HNHacker News
TopNewBestAskShowJobs

kjksf

6,940 karma · joined September 5, 2009

My website is https://blog.kowalczyk.info

Working on https://filerion.com/, a web based file manager (s3, Dropbox and more).

I created https://www.sumatrapdfreader.org

submissionscomments
kjksf··on Why Go and Not Rust? (2019)
Because there are lies, damn lies and benchmarks.

I looked at first benchmark (funkuch-redux) where C# significantly beats Go (8.3 vs 32).

It turns out that the winning C# programs uses SSE2 intrinsics.

The second best C# programs virtually ties with Go implementation, which is not a surprise: those are pretty straightforward, calculation heavy programs. Easy to optimize so both .NET JIT and Go compiler do similarly good job generating code.

So all we've learned that using SSE2 can significantly speed up certain kinds of code but it has nothing to do with C# vs. Go.

Arguably SSE2 intrinsics are actually part of C#. Apparently they were added as a library in .net core 3 (in 2019?).

You will typically not use intrinsics in your code because a) you don't need them b) they're pain to use

And I could write a Go version that uses intrinsics and it would be equally fast.

The same applies to n-body and spectral-norm at which point I've lost interest in those benchmark because all they show is that someone spent extraordinary amount of time optimizing C# code with intrinsics while the same effort wasn't applied to Go programs.

But by constructions you average Go program will be faster that your average .NET JITted program.

JIT has constant overhead and can't spend as much time generating best machine code as offline Go compiler.

Plus .NET is more memory hungry (.NET object have memory overhead not present in Go structs), UTF-16 strings take up ~2x more spaces than Go's byte strings etc.

Doesn't mean every Go program will be faster than equivalent C# program but it's more 80% in favor of Go.

kjksf··on Why Go and Not Rust? (2019)
I wrote more Go programs that I can remember.

I never worked at Google and presumably I don't have Google's problems and yet Go is the best language I found to wrote those programs. Simple, fast and productive.

Could you connect the dots for me: what exactly is that Google needs that I don't and how does that map to design of Go?

Is it compiled because only Google needs statically compiled, easily cross-compilable, easy to distribute executables?

Is it portable to almost every OS / arch available because that's something only Google needs?

Is it statically typed because only Google cares about catching type errors at compilation time?

Is it garbage collected because only Google cares about the productivity and memory safety of programs?

Which Go design features solve problems that only Google has?

kjksf··on Why Go and Not Rust? (2019)
Name a program you can write in Rust that I can't write in Go.

Both Go and Rust and general purpose languages. You can write most (but not all) kinds of software in both of them so a claim that Go is more "limited" than Go is bs.

As to "mastery": I've spent significant amount of time mastering C++ and got pretty good at it.

Go was both significantly easier to master and I'm significantly more productive in Go.

So what exactly is the justification for C++ (and Rust) complexity?

kjksf··on Why Go and Not Rust? (2019)
Community (as used in this context) is a derivative of a language.

The reason inheritance is not popular in Go is that it's not possible.

The reason IOC containers are popular in Java and not Go is that Java makes them relatively easy and Go doesn't.

Go is drastically less complicated language than Java, it has significantly less foot guns.

As a result Go community self-selects to people who value simplicity and as a result you do get a different community.

But community is shaped by the language so it really is about the language.

kjksf··on Why Go and Not Rust? (2019)
That doesn't mean that JVM or .NET JIT is faster than Go compiled code.

They will almost always be slower because they have 3 structural disadvantages.

1. Compilation bytecode => machine code happens at runtime so it's a constant overhead. There's also other overhead related to keeping invocation counts to decide when JIT should kick in.

2. Because compilation happens at runtime, they can't spend too much time on compilation so they are limited at how good the generated code can be. Go doesn't have those restrictions.

3. Modern performance is heavily influenced by memory use. By design JVM and .NET are more memory hungry. Every object in JVM has 8-16 bytes overhead compared to Go struct etc. Value types are relatively recent and not commonly used. In practice it adds up.

