telendram

Python is part of the official bindings provided in the repository

As in many large projects, even more so with a large company, the point is not to react when the problems are happening, it is to preempt these problems, foresee them and prevent their happening.

So indeed, by 2018, even though Intel has not yet fallen, it's actually already late. The roots of the problems seems to be earlier, and that's where the CEO, and the Board, should have reacted.

Not even. Zstd offers the best decompression times using just one single core.

Multi-threading is a feature for the compression side.

Wow, LZ4 high compression mode has become really fast, at least on my system. I don't need to settle on some lower compression setting now. That's -9 all the way.

The probability is much lower than that.

To begin with, it requires the distance between 2 consecutive matches to be exactly 65536. This is extremely rare. Mountains of files never generate such a situation. Then it needs to employ a repcode match as the match following the 65536 literals. Repcode matches are more common after short literal lengths (1-3 bytes). Needless to say, 65536 is far from this territory, so it's uncommon to say the least. Finally, the block splitter must be active (hence only high compression modes), and it must decide to split the block exactly at the boundary between the literals and the repcode match.

So this is not 0, since Google found a sample, but all these conditions _together_ have an astronomically low chance to happen, as in competitive with winning the Powerball jackpot. I wouldn't worry so much for my own dozens of archives.

Interesting, nVidia selected to go with zstd instead: https://developer.nvidia.com/nvcomp

This guy has no idea what he's talking about. Yet again another click-baity title, followed by an angry rant, the topic of which has nothing to do with C after all (but readers have to go through an impressive amount of noise to deduct that).

Quite the reverse actually, the btrfs estimator is way worse than the zstd one. By using the btrfs estimator, you leave a ton of compression ratio on the table, for no good reason.

Finally, someone dares to talk.

My own experience with this group has been remarkably similar. Oversized ego, dismissive of others, overtly aggressive to competing ideas and technologies, the Rust core team is simply one of the worst I've ever had to make contacts with. The overall feeling when dealing with them is that they are so sure that they represent the "future of programming" that they want the rest of the world to burn now, and no one else than themselves to benefit from the aftermaths.

(I used to be active on Rust in the 2014-2018 era, but then moved away, essentially disgusted by the attitude).

Well, indeed, there are limits to copyrights that are supposed to be applicable when using trivial common names.

But ever since a certain company decided to call itself "fruit", and then go on rampage attacking small players which themselves used to be called "fruit" under the petty reason that they are actually producing and selling "fruit", there is a sense in US that, as long as you got the balls and the lawyers, there is just no limit. Limits of the rules, even when written, are for "others".

Or maybe this is a classical troll scheme, like we have seen a hundred times, black money used to confiscate a public good (a common name) and ransom it to highest bidder, all this with the entire backing of Uncle Sam's own lawyers.

And maybe you are yourself a troll.

I think the pb is that it's so barebone that pretty much nothing works out of the box. One has to add a few more components, depending on which application needs to run (as an OS alone is, well, not an end in itself). However, that's where it stings : adding "components", yet but which ones ? in which order ? how ? Componentization was not a topic, so there's no tool, almost no doc, just tribal knowledge on the topic. That makes it essentially useless for the vast majority of humanity.

Well, maybe you have to remember the capacities of these days. If my memory serves me right, a "good" computer for win98se was 64 MB of RAM. But of course, there were plenty of "less powerful" ones still running, with 32 or even 16 MB of RAM. With such limitation, I can see how a 800x600 BMP image can become a drain on RAM. The swap engine tries to make up for it transparently by using swap space on spin-disk HDD, but then, it becomes sloooow.

Why does it sound so much like a scam ?

Interestingly, the author manages to spend the entire blog entry without even mentioning Apple's own Ad Network. If there is a single example where you pay (the devices, the apps) and are nonetheless tracked (without even receiving a warning selection prompt, in contrast with _all others_ Ad networks), it should be this one.

The part where he is wrong is the premisses that every new computer must more or less gives up all the software of the previous one in order to reach a new trajectory. This used to be more or less true with early 80's computers, such as Apple II or C64, but that's exactly what changed with the PC. 32-bit arrived, and kept compatibility with 16-bit alive. Windows arrived, but (initially) kept compatibility with DOS. etc. Fast forward today, and the Mac M1 arrives, and can still count of Rosetta to continue running x64 applications.

This is a key change, and a necessary one, as rewriting everything for a new system is just a killing blow.

Unfortunately for him, this mistaken analysis is exactly what lead Jobs to the NeXt, a new computer re-invented from scratch, dropping everything that existed prior to rebuild it. This costed too much time and energy, and the NeXt was late, expensive, incomplete, a commercial failure. Jobs was lucky to be bought by Apple before going under.

And arguably, that's a lesson he learned well. Immediately after, while taking command of Apple, he states basically the opposite : re-use everything that can be re-used, especially open source. Only redevelop what you think you can do better.

