Hacker News Reader

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

For the same reason Trump has opened so many legal challenges in his trials. It's a stalling tactic. Every year Apple can continue their App Store business monopoly without adequate competition is another year Apple earns $100B (or whatever their profit is - Apple does not disclose the figures [1]). They are trying to delay any threat to that as long as possible. No doubt they have done their cost/benefit calculations very carefully, weighing the potential fines and legal fees against the revenue they will be able to keep earning this way, and they have decided this is the best way forward.

[1] https://www.marketwatch.com/story/how-profitable-is-apples-a...

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

$45K per month.

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

Great point, I will revise my statement to "no developer who hopes any app they ever create will one day have more than 1M installs will choose these new terms, and I have no doubt that is exactly Apple's intention here.".

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

Luckily, the EU included substantial anti-circumvention provisions in the DMA. I'm not a lawyer, but based on the excerpts from the regulations below, it seems likely the EU would have grounds to open legal proceedings against Apple's new rules.

https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELE...

"Article 13: Anti-circumvention

4. The gatekeeper shall not engage in any behaviour that undermines effective compliance with the obligations of Articles 5, 6 and 7 regardless of whether that behaviour is of a contractual, commercial or technical nature, or of any other nature, or consists in the use of behavioural techniques or interface design.

7. Where the gatekeeper circumvents or attempts to circumvent any of the obligations in Article 5, 6, or 7 in a manner described in paragraphs 4, 5 and 6 of this Article, the Commission may open proceedings pursuant to Article 20 and adopt an implementing act referred to in Article 8(2) in order to specify the measures that the gatekeeper is to implement."

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

They may not like it, but the question is: is it actually illegal under the new DMA regulations as they are currently written?

The DMA is 66 pages of legalese, otherwise I would have read it to find out:

https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELE...

mil22on Apple announces changes to iOS, Safari, and the App Store in the European Union

"Developers who adopt the new business terms at any time will not be able to switch back to Apple’s existing business terms for their EU apps."

from https://developer.apple.com/support/dma-and-apps-in-the-eu/#...

I have not seen this important aspect mentioned yet. Once you have opted in to Apple's new business terms, you will be locked in to those terms forever.

Because of this restriction, given the ridiculous €0.50 per year per install fees, no developer who hopes their new app will one day have more than 1M installs will choose these new terms, and I have no doubt that is exactly Apple's intention here.

mil22on AI to hit 40% of jobs and worsen inequality, IMF says

Whenever I read stories like this, I'm always reminded of Chapter 7 of Hazlitt's book Economics in One Lesson, "The Curse of Machinery."

mil22on Japan to crack down on Apple and Google app store monopolies

They have deliberately and knowingly made it difficult by showing warnings and making users jump through hoops. I think evidence that it was deliberate and intentional was revealed in one of the many lawsuits in the form of meeting notes and reported speech, if I remember right.

https://www.theverge.com/2023/11/6/23948990/and-were-on-to-s...

mil22on Google Gemini Pro API Available Through AI Studio

60 QPM free is great, but the pay-as-you-go pricing is the same.

Courtesy of GPT4:

"To determine which option is cheaper, Gemini Pro or GPT-3.5 Turbo, we need to consider the average length difference between tokens and characters and the pricing structure for each.

Gemini Pro Pricing: Input: $0.00025 per 1,000 characters Output: $0.0005 per 1,000 characters

GPT-3.5 Turbo Pricing: Input: $0.0010 per 1,000 tokens Output: $0.0020 per 1,000 tokens

Average Length Difference Between Tokens and Characters: A token in GPT-3.5 can be a single word, part of a word, or a special character. On average, a token in GPT-3 models is roughly equivalent to 4 characters (this is a rough estimate as the length of tokens can vary significantly).

Given this information, let's calculate the effective cost per 1,000 characters for both Gemini Pro and GPT-3.5 Turbo. For GPT-3.5 Turbo, since 1 token ≈ 4 characters, the cost per 1,000 characters would be a quarter of the cost per 1,000 tokens. We'll calculate the cost for both input and output for each and compare.

The total cost per 1,000 characters for both Gemini Pro and GPT-3.5 Turbo, considering both input and output, is the same at $0.00075. Therefore, based on the provided pricing and the average token-to-character ratio, they are equally cost-effective."

mil22on Reflecting on 18 Years at Google

https://www.jetbrains.com/lp/devecosystem-2023/development/#...

Flutter: 46% React Native: 35% Unity: 10% Cordova: 10% Ionic: 9% Xamarin: 8%

Just one datapoint, of course - a survey of 26,384 developers.

mil22on Greg Brockman quits OpenAI

It means there was no shifty business.

mil22on Simple fragrance method boosts cognitive capacity: study

Weak science.

