HNHacker News
TopNewBestAskShowJobs

seiji

8,711 karma · joined February 19, 2007

matt@matt.sh will work if your email doesn't look like spam.

subscribe to my newsletter: newsletter@matt.sh

Nobody visits HN anymore. It's too crowded.

https://github.com/mattsta

https://matt.sh/

https://matt.io/

What do you do?

submissionscomments
seiji··on It’s open-source software that’s eating the world
Are you calling me insane in a roundabout way?

It takes one to know one.

But it's pretty rare.

Pretty rare? That's back to the beginning, because every acquisition means the bought company has explicitly failed to become "real" in a make-more-than-we-spend (and continually grow the "make more") sense.

But better than chapter 11 if that's the alternative.

I'd rather see 20-somethings lose everything and learn tough, introspective, life changing lessons, than see them running around with their "soft landing" $10 million paydays just because they had the right connections.

seiji··on It’s open-source software that’s eating the world
Apple "owns" LLVM which has changed our computing landscape from the bottom up in more ways than I can count. GCC was happy being stagnant until LLVM came along and GCC could finally see how awful and behind the times they had become. Now there's at least some competition again.

(competition-free platforms are never a good thing, no matter how many times you pray to your zero-to-one god)

seiji··on It’s open-source software that’s eating the world
It's difficult to buy stock in a product. It's difficult to make an index fund of products. It's difficult to retire on "product dividends."

Companies exist as an attempt to get all the money in the world, and most fail that goal. Apple is pretty close. Refusal to acknowledge failure as failure is pathological insanity. Getting paid to fail (or to "soft land" and save face on an outright failure) doesn't mean the fail didn't happen.

seiji··on It’s open-source software that’s eating the world
Imagine if programmers unionized and enforced corporate usage-based givebacks to independent software creators through collective bargaining.

When the default position of management is to fight progress (or hoard private profits enabled by using public goods), playing the role of docile puppies and kittens won't enact change.

seiji··on It’s open-source software that’s eating the world
It isn't about product success, it's about company success. Even for one-trick-monkey companies like Android (see how their prior sale to Microsoft went... smashing success)—selling a company is the literal death of the company and the absence of success outside of private financial engineering.
seiji··on It’s open-source software that’s eating the world
If you are a cult-of-pg founder with "money money money startups exist to make me rich rich rich!" mindset, sure, it's a personal success.

But it's a blatant, raw, undeniable failure when a company lets itself get acquired no matter how much money changes hands.

(this is one of the topics that induce harsh West Coast Cognitive Dissonance because it goes against what the billionaire priests of finance tell you to believe about your life)

seiji··on It’s open-source software that’s eating the world
It's up to you to not accept that excuse.

Except, the company endpoint is a low status programmer.

Imagine how quickly the average programmer would get laughed out of the room if they went to their manager and asked for $100,000 to pay an external contractor to implement a feature they need. Oh, also the feature will be open source and other companies will be able to use it after it's written too.

The world doesn't operate on that level yet outside of maybe seven companies in the world.

seiji··on It’s open-source software that’s eating the world
we can't have negotiations with all 100k library authors between themselves

Yeah, it's the "nobody is ever going to pay for bash, vim, top, netcat, ..." problem.

Apple has found a good middle ground though. Pay developers millions of dollars internally (infinitely growing stock) then they release what they work on as open source code. It solves the (reasonable) compensation problem for individuals while letting new software spread purposefully throughout the world.

We're just missing a middle ground where you can write open source software while not being owned by the most valuable company in the world.

seiji··on It’s open-source software that’s eating the world
making any proprietary stuff impossible to sell.

That is the end goal, right? Your margin is my opportunity. The problem in the middle is: open source software is considered free-as-in-never-pay-me free.

I have multibillion dollar companies using some of my open source software. They've never paid me anything. They ask for bug fixes and feature improvements that would take me six months (or more) of full time work. They don't offer to pay (or then even balk at the thought), they just keep saying "you should support us for free because it's open source."

Part of the "we won't/can't pay" is due to programmers being very low status people in organizations. Programmers don't have budgets to buy anything because "lol free open source" and "it's just code, do whatever." If programmers could allocate resources towards getting features/fixes they need from external projects, the software world could be more... synergistic.

seiji··on It’s open-source software that’s eating the world
Open source companies that are successful are acquired

Acquisitions aren't success though. It's just "failing upwards" for the founders. Successful companies are self sustaining.

seiji··on It’s open-source software that’s eating the world
If you cheat the definition, you could say Apple has opened some code, therefore they are an open source company.

But, clearly that's not what's in play here. It's difficult to be an open source startup because it's easy to get a big ego and start doing software wrong. You don't get people to use your software by saying "LOOK! We have VCs. Trust us. Buy our support contracts!"

You get people to use your software by making good features, growing by word of mouth, then you're entrenched and can start milking the support cow. Your word-of-mouth features don't even have to work (coughmongocough), they just have to be easy to brag about so developers get ego boosