplaces.
I do try to buy stuff directly but so many times I order it and then it comes 10+ days later and often very little info upfront or during about when I will receive the order. I ordered from a website the other day where it said Dec 22nd delivery, it came mid-January and support was unhelpful (parroting the FAQ, might have been a LLM).
Now, I've head great experiences purchasing directly as well. The Smartest House [0] is where I buy everything I can smart-home-related. Their pricing is excellent, their shipping is fast, and their support is top-notch. I had a problem with 1 device, they got on a call and tried to fix it, then when that didn't work, they shipped me a replacement.
So I agree, there are great alternative out there and I also like B&H. I just wish it was easier to be (more) sure about how quick something will ship or how easy returns will actually be. Finding consistently good merchants online sometimes feels like trying to find a needle in a haystack.
When the App Store first launched I think 30% was pretty fair fee for Apple to collect, but that was a long time ago, and before IAP/Subscriptions. Apple might still be entitled to some percentage but they've expanded to cover more and more things (like this Patreon change or Kindle back in the day) and now we have moved far, far beyond the pale.
Apple (perhaps like all corporations but I'm focusing on Apple) is a greedy company that has massively lost it's way. Tim Cook support fascists and/or anything to improve the bottom line, especially if it increases "services" [0]. Alan Dye (thank god he is now busy screwing up Meta) shipped the worst UI revamp I've seen in a while from a company Apple's size and the iOS/iPadOS/visionOS/macOS software is all in dire straits. And they managed to do all of this while alienating developers left and right and playing chicken with governments around the world [0] instead of relaxing their hold on their platforms.
But who cares? The stock price went up. /s
I was overjoyed to see Alan Dye leave (and Jony Ive) and hope that we don't have to wait too much longer to bid Tim Cook adieu. Whoever takes over next has a lot of work ahead to dig out of the hole Tim Cook dug for Apple.
Tim Cook might be the best thing for shareholders but he has been horrible for product quality (software and hardware) and for democracy.
[0] Pay no attention to how much of services revenue came from the Google search deal with the majority of the rest coming from casinos for children and adults alike.
[1] Like the EU DMA, which, I have publicly and privately voiced my dislike of parts of it but Apple has no one to blame but themselves. By keeping a white-knuckle grip on their revenue they forced governments across the world to pass laws (often bad IMHO) that fragment and confuse the entire iOS market.
I suspect developers are looking for these workaround because of the 30%. If Apple had asked for, say, 10%, would there be as many developers looking for loopholes?
I don't know. Apple perhaps should ask for compensation for "vouching for" the developer's app, hosting the app, distributing the app. But Steam shows us another model where the developer themselves pay a modest up-front cost to have their app hosted ($100) and then Steam steps out of the way.
I wonder if this would go a long way too to thinning the herd so to speak from the Apple App Store—perhaps improve the overall quality of the apps submitted.
I think a lot of developers were willing to let it slide when App Store was a luxury market. You could just ignore it and make regular webapps and/or desktop software.
But now iOS is the most popular computing platform in the US. We no longer _have_ an option to ignore it.
And 30% is just crazy. And it's _on_ _top_ of all other expenses: Apple hardware that you need to buy to develop for iOS, $100 per year subscription fee, overhead of using Apple's shitty tools, etc.
I have to respond to your point, though. Whether 30% cut is excessive depends on whether devs feel like they are getting a good deal. As far as I can tell, game developers don't seem to complain about Steam cut very much, it seems like the value you get is worth it.
For example, this thread https://www.reddit.com/r/Steam/comments/10wvgoo/do_you_think... seems like majority is positive about it, even though people debate. When Apple tax is brought up, there's almost never even a discussion there, it's pretty universally hated.
Apple seems to have almost adveserial relationship to its developers. I deploy to App Store and I feel like I'm getting screwed. Even compared to Google, which takes the same cut, but does bahave a lot more nicely to its developers.
May I introduce you to years he let Jony Ive control that. Which brought us things like the butterfly keyboard, thinness at all costs (battery life), and loss of ports (in part due to thinness) that had to be walked back.
