I hate it so much. So arbitrary and capricious. I would say this is currently the number one blocker for the web as a serious platform. And they're doing it on purpose.
I guess the policy is that tabs can use 100% of the available resources on low end devices, but only 10% of the available resources on high end devices.
I think the desktop policy might be better. In the tablets I've used, tabs sometimes get killed when I switch tabs and visit another website with a lot of ads. It's an annoying way to lose work in an unsubmitted form. It doesn't seem to happen for desktop.
That's because most (all?) phones don't swap to a pagefile whereas every desktop OS has swap enabled by default. The only practical solution is to buy a phone with more memory. IMO 6GB is the bare minimum in 2024
Ad blockers certainly shave off a few MBs, but in my experience the vast majority of RAM usage is not caused by ads. Unlike first party content, ads are automatically benchmarked by ad exchanges and penalized for using too much resources. I also don't think a 200gram phone is the kind of waste that we should be concerned about. Think bigger
> I also don't think a 200gram phone is the kind of waste that we should be concerned about. Think bigger
A lot of raw ore is processed to get those 200 grams in your hand.
That said, a quick search tells me the carbon footprint of producing a phone is around 55kg, which is about 320km of car travel; it's not trivial, but it's not as much of a bottleneck I thought it might be.
It's not uncommon for me to drive 320km in a day and take flights that Google Flights claims emits 579kg per leg. So I'd place 55kg from phones in the "fuck all" category even if I upgraded every year.
Like, I don't know your life, maybe you go to your family or important business meetings or something with these flights, but in any case none of this is going to be "fuck all" if we want to have any chance to stay under the +1.5°C (or even +2°C) bar.
Video editors are a big one. I've heard of people crashing a browser tab with Figma as well.
For data exploration tools it's very easy to want to use 4GB+ of memory. I found the limit cumbersome while working on financial tools. It usually comes up in internal tools where you reliably have a fast internet connection; it's harder to reach the limit for public-facing tools because there the slowness of sending 4GB+ to the browser is the more limiting factor.
The annoying part isn't just that the limit is there, but that you can't really handle it gracefully as the developer -- when the browser decides you've hit the limit, it may just replace the page with an error message.
For a video editor, only a small portion of the video needs to be in memory at any given time. The rest can be squirreled away in an IndexedDB store, which has no hard size limits on most browsers.
It's one of our big barriers over at Figma. Creative tools in general hit this limit pretty quickly. For context, I was a very heavy user of Photoshop back in the day. Even a decade ago I remember hitting 20GB of active memory use for Photoshop.
Things get really big really quick, especially when you're storing uncompressed versions of raster elements in memory. To frame things in a different way, 4GB is 22 seconds of 1080p video if you're loading the raw frames into memory.
Processing credit/debit- cards isn't that expensive, 30% is a joke, even if you try to justify it with by saying that it helps pay for the store infrastructure.
Spotify pays no feed to Google when they handle the payment processing themself, via some credit card processor and honestly why the hell should they. So Google decide that in an attempt to get even the tiniest amount of payments to go through them, they'd drop the price to 4%... Why is that so important to them?
They aren't charging the fee for payment processing, they're collecting the fee via payment processing. I don't understand why people keep making this mistake. Hell even Epic made it and lost horribly when they made it. The fee is a sales commission, you pay it because you want access to the customer base. They are allowed to deliver you literally nothing at all in additional services and still charge you a fee.
The other stuff they provide exists to sweeten the deal.
I believe the article cited 15% rather than 30% for subscriptions, or 11% if you process your own payments. So [.15, .11] —> [.04, .00], or an overall discount of 11% off subscription revenue they take through the platform.
It sounded to me that, in exchange for that favorable per-subscription rate, Spotify paid $50,000,000 into a “success fund,” whatever that means. $50mm / .11 = $454mm... sounds like they’re basically paying the equivalent of Play Store commission on half a billion’s worth of revenue, without any guarantee that they’ll actually be able to sell that much new growth through Play Store. That feels like sharing the risk in a way that isn’t too out of the domain of reason.
From [0], they added about $2bn in revenue year-on-year in 2022. Although we don’t know the term of this Play Store agreement (or what a “success fund” is), paying an amount equivalent to the normal commission on 1/5 of that total annual growth figure seems like a pretty nontrivial amount to commit.