You obviously haven't lived a single of these "trainings". Once they are in place, it's part of a broader culture. People receiving these mandatory trainings are afraid. During the training, they comply, and look at each other. After the training, they are still afraid. They know it's part of a general strategy to prey on the scape goats, so if any of them "laugh" about the "training", he risks being reported or complained about by another employee, even outside of working place and working hours, because mandatory Reporting of any Deviance under penalty of job termination is also Part of the Training !

But hey, I get it, as long as "others" have to suffer it, you don't care.

That's not the reality. Once they arrive, these kind of trainings become Compulsory for everybody. Those who refused are fired, with cause. Well, everybody "white", of course, thus worth being "reminded" of their "privilege". And also their flaws. I mean, if they are white, they are necessarily racist. If they seem not to be, they are just hiding it. If they defend themselves, it's a proof of their privilege, and by extension their racism. If you are a man, you are also a rapist in hiding. Etc. Just pass a few hours receiving such kind of "training", get it repeated every year just in case, and we'll see how you think about it afterwards.

> I’m buying an iPhone because Apple curates the App Store

Ahah, Ahahah.... AhaHAHAHAHAHAH

Sure, the App Store is just full of Scam Applications that are "so hard to detect" that they top the App Store revenue chart,

https://www.breitbart.com/tech/2021/02/09/developer-uncovers...

but Hey, THANKS APPLE, they CURATED the App Store !

(PS: if you got scammed and get your bank account siphoned off, or lost all your bitcoin, this is obviously Your Fault. Now that you have a problem, let's correct the previous statement : Actually, Apple Never promised anything, right ?)

What @shkkmo said is right : the problem is the total absence of consumer choice.

It's strange that he got downvoted for telling something so obvious.

> in the aggregate Apple does do more for privacy than Facebook

Then you are extremely naive.

Apple takes away from you as much private data as it can. It's just very bad at exploiting it, primarily because the company already earns so much money, it's not even incentivized in recognizing the benefit of data mining.

But with iPhones sells diminishing every year, the situation is changing. Growth must come from services.

Apple will continue to slurp away your private data. Apple will invest more into this resource, and make more and more use of it, including for the targeted Ads that used to be "so evil" when they came from another Ad network, but hey, if it's Apple, it's different.

Ever heard of the concept of "selling your business" ? On what ground is it valued, at sell time, if not based on the promise of future equivalent revenues ? This is hard, concrete, evaluated, value.

The problem is about ATT being only triggered for _all other_ Ad networks, but _not_ for Apple's own Ad network, which is sneakily enabled by default.

Sounds like an obvious monopoly abuse to me. I hope some government (European maybe?) do something about it.

How come no one complains against CC companies about this obvious sell off of our privacy ? How come they are not targeted, neither by NGO nor government agencies ?

It mostly shows that there is no shortage of rumor generators and listeners

Nowhere in the message is there even a single vague hint that the product will "charge" anything in the future, let alone a "threatening" (with a knife) statement. This interpretation is pure bad-faith speculation if not downright misleading libelous statement from article's author, iMore.com, yet another (paid for?) Apple bitch, desperate to make people look anywhere else than Apple's ongoing litigation for monopoly practices and abuses.

Yeah, also don't say that Apple requires all other Ads business to go through a "scare screen", to "protect users", while sneakily making _their own Ad network_ completely free of such shenanigan. Nothing to see, absolutely not another disgusting monopoly abuse.

https://twitter.com/eric_seufert/status/1387044025442594819 https://techcrunch.com/2021/03/09/french-startup-lobby-targe...

> It was astounding how large the disconnect had grown between management and the front line

Isn't this the core of the MBA-hate argument ? That they occupy middle-layer positions, and actively obfuscate messages and merit from front line in order to redirect all the benefits for themselves ?

There is also a darwinian component there, where such behavior is beneficial to game company evaluation, so that only people with similar principles end up being promoted, then actively scuttling any other middle managers that don't adhere to the same set of principles since they are effectively representing a dangerous rival set of values, thus ensuring that after a few cycles, only they remain.

That's how culture evolve in (most) big companies where upper management don't make an active effort to keep middle management in check. A begrudging hard work, that most prefer to skip (get the quick compensation, and move on).

That's also why there is a big difference when companies still have their founders in, as they tend to pay more attention to these issues since they are personally invested in company's future and brand, than recently promoted middle managers which are just temporarily in for the check.

Way faster than Intel i7-8700B, and this is __ while running the x64 code emulated by Rosetta 2__ ... Impressive ...

> the main reason we settled on Brotli was the second order context modeling

This advanced feature is only relevant on reaching compression levels 10 or 11, which are extremely slow. Below that, it's barely used by the encoder, due to memory and cpu taxes.

Given your application has reached speed concerns, and ends up using brotli at compression level 1 in production, you would be surprised to notice that in this speed range, zstd compresses both faster and stronger, by a quite substantial margin.

A pure-rust implementation of zstd decoder already exists in production : https://github.com/KillingSpark/zstd-rs