Incredible is stretching things. Apple had to catch up with AMD in efficiency, and they did that. Outside the mobile market, Apple is basically a non-entity.
Apple doesn't have huge sales volume for Macs because of macOS and their astronomical pricing schemes, but it's not because of the hardware. Macbooks are easily the best laptops you can buy for most purposes, and they have been since the M1 came out. That has never been true of Apple computers before.
It's because of the hardware. For mobile Apple is competitive, for desktop applications they don't even show up on most benchmarks next to AMD/Nvidia hardware.
That's also because of software. Apple deprecated OpenCL in MacOS eight years ago. In productivity software with solid Metal implementations, like Blender, the M4 Max is on par with the top of Nvidia's (mobile) 5xxx line, except with much more VRAM.
No software fix exists, Apple's GPUs are architecturally limited to raster efficiency (and now, matmul ops). It's frankly bewildering that a raster-optimized SOC struggles to decisively outperform a tensor-optimized CUDA system in 2026.
I get the feeling you had a specific use case that didn't work well with Apple GPUs? I'd be curious what it was. The architecture does have some unusual limitations.
By software problem, though, I meant referencing OpenCL benchmarks. No one in 2026 should be using OpenCL on macOS at all, and the benchmarks aren’t representative of the hardware.
I agree that the early days when every app was a single purchase and the prices were much higher it made more sense. A lot of people got rich from the App Store. So 30% wasn't a huge piece when you were seeing consistent growth every year in the user base.
I think the most annoying thing is how unevenly the policy is applied. Some megacorps pay the 30% and others like Amazon get sweetheart deals. So it unfortunately comes down to who benefits more. If you have something Apple really wants then they will cut a deal. But if not then you pay the high tax. They've at least cut it down somewhat for smaller devs and teams, but the whole industry needs to change. IAP/Subscriptions shouldn't just inherit the pricing systems of old.
I have a feeling Tim is just going to tank the Trump stuff and then peace out next admin so he gets all the blame. Much like Ive and Dye have been.
> I think the most annoying thing is how unevenly the policy is applied. Some megacorps pay the 30% and others like Amazon get sweetheart deals.
I agree, there were deals down to 15% I think (maybe lower) but I don't think that's still happening? I mean, Netflix finally gave up but only after increasing their IAP fee to cover the difference for many years. I might be behind the times on this but I didn't think they still had better cuts for larger corporations. I do know not all developers are treated the same (see Meta still being on the app store after all the shenanigans they pulled with enterprise certs, or Uber), and that does suck. It means that if you are big enough you can break the rules while an indie dev can have everything taken due to an automated system or mistake, even when it's not their fault.
> I have a feeling Tim is just going to tank the Trump stuff and then peace out next admin so he gets all the blame. Much like Ive and Dye have been.
I agree that's likely, though the thought of him staying till the "end" of that is not attractive.
Apple and the contracted company are very very unlikely to tell you they have a secret contract for lower prices in effect unless they are forced to under court disclosure.
Oh, I 100% agree. I was wrong, I thought they got in trouble for doing that but I think I am only remembering things that came out in discovery for the Epic case, which didn’t center on that or prevent Apple from having such arrangements.
There's little assurance of safety or 'fitness for purpose' for apps in the App Store. Apple takes 30% for distribution, and you're basically on your own.
It was the opposite. US mobile operator stores charged upward of 50% to sell stuff on their feature phones, with cherry on top in the form of paid submissions.
You think that's bad? Grugnar charge 80% to sell rocks in front of cave, but Grugnar killed by Bugluk and then cave belong to Bugluk. Bugluk eat you and take rocks if you try sell in front of cave.
I'm replying to the statement that 30% was always a bad deal, by providing an example that shows that it was a clear improvement on the market of mobile development (as others did the same in this comments section).
In your cavemen logic the closest example would be that nobody killed the first guy; he was forced out of business because a new cave opened nearby and they were selling rocks much cheaper.