30% is because most transactions, early on, were around $1. And processing fees are usually $0.25 + ~3% = so 28% of the average transaction.
When you start applying that 30% to $15 spotify subs, $50 uber rides, $60 video game, then a bunch of in-app purchases ranging in value; the math falls apart and if you have scale to do it, the negotiated rate makes more sense for those businesses given the average transaction amount.
The article was obviously horseshit -- and I hate google. But if your search for "running clothes" was getting replaced with "addidas running clothes" you would see clearly in the results. Degraded as it is, I still get results that include the specific words I searched in most situations. That type of replacement would be immediately and obviously apparent.
Let me describe the situation the other person is talking about, with two examples that happened to me in the past year.
1. I scheduled a colonoscopy with a hospital. They charged me $300 for the procedure, which I paid up front. Everything went smoothly and there were no delays, complications, or followups. A week later they sent me a bill for over $1k. They had not applied my $300 payment towards the total. They had not waited for insurance to reimburse them. They simply decided they could arbitrarily charge me more, and they did. Fortunately, my wife is somewhat familiar with medical billing and we did not pay. Insurance reimbursed them, then we called them and told them to apply the $300 we had already paid, at which point we owed nothing. If we had paid the $1k+ bill, I am sure we would not have been reimbursed.
2. I went to a clinic for a minor sports injury. They charged me a $60 consultation fee which I paid up front; I then spoke to the doctor for 15 minutes and got a prescription. A week later, I got a bill demanding an additional $180, and in that bill they claimed they had already been paid by my insurance. I checked with my insurance provider and found they had not even filed the claim yet. I ignored the bill, insurance reimbursed them, and they have not contacted me since.
Your characterization is ignorant of how scummy medical billing actually is.
> Your characterization is ignorant of how scummy medical billing actually is.
Concur completely. I once lived in a state where "Balance billing" was illegal, yet despite having a bluechip insurance plan, seeing some physicians in-network AND paying the co-pay at the same time, I could always count on being balance billed.
GRRRR. Talk about a seriously ticked off customer. Given what I know about medical billing though, I would not be surprised to hear that insurance companies were likewise trying to bilk the practices, forcing these shady responses (My wife's practice has an in-network private practice model, but insurers refuse to accept the billing rates they agreed to as a part of contract with the practice, and they play an intentional game of denials after previous authorizations, or, they intentionally send patients checks which the PTs go and cash, and fail to reimburse the practice)
FWIW for #1 if you had paid the fraudulent $1k bill, after your "insurance" paid you could have likely spent more time on the phone with the provider and gotten them to refund you. Of course if you had paid the fraudulent bill promptly they might not have even bothered billing your "insurance", meaning even more time/rounds on the phone in order to get them to do that.
I'll provide an example...while heavily medicated after breaking both of my legs I signed something on an Ipad, once. I asked for a copy and they weren't able to provide me one. After more than a dozen phone calls I was finally able to get a copy. But that involved figuring out whether it was the company that rented me the wheelchair, the company that billed me, the company that billed my insurance, or the company that owned all of those companies. They provided me three documents I had signed with not the same signature on them. Then when I asked them to pickup the 'rental' that they never told me was a rental, it took five phone calls and over two weeks (during which time they billed me and my insurance for another month of renal). They finally came and picked it up within three hours when I had the news reporter who covered my accident get in touch with the CEO's office.
I reported all this, with call recordings, logs, notes, etc. to my insurance's fraud hotline and they closed the file without doing anything.
Much of health care in the US is wrapped in a blanket of outright fraud. You can't agree to be a victim of fraud. Until I get to charge them for the time I spend correcting their billing mistakes I am totally fine deciding what bill is appropriate and what bill is not on my own.
I can count on one hand the number of times in my entire life that I've agreed to a price for a healthcare service in advance and they actually billed me the agreed upon price and no more.
Dentists tend to work that way. Now, the price they'll charge for any given service is super-arbitrary, as is what they'll claim you need. But they will quote you up-front for what they'll be charging you.
> nothing better than someone who agrees to a service then decides later that they aren’t going to pay because now the deal is “unfair.”