They ran a battery of 12 tests and looked for one with a statistically significant difference at the 5% level.

If there were no actual differences, how many tests would you need to run to find at least one that by chance shows significance? If you run 20, you’d expect 5% to do so, on average, which is 1. Running 12 and finding 1 is not a significant finding.

Then, suspiciously, the only statistically significant cognitive test improvement they found was on trial 5 of a verbal learning test, which showed little to no improvement on any of the other trials. That makes no sense.

Finally the brain changes had a p-value of 0.046, which is barely significant by even the weak 5% threshold, and the brain changes weren’t correlated with the test result improvements.

Perhaps interesting, but only if replicated.

mil22on A single weekly bout of exercise may reduce cardiovascular mortality (2006)

This study is 17 years old and from an era before physical activity tracking devices like smartwatches became widespread. As such the findings were based on subjective answers given by participants to questionnaires and in interviews. There are likely more recent and more informative studies published now.

mil22on Aromas while sleeping spark cognitive increase: study

Weak science. They ran a battery of 12 tests and looked for one with a statistically significant difference at the 5% level. If there were no actual differences, how many tests would you need to run to find at least one that by chance shows significance? If you run 20, you’d expect 5% to do so, on average, which is 1. Running 12 and finding 1 is not a significant finding.

Then suspiciously, they found one significant result on trial 5 of the verbal learning test but little to no improvement on the other trials. That makes no sense.

Finally the brain changes had a p-value of 0.046, which is barely significant by even the weak 5% threshold, and the changes weren’t correlated with the test result improvements.

Perhaps interesting but only if replicated.

mil22on Lima: A nice way to run Linux VMs on Mac

You can use VMware Workstation Player and an unlocker like Auto-Unlocker (https://github.com/paolo-projects/auto-unlocker) to enable Mac OS as a guest. It works, but it's very slow because the Mac guest runs without GPU acceleration.

mil22on Google CEO: “can we change the setting of this group to history-off” [pdf]

At that time, all corporate email was already being hosted on an internal version of Gmail, so yes, both were technically possible, and external Gmail accounts were in fact backed up. To do so would have rendered any instructions not to delete any emails moot and unnecessary. Evidently they decided not to do so, and instead, tell everyone to NOT delete any sensitive emails. One can only speculate why that might have been.

mil22on Google CEO: “can we change the setting of this group to history-off” [pdf]

Personal anecdote, personal interpretation (though shared by colleagues I discussed this with at the time who also heard the tone and phrasing of the words first hand). The SVP would never tell anyone to delete emails, and he did not do so! Did I make that clear? He did NOT do so. He definitely DID NOT want us to DELETE ANY EMAILS! No SVP would have publicly instructed anyone to do so.

Unrelatedly, I would not have believed you at the time if you told me that the then CEO had entered into an illegal mutual non-solicit agreement with Apple (and Intel, Adobe, and others), suppressing employee pay by agreeing not to recruit or hire each other's workers. History and a successful class action lawsuit revealed a different story.

mil22on Google CEO: “can we change the setting of this group to history-off” [pdf]

At Google in 2006, the SVP of Engineering was presenting to the entire engineering org during an all hands meeting. A question came up as regards retaining emails related to a pending lawsuit. He told everybody:

"Do NOT delete any emails. <chuckles>. Let me say that again to make sure everyone has understood me correctly: do NOT delete any emails!"

The way it was said, the chuckle, and the way it was repeated with the extra emphasis on NOT, sent a clear message: delete all such emails.

mil22on Low and moderate alcohol consumption not associated with increased mortality

Thank you, that's super interesting information I wasn't aware of. I did get 23andMe back in 2009 and was able to locate the ApoE4 SNP. For anyone with 23andMe or WGS etc. and wants to do the same, look up rs7412 - the C/C combination is "normal", while one T increases your AD risk by 3x and two Ts by 10x.

As it turns out, I do have one T, so I have another really good reason to avoid alcohol entirely - and also to take krill oil, quercetin, resveratrol/pterostilbene, D3, K2, the Bs, and lithium (the recommendations from the study you linked).

mil22on Low and moderate alcohol consumption not associated with increased mortality

The title of this HN post is misleading and the conclusion of this study is actually bad for alcohol sales. They found low and moderate alcohol intake was not protective, confirming other similar studies, and at higher levels it increased the risk of death (no surprise). So the conflict of interest does not seem relevant. It's hard to see how getting paid to consult for a company in the business of selling alcohol is a conflict when you're publishing a paper confirming alcohol consumption isn't good for you.