Cook is a coward and has little to no morals. That has been laid bare for all to see. I will celebrate the day he is no longer the head of Apple. I don't care if he is doing it to shield the company, there has to be a line somewhere and from my perspective we passed it a few hundred miles back.
He will go down as a fascist supporting coward who sold his soul for 30 pieces of silver.
Come on now, you didn't expect someone linking to that trash website to actually read any of it did you? Grokipedia tries to downplay the progressive part but does still mention it.
Maybe I'm just a fuddy-duddy but my eyes about rolled out of my head reading this. The same article could probably be written about multiple companies and it'd be just as uninteresting. It's my understanding that there isn't anything special about WD-40, as in alternatives exist that can work just as well. Now, I think WD-40 is a brand name that can be trusted to work well more often than most alternatives but that is more about process than recipe (I would think).
I've long thought that every restaurant/bakery/etc could publish their full internal cookbooks and not see a drop in sales. People don't buy it because they are incapable (or think they are) of making something, they do it because it's faster, they don't have all the ingredients, they don't have the time, they don't have the skill, the list goes on. I bet I could give the instructions, the equipment, and the ingredients to people and they'd still choose to buy it. Sure, you might lose a tiny bit of sales to "home bakers" [0] but I think it'd be eclipsed by people that saw/read/heard about the cookbook (maybe never even saw it) and that was enough "marketing" to get them in the door.
I've always found "secret knowledge" to be a little silly. A sort of, security through obscurity. Knowing a recipe doesn't make you special, being able to build/run a company around it and make it consistently good does.
[0] I love to cook, I sometimes like making copy-cat recipes. I cannot think of a copy-cat recipe that I made more than 2-3 times. While it's fun to do, it's never exactly the same, and I also believe that "food tastes better when someone else makes it". Also it can sometimes be just-as or more expensive to make some food items due to needing a bunch of ingredients that they don't sell in exactly the quantity the recipe calls for.
> I've long thought that every restaurant/bakery/etc could publish their full internal cookbooks and not see a drop in sales.
Makes me think of all those stories[0] employing a "secret recipe" plot. Some baking/cooking recipe (or a whole cookbook), written down by grandma and passed down in the family, or such, is critical to the fate of a bakery/restaurant/Thanksgiving dinner/etc.; predictably, it gets stolen, and suddenly the meal everyone loves cannot be made anymore.
It's a dumb idea if you think about it for more than a second - even the worst home cook will naturally memorize all the ingredients and steps after using the recipe more than couple times. If the process involves more than one person, there's bound to be copies and derivative documents (e.g. shopping lists) around, too. Recipes are good checklists and are particularly helpful when onboarding new cooks, but losing an actively used one isn't a big deal - it can be recreated on the spot by those who already know it by heart.
--
[0] - One I've watched recently was Hoodwinked! - https://en.wikipedia.org/wiki/Hoodwinked!. Great movie, but out of all the absurdities in it, by far the biggest one was the whole "stealing recipes to put bakeries out of business" plot driver.
Recently Fallow posted video on how they made demi glace. Bit specialized ingredients probably if you ordered far enough from butcher doable. Bigger issue was the larger scale and time effort.
I really feel main difference is the scale and then getting right ingredients and then actually using all of them. Later making thing somewhat cost effective.
I have no doubt any serious company couldn't make something like WD-40. Not exactly same stuff, but in general close enough. Probably close enough that if you labeled over nearly all users would not notice.
Nothing gets gearhead nerds going more than arguing about lubricants and gas. Ask the wrong group of dudes about when to change your oil at breakfast, and they will still be going at dinner.
> Unless previously approved, Anthropic does not allow third party developers to offer claude.ai login or rate limits for their products, including agents built on the Claude Agent SDK. Please use the API key authentication methods described in this document instead.
Which I have interpreted means that you can’t use your Claude code subscription with the agent SDK, only API tokens.
I really wish Anthropic would make it clear (and allow us to use our subscriptions with other tools).
reply