I would agree for any medical transaction where the price for the service was disclosed prior to the transaction. Hell, I’d even agree if an estimate with a low and a high range were provided.
I don’t know about you, but in my lifetime the number of times this has happened rounds to 0%. And I ask for the price at a rate significantly higher than the average medical consumer.
So, since the entire industry has completely and totally failed at price transparency, then I’m completely on the side of someone who makes this decision.
When they agreed to the service 12+ years ago, were they shown the costs? Would you say their purchase was fully informed? Hell for all we know it was an ER visit and they had absolutely no say in the matter.
> Hell for all we know it was an ER visit and they had absolutely no say in the matter
This has been tried in case law, and failed. Even if you're documented unconscious, the EMTs are allowed to make a good faith assumption that you want reasonable efforts of resuscitation. Look up "implied consent" laws.
The EMTs assess your financial situation before deciding if you can afford their actions?
(Charging people for emergency treatment is one of the most insane aspects of the US system, and we need to fight it as long as possible in the UK before it gets inflicted on us)
I understand how it works, I think you’re missing the thrust of my point. Buying a car for an agreed-on price is one thing. Opaque medical billing where we don’t even know what is covered, let alone what it costs, is very different.
You plead poverty with the ambulance company and hope they'll settle on a reduced amount with you. They may also put you on a payment plan, or report non full on time payment to the credit bureaus. Sometimes all 3.
More like you went to your favorite restaurant and wanted the fish but there was no price on it, and when you asked how much it would cost they would ask if you had insurance or not because that would determine your price, so you say no insurance and they say okay well it could cost anywhere between $10 and $1500 and then you get the fish and enjoy it and they give you the bill after and it's actually $10,500.
Don't forget the social shaming where the waiters act like you're ridiculous for possibly caring about how much the fish might cost. No well defined consideration means there is no contract. No contract means there is no legitimate basis to bill for anything more than reimbursement of costs that have been actually incurred.
That reminds me of the time I went to my favorite local restaurant for my annual complimentary dinner and received a bill the following month for $150 then received another bill several weeks later for a further $175.
You standing on this system like it is worth defending is pathetic. You are lower than a worm if you can't see the inhumanity of the system and how defending it is anti-human. If I heard you say something like this in person you wouldn't do it again.
Until our latest insurance, Ive never had a medical interaction in the US that didn't involve my wife fighting some midwit over the phone for an hour over billing mistakes.
She's probably saved us $5-10k, probably $50k for her extended family (about five nuclear families)
I’ve seen this argument repeatedly. The example is just that - an example. It doesn’t cover anything related to productionizing the application, dealing with user management, rate limits, cost controls, hosting the service, data persistence, documentation, billing, marketing, or any of the other 100 things you need to do to turn a simple script into a product that people are willing to pay for, trust with their data, and embed into their own websites.
So sure, be shameless, but don’t be surprised when your copy/pasted version of the example doesn’t take off.
I beg to differ. They worried about marketing (this article is marketing. Their tweets about building it are marketing). And you expect me to believe that they just lumped the example from the docs straight to production? No user management? No website? No auth? Everyone’s data just commingled so the chat bot is responding to user A about user B’s website.
Why does the audience of HN insist on being so dismissive of effort. Yes they moved quickly. No, that doesn’t mean they just copied an example with no other work.
theres no argument. They productionized a docs example, something I did consider too shameless to try. Wish I had tried. Kudos to this dev he is rolling in the cash.
I wrote more about this in another comment ("Postmortem"), as I developed a similar idea from Mid 2022 to early 2023, but made different decisions.
OpenAI had a serious focus on this that did see adoption; as early as Spring 2021 this was in their docs (linked by others), but more crucially, they quickly added an Answers endpoint(https://platform.openai.com/docs/guides/answers) specifically for this use case (query against uploaded files).
The creator of this project did almost everything perfectly, for sure. They had speed, UI, a converting page, found virality. But they also had timing: OpenAI had banned this open-ended use case till that time, they got lucky that people did not develop Explainpaper competitors, and ChatGPT improved quality while reducing run costs by 10x right when they came into this area, meaning many companies that fore-aware of these changes would have taken it's place, were not trying to enter the market (OpenAI approval was hard, and the costs of models before December 2022 were literally 10x as much for decent quality, making it unprofitable)
