srg0

It appears there is space for "packaging and distribution in EU as a service" business model.

Creators, producers and manufacturers in many fields do not distribute or sell directly to consumers (musicians can't directly upload to Spotify and Apple Music; film producers don't upload to Netflix, writers don't print their books).

A small-time maker may pay a fee to the established distributor of the physical products, who takes care of compliance, and uses their packaging and distribution channels.

Plastic is made from the same stuff as gasoline.

Are there good ESP32-based starter kits with manual books which a kid can learn from? I was looking for an Arduino-like kit as a Christmas gift, and it seems that Arduino kits are unbeatable. The starter kit is available in 10 languages and comes with a project booklet. All ESP32-based products seem to be better suited for more advanced users and seem to have a steeper learning curve.

At some point this year I was getting 14-20 spam calls per hour. I'm in Italy.

Pro Spotify: existing playlists and history, better artists info, better UI.

YouTube Music is both better and worse: UI has some usability issues and unfortunately it shares likes and playlists with the normal YouTube account, as a library it has lots of crap uploaded by YouTube users, often wrong metadata, but thanks to that it also has some niche artists and recordings which are not available on other platforms.

So, basically, another smart speaker skinned as a toy robot.

That's probably driven by some kind of an AR headset. AR can't properly render solids, so it is stuck with having everything transparent. Now it won't look worse than everything else.

Exactly. Some copyright holders do license their work for AI training. It certainly happens in the music industry, but I don't see why texts would be any different. The exception would harm their business.

> most LLM users will ~always choose the smartest model

Most LLM users will choose the cheapest model which is good enough.

I think that LLMs' performance is already "good enough" for a lot of applications. We're in the diminishing returns part of the curve.

There are two other concerns:

1. being able to run the model on trusted infrastructure locally (so some jerk won't turn it off on a whim, and the data will remain safe and comply with the local data protection laws and policies)

2. having good tools to create AI applications (like how easy it is to fine-tune it to customer needs)

> how much the addition of copyrighted material affects how smart the resulting model is

Copyrighted material improve the models, not by making it smart, but more factually correct, because it will be trained on reputable, reliable and up-to-date sources.

Copyrighted material includes works by authors from outside the US. By Berne convention, the exceptions which any country may introduce must not "conflict with a normal exploitation of the work" and "unreasonably prejudice the legitimate interests of the author". So if at least one French author does license their work for AI training, then any exception of this kind will harm their legitimate interests and rob them of potential income from normal exploitation of the work.

If the US can harm authors from other countries, then other countries may be willing to reciprocate to American copyright holders, and introduce exceptions which allow free use of the US copyrighted material for some specific purposes they deem important.

IANAL, but it is a slippery slope, and it may hurt everyone. Who has more to lose?

And I hope that Mistral.AI takes note.

My first reaction was like -- wow, a site that runs on a reverb pedal.

I would also like to point out that English spelling is obsolete and should be abolished (/s). The text of the CRLF abolition proposal itself contains more digraphs, trigraphs, diphthongs, and silent letters than line-ending sequences. The last letter of the word "obsolete" is not necessary. "Should" can be written as only three letters in Shavian "𐑖𐑫𐑛".

According to ChatGPT, the original proposal had:

Number of sentences: 60 Number of diphthongs: 128 (pairs of vowels in the same syllable like "ai", "ea", etc.) Number of digraphs: 225 (pairs of letters representing a single sound, like "th", "ch", etc.) Number of trigraphs: 1 (three-letter combinations representing a single sound, like "sch") Number of silent letters: 15 (common silent letter patterns like "kn", "mb", etc.)

For all intents and purposes, CRLF is just another digraph.

The latter. Plus 2 EUR/month for ad-free video. On top of 50 to 90 EUR/year for Amazon Prime.

If we consider it as a sneaky way to increase prices, it's less steep than YouTube Premium price hike. At least you can ignore Prime Video an get all the other benefits for the same price as before.

From a user perspective, isn't it an improvement of privacy of iOS users?

The bug report referenced in this issue is the case of invasive tracking (basically enterprise spyware) breaking in iOS18 https://bugs.webkit.org/show_bug.cgi?id=279153

Normal web usage is not affected.

```

- WebKit is honoring the cookie's SameSite=None attribute when the cookie is set by server in this case the IdP

- Attempts to set this attribute from the client side (from the app interacting with the iOS cookie store) have been unsuccessful. For example, by setting `.sameSitePolicy = "none"`

- Safari Web Inspector shows the option to set the cookie's SameSite attribute to None, it however, does not get honored either, and is immediately reverted.