"In this updated systematic review and meta-analysis, daily low or moderate alcohol intake was not significantly associated with all-cause mortality risk, while increased risk was evident at higher consumption levels, starting at lower levels for women than men."

mil22on Low and moderate alcohol consumption not associated with increased mortality

Yes, this is interesting research. If I was going to drink again, it would be limited to red wine and beer (which contains other beneficial compounds). Personally I just take pterostilbene, which is structurally similar to resveratrol and has similar health benefits but is much more bioavailable.

Cognitive decline and dementia are multifactorial, and with alcohol consumption there are trade-offs between neuronal death, potential improvements in vascular function, and other things known and as yet unknown.

Here's an interesting recent finding from a study of 20K BioBank participants: "Alcohol consumption above 7 units weekly associated with higher brain iron. Iron accumulation represents a potential mechanism for alcohol-related cognitive decline."

https://journals.plos.org/plosmedicine/article?id=10.1371/jo...

mil22on Low and moderate alcohol consumption not associated with increased mortality

I am not aware of any such studies but I would be curious if you have references. There are only a few isolated parts of the brain that show neurogenesis (the birth of new neurons) in adulthood - the hippocampus and the olfactory bulb are the primary two examples. Alcohol kills neurons across the entire brain, with volume reductions and microstructure damage in the white and grey matter, and the damage is not limited to any specific areas.

There is some functional recovery after abstention from alcohol consumption due to regrowth of synapses and neuroplasticity, much in the same way our brains can partially recover from a stroke even though the neurons in the part of the brain that was deprived of oxygen are permanently lost. However the recovery is not complete and some of the loss of brain volume is in fact permanent.

mil22on Low and moderate alcohol consumption not associated with increased mortality

Over the last two years, yes.

mil22on Low and moderate alcohol consumption not associated with increased mortality

You seem to be misunderstanding the points made in my original comment. The statement "They don't grow back" refers to the fact that alcohol can kill neurons, and these neurons do not regenerate. The subsequent statement, "Since quitting in 2021, I have become noticeably sharper," does not conflict with the previous statement, as it implies that cognitive abilities can recover after alcohol cessation through mechanisms other than neurogenesis.

Firstly, it's important to consider neuroplasticity, which is the brain's ability to reorganize itself by forming new connections between neurons. Even though the number of neurons may not increase, the brain can adapt and compensate for the loss by strengthening existing connections or forming new ones, leading to an overall improvement in cognitive function.

Secondly, alcohol impairs cognitive function through various mechanisms other than neuronal death. For instance, alcohol disrupts sleep quality, which in turn impairs cognitive function. When someone stops consuming alcohol, their sleep quality is likely to improve, leading to better cognitive performance.

Lastly, regular repeated brain insult from alcohol consumption leads to chronic cognitive deficits, as the brain is consistently affected by alcohol's neurotoxic effects. When alcohol consumption ceases, the continuous damage stops, allowing the brain to recover to a certain extent, even without replacing the lost neurons.

In conclusion, the statements in my original comment are not conflicting. Improved cognitive abilities after quitting alcohol can be attributed to factors such as neuroplasticity, improved sleep quality, and the cessation of ongoing damage to the brain.

mil22on Low and moderate alcohol consumption not associated with increased mortality

While they found low alcohol consumption did not reduce all-cause mortality relative to abstention, other studies have shown a shift with reduced cardiovascular mortality balanced out by increased cancer mortality.

For me the primary reason to stop drinking was preservation of brain cells. Recent studies have shown a simple dose response relationship between neuronal death and alcohol consumption, with no minimum threshold. Even one drink will kill a few. They don’t grow back.

Since quitting in 2021, I have become noticeably sharper, my code quality has improved, and I feel better. I do not miss it at all.

mil22on Show HN: Alpaca.cpp – Run an Instruction-Tuned Chat-Style LLM on a MacBook

Runs fine on WSL2:

main: seed = 1678990693 llama_model_load: loading model from 'ggml-alpaca-7b-q4.bin' - please wait ... llama_model_load: ggml ctx size = 4529.34 MB llama_model_load: memory_size = 512.00 MB, n_mem = 16384 llama_model_load: loading model part 1/1 from 'ggml-alpaca-7b-q4.bin' llama_model_load: .................................... done llama_model_load: model size = 4017.27 MB / num tensors = 291

system_info: n_threads = 4 / 32 | AVX = 1 | AVX2 = 1 | AVX512 = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | VSX = 0 | main: interactive mode on. sampling parameters: temp = 0.100000, top_k = 40, top_p = 0.950000, repeat_last_n = 64, repeat_penalty = 1.300000

== Running in chat mode. == - Press Ctrl+C to interject at any time. - Press Return to return control to LLaMa. - If you want to submit another line, end your input in '\'.

> What is an alpaca? An Alpaca (Vicugna pacos) is one of two species in the genus Vicuña, which are both domesticated animals from South America and Central Asia that have been bred for their fine wool. >