The are 2 advantages of advanced JITs:

1. It can make better inlining decisions based on actual execution counts of a function vs. a heuristic in a Go compiler. (Go 1.20 will have preview of profile-guided optimization which provides the same optimization).

2. A generational GC used in most JVM / .NET implementation is faster for certain allocation patterns (high rate of allocations / frees) than Go's allocator.

But your average Go program will be both faster and significantly less memory hungry than your average Java / .NET program.

The memory overhead is especially bad for Java / .NET. Go only compiles the functionality used.

Java / .NET have very heavy runtimes / standard libraries that are packaged into one library and have to be loaded into memory.

What it means in practice is: I can run a Go server comfortably on the cheapest shared server with 512 MB and a Java / .NET server would push it to the limit.

kjksf··on Companies use drip pricing to overcharge consumers
Starlink?
kjksf··on GM is investing almost $1B in new V8 engines
Kerosene? Maybe.

A net-zero kerosene? No, because it's a made up stuff, like Toyota's self-charging ICE cars or "clean" coal, gas or whatever.

BTW: Tesla just released a semi truck that does 500 miles fully loaded and it's a v1 technology. It's already good enough for most trucking and an EV semi 10 years from now will be both cheaper and longer range than that.

kjksf··on GM is investing almost $1B in new V8 engines
Combustion and hybrids will be banned in most markets by 2035-40. Until then, it'll market of declining sales year after year.

Investing in anything in them is a huge mistake. It's like investing in improving black and white tv at a time when color tv have 10% market share and growing 50% a year.

Car companies have no choice but to sell ICE for a while. The demand is still there and they can't convert to EV production overnight.

Yet, the winner of tomorrow will be companies that convert to EV quickly and without going bankrupt. Spending $1B on non-EV technology is not going to help GM do that.

kjksf··on Why to start a startup in a bad economy (2008)
A better comparison is something like Hetzner dedicated server.

Here's what you can get for $89+: https://www.hetzner.com/sb?price_from=89

For 92 euro: 128 GB RAM, 3 TB SSD, 8 core i9-9900K, unlimited bandwidth.

You get 50x value for the 2009 price.

kjksf··on Hard truths I learned when I got laid off from my SWE job
More accurately: your government takes your money and then, maybe, gives it back to you in the form of unemployment benefits and the like, minus significant processing fees.

The alternative is to use the money the government doesn't take and save it for your future "unemployment benefit".

kjksf··on Toyota chief says ‘silent majority’ has doubts about pursuing only EVs
That's the theory. Here's the sad, sad, extremely sad reality:

bZ4X sales in US in 2022:

  05: 199
  06: 33
  07, 08, 09: zero (they recalled every car and stopped production and sales)
  10: 351 (resumed production)
Those numbers are not in thousands. That's 351 cars in November.

https://carsalesbase.com/us-toyota-bz4x/

kjksf··on Toyota chief says ‘silent majority’ has doubts about pursuing only EVs
To put this in context: this is not the first time Toyota advocates anti-EV point of view.

In the beginning of this year the same guy (Akio Toyoda) was spreading all the EV FUD there is: electrification will tank the (Japanese) economy, the electric grid will collapse, no-one needs electric cars etc.

He just might be right about tanking the Japanese economy but not because "EV bad" but because Toyota and Honda and their suppliers are big parts of the economy and they are laggards among laggards on EV adoption front.

And it's not just words but money and lobbying: Toyota joined Trump's lawsuit against EPA (now dropped) and is among the largest anti-EV lobbyists, in the great company of oil giants. Anti-EV lobbying means: lobbying against increased fuel efficiency standards and ICE bans.

Given that China is already 30% EV, Europe ~25% their position doesn't seem rational but purely talking their book and trying to not get killed by mistakes they made in the last 5-10 years.

Until mid-2022 they were selling more hydrogen cars (1) than EV cars (0). Hence their bizarre "we're pro EV but ALSO pro hydrogen" stance. Hydrogen lost and there's no path to hydrogen becoming a thing for passenger cars.

They are also waiting their time working on manual transmission for electric cars. It's peak stupid.

But most importantly they only have 1 EV on sale (with a genius name of bz4x) and it doesn't sells well.

Rumor has it that they just decided that their e-TNGA platform used in bZ4X is not cost-competitive with Tesla and started working on a new EV platform that won't be ready in 4 years.

https://www.reuters.com/business/autos-transportation/exclus...

kjksf··on Paul Graham is leaving Twitter for now
No one wanted to work at SpaceX initially. It took Musk's persistence and ability to sell a vision to hire first employees. And yes, also money. But just money gives you Blue Origin, not SpaceX.

It took SpaceX many years and a few rockets blowing up before they had first successful lunch so what you expect Neuralink to have achieved by now?

They're making progress. Let's revisit the "Neuarlink is a failure" 10 years from now.

kjksf··on S.F. police announce dozens of arrests in crackdown on retail theft
That blog post doesn't show that retail thefts were not a reason to close SF Walmart.

The guy found out that in 2017 Walmart decided to close some stores and concluded "aha! closing SF store has nothing to do with rampant and obvious theft but it was part of that program to close stores".

Except the program ended in 2020 and SF store closing was in 2121. It's right there in text supposedly "proving" it's not because theft.

Plus Walgreen explicitly said it's because theft.

There's willful ignorance involved in parsing this information to conclude that it's not theft.

kjksf··on S.F. police announce dozens of arrests in crackdown on retail theft
San Francisco budget is $14 billion. Of which $672 million is for "solving" homelessness.

Austin City budget is $1.3 billion.

Yes, half of Austin budget is spent by SF on homelessness alone.

Money is not the problem.

SF is already getting gigantic chunk of tech money.

They just demonstrated, year over year, that the only thing they are extraordinarily good at is wasting this money.

What we need is to fire ever single government worker and rebuild the government from scratch. And yes, I know that this is utterly delusional proposal, which is why SF will continue spending like Saudi Prince on crack and SF streets will continue to be lined up with homeless people.

kjksf··on S.F. police announce dozens of arrests in crackdown on retail theft
What if a thief resists your citizen arrest and pulls out a knife or a gun?

When the fight instinct kicks in you get a very dangerous situation for both.

You don't want the security get hurt but you also don't want the security to hurt the thief.

kjksf··on S.F. police announce dozens of arrests in crackdown on retail theft
Having Walmart hire private security is bad for both Walmart, the people they would hire and the society.

For Walmart it's obviously a cost, probably higher that the cost of goods stolen.

For the private security: they are literally risking their lives and they are probably not paid a lot. Shop owners get beaten up and sometimes killed when defending their stores.

For the society: we don't want privatized violence.

What is the security guard supposed to do when some big dude (or worse: a group) starts stealing stuff?

Get in a fight? What if they hurt the thieves? It's an assault and they could end up in jail for that.

And in SF the police doesn't bother to arrest thieves because D.A. was refusing to charge them.

So what would even be the point of e.g. detaining the thieves?

kjksf··on Graph showing level of Covid, flu and RSV in US wastewater systems, by city
Being killed by a falling brick is a very clear cost.

How would you characterize the costs associated with wearing a hard hat most of the time when in public?

Talking about cost without normalizing for probability of prevention is meaningless yet you want to frame the discussion to be purely about (low) cost of wearing masks.

We had a year of a pretty tough lockdown, most people wearing masks, distancing in grocery shops, restaurants and "non-essential" shops and companies closed, no public events, work from home, most people staying at home etc.

And yet people were getting sick with covid, even those wearing masks.

Masks alone won't protect you from covid. Even pretty hard core isolation might not protect you.

We basically have no idea how effective just wearing masks is. It surely reduces the probability but by how much?

Feel free to wear masks and hard hats but don't pretend you know for sure it's a rational choice.

kjksf··on Twitter applies 7-day suspension to half a dozen journalists
And how exactly do you know how much time Musk spends with his family?

This torrent of baseless confabulations about him is truly disgusting.

kjksf··on Twitter applies 7-day suspension to half a dozen journalists
Jesus Christ, are you seriously calling Musk a liar based on who held the child the exact moment the child died?

Like he has to provide forensic level testimony with a timeline of who's holding a child while it's dying?

That's some psychopathic level lack of empathy for someone who's child died.

kjksf··on Twitter applies 7-day suspension to half a dozen journalists
What happened was: Musk was unhappy with progress so he fired bunch of executives and promoted others to take their job.

And voila: the new team successfully launched Starlink.

Remind me, what was the other company where Musk fired a bunch of people? It's been in the news recently.

kjksf··on Twitter applies 7-day suspension to half a dozen journalists
Latest rumor is that they pushed it from 2025 to 2026.

By 2026 it'll be next to impossible to make any serious impact in EV market, certainly not serious enough to affect Tesla.

In 2026 Tesla will be at run rate of 5+ millions cars.

There's no magic in this business.

Even if Apple has a car with that kind of demand, it takes 1 year to build a factory and 3 years to ramp it to 1 million cars a year. This is what Giga Shanghai did and that's faster than anyone ever done it.

So we're talking 2030 for 1 million cars, if somehow Apple can build it's first factory at the same scale and speed as Tesla it's second factory, after lots of painful learning scaling Fremont production.

Plus, without robotaxi what's the point? Luxury brands like BMW / Audi / Mercedes top out at ~2.5 million a year. That's a business, but it's not a Tesla destroying business.

kjksf··on Twitter applies 7-day suspension to half a dozen journalists
Quite a tale, if entirely fictional.

> Capital that is going going to be harder to raise

Tesla has over $10 billions in cash and adding few billions every quarter. They don't need to raise ever again.

> They had early-adopter success

Yes. Also, they became the largest EV company in the world with 2x margin of other car companies.

> when they had the field to themselves

Nissan Leaf and Bolt EV launched before Model 3.

Jaguar i-Pace, Audi eTron, BMW i5, VW ID.3 and ID.4, few models from Hyundai and quite a few more.

Model 3 and Model Y had plenty of competition for several years.

That competition didn't sell many cars and Tesla did.

> But now every major car company ... are coming for them

More like desperately trying to catch up. Tesla is still ahead of everyone in things that matter, like securing raw materials for batteries, building battery cells, securing battery cells from suppliers, manufacturing (gigacasting, spending less time and money to build a car), building more factories (ramping up 2, soon announcing 2 or more), Tesla Semi with best specs by far, still the best motors, the most efficient cars, the safest cars, building insurance business, shipping more software updates than anyone. This is not a complete list.

The question for the future is not: will Toyota or Honda kill Tesla.

It's: will Toyota and Honda keep up enough to not go bankrupt.

kjksf··on Twitter has re-suspended ElonJet account
https://twitter.com/scottwww/status/1490553502640140288

> Elon uses the FAA PIA privacy program for a private plane ID. When using a PIA address, the owner is anon and private, not public. Sweeney’s workaround is (likely) to spot a (rare) ICAO plane resembling Musk’s & noting the private code.

You can no longer claim it's public. It was de-anonymized in some way.

It's no longer a case of reposting public information but someone going considerable lengths to post this information.

The @ElonJet guy even bragged about it:

> Elon got PIA but I've already identified it!

kjksf··on Twitter has re-suspended ElonJet account
Plot twist: Elon applied for and obtained privacy protection from FAA so it was no longer public information.

The owner of @ElonJet account de-anonymized the info and bragged about it on Twitter.

Now Elon is suing him and people who helped him to obtain that no-longer-public information.

See https://twitter.com/scottwww/status/1490553502640140288

kjksf··on Tesla Market Cap falls below $500B
Tesla EPS growth: 94.96%, 660.27% Toyota EPS growth: 15.49%, 6.95%

The market looks to the future. A company with 95% earnings growth deserves much higher P/E than a company with 16% earnings growth.

Tesla and BYD are clearly dominant players in a rapidly growing EV market.

Toyota is a laggard among other laggards. With a single EV only introduced for sale this year they are behind pretty much everyone.

kjksf··on Tesla Market Cap falls below $500B
> in case of Tesla a lot of this expectation of future profits and growth was based on charisma of its owner

Not even a little bit. Musk has a charisma of a bag of potatoes.

What matters is that Tesla had 10 years of 50%+ revenue and unit growth, on average. They performed better than any of the FAANG companies.

With Giga Berlin and Giga Texas ramping up, Cybertruck, refreshing Model 3 for lower cost of production, ramping 4680 battery production, lithium refinery in Texas, Inflation Reduction Act in US => the 50% growth the company is guiding for in the next several years seems to be in the bank.

kjksf··on Tesla Market Cap falls below $500B
Where do you get Toyota's $40 B in profit?

Last fiscal year Toyota made $17 billion profit, which was less than $19 billion profit 6 years ago.

Profit growth rate is the crucial part of the info you skipped.

To simplify: market cap is ${CurrentProfit} * ${P/E}.

P/E is a reflection of profit (earnings) growth rate.

Since becoming profitable, Tesla has been growing profits at 50%+. They are the fastest growing large company in the stock market.

That growth is going to continue due to ramp of Giga Berlin, Giga Texas, introduction of cybertruck, Inflation Reduction Act which will turbocharge EV demand in US.

Toyota is not growing earnings at all.

Things are likely to go even worse in the near future. The EV market share is skyrocketing (25% in Europe, 30% in china) but Toyota has only 1 EV currently that isn't selling well.

It's rumored that they are scraping their old EV plans (make EVs based on current platform that was designed for easy manufacturability of both EVs and ICE cars in the same factory) and considering developing a new platform.

Platform that won't be available until 2026.

EV market share is likely to grow 50% YoY and during that time Toyota won't have EVs to sell. That won't end well.

Tesla is severely undervalued based on their forward P/E ratio, which is lower than for companies with much lower growth rate.

kjksf··on Apple considering dropping requirement for iPhone web browsers to use WebKit
Those rumors are not really providing new information.

The EU law has passed. It's pretty clear that it requires Apple to do the things those reports say Apple is "considering" (allow truly alternative browsers, allow other app stores etc.).

The law is taking effect soon. Apple has little choice but to comply.

It would be news if Apple was considering not complying with EU law. Which would likely be suicidal but it's not like Apple wasn't jerking around South Korea and some european country with fake compliance.

What I would like to know: will Apple give US a big middle finger by only implementing this in EU or globally.

Again, I think excluding US from those changes would lead to blowback but Apple was so arrogant about those things in the past that I wouldn't consider that impossible.

kjksf··on Tesla FSD data is getting worse, according to beta tester self-reports
No, I'm arguing that Waymo, Cruise and others following similar strategy will go bankrupt before delivering a working product and Tesla / Comma.ai won't.

As to crashes: the disengages part of your rebuttal is implied claim that Waymo / Cruise are perfectly safe.

Which they are not.

FSD have been deployed on 160 thousand cars. No fatalities so far. No major crashes.

Cruise has 30 cars in San Francisco and you get this:

> Driverless Cruise robotaxis stop working simultaneously, blocking San Francisco street

> Cruise robotaxis blocked traffic for hours on this San Francisco street

Another Cruise robotaxi stopped in the muni lane.

Waymo car also stopped in the middle of the road.

Neither FSD or Cruise or Waymo had fatalities.

They all had cases of bad driving.

This is not Safe-but-will-go-bankrupt vs. not-safe-but-won't-go-bankrupt.

It's: both approaches are unsafe today but one has a path to becoming safe eventually and the other doesn't, if only because of economic realities of spending $2 billion a year without line of sight for going break even.

https://www.theverge.com/2022/7/1/23191045/cruise-robotaxis-...

https://techcrunch.com/2022/06/30/cruise-robotaxis-blocked-t...

← PreviousPage 14 of 34Next →