```

Thank you!

Slightly offtopic. What would you suggest as an introductory text on modern CPU architectures?

If learning durations were log-normally distributed, how would people be able to graduate from universities and finish studies, mostly in time? Or accomplish anything substantial in their limited time span?

I agree that distribution of most human tasks' duration is skewed (not necessarily distributed log-normally), but these tasks can still have a reasonable upper bound for completion. The success is not binary. Like in grading, we need to accept that some projects will get an A, and some will get only B or C, and it's still OK. Some may fail.

Visit reddit, and you'll see that many of these Europeans who are against immigration are specifically against male immigrants with Black African and North African heritage. I think that there's a lot of racism in these sentiments. More than just a generic xenophobia.

With Logic Pro for iPad they now have applications for all their traditional Mac use cases on iPad. If anything, it feels like Apple is pushing for a switch from low-tier Macs to iPad Pro.

And they surely can sell more gadgets and accessories for an iPad than for a laptop.

I beg to differ. Quick search shows that people animate checkboxes too:

https://stackoverflow.com/questions/44876144/how-to-style-th...

Devil's advocate.

Private offices facilitate sexual harassment and some other undesirable behaviors.

It's a good news bad news situation.

AI assistant is getting a dedicated app. Great. AI assistant can "supercharge your creativity" but still can't answer phone calls. Well, it's going to be as useful as Cortana and Siri.

Citroen Ami. Not 1200, but probably the closest to an e-bike with a roof.

At some point platforms should be held responsible for the ads they run and the damages they cause. If the ad is to be broadcasted to thousands of people, or worse targets groups of people, it's reasonable that it must be approved and vetted by the platform. The cost of such vetting, even if done manually by a human, is likely a fraction of what is the cost of creating an ad anyway. If necessary Google can charge advertisers with an upfront cost.

Edit: grammar

That's a valid observation, but if getting a 12-pack of identical decks is not a problem, it's enough to build 13 custom decks (all Aces, all 2s, etc).

Edit: if the players can announce only what suit they are buying and selling but do not show cards to the others, it may end up difficult to deduce what cards are missing, because only the buyer and the seller know the value of the traded card.

Shuffling piles is difficult to do in a way that nobody, including the dealer, knows which is which. And it requires more table space. It also requires an additional shuffle of the final deck. If the final shuffle is not good enough you can guess which one was the short suit by looking at the sequence of the dealt cards (the deck was not cut, and the last 2 cards are the same suit? chances are the dealer is bad at shuffling and it's the short suit). Maybe passing a tray with four piles and letting each player to swap two piles without letting other players see which piles were swapped is good enough. But it's still fiddly.

Normal shuffle and discarding face down is easier to explain and faster to do. And it can be done where the table is not available or is not big enough. Starting with a 44-card deck and discarding 5 down to 39 should make it more similar to the original 40-card version, but 39 is not divisible neither by 5 nor by 4, so one player will always be one card short.

Being one card short might not actually be a problem in practice if the game is played across several rounds. It may even add some interest to the game by creating an asymmetry.

That's a very interesting game mechanics.

It wonder if it can be adapted for playing with paper cards. One variation can be this one: shuffle a 40-card deck, discard 5 face down, it guarantees that there will be an unknown short suit (with 8 to 5 cards), the goal suit will have from 7 to 10 cards. Deal the remaining 35 cards, and start trading. With 4 players one player will be one card short, so they need a compensation like being able to reduce ante by some amount.

The balance of the game with 35 cards and random suit length will probably be different but the trading mechanics should be able to address that. So ante and payout per goal card may have to be changed.

> MtG saw a huge increase in demand while people were mandated to stay at home

Did it? If anything, MTG Modern prices were going down until the end of 2020, both in paper and online. And it's one most popular formats.

https://www.mtggoldfish.com/index/modern#paper

Even without a cut, there are several virtual distribution services like DistroKid and TuneCore which for a flat fee will take care of submitting whatever WAV one created to all streaming platforms. Getting it noticed is a whole other story.

Awesome slides, but then I tried to subscribe to growth.design newsletter using a Firefox Relay email, but mozmail email masks are rejected. So much for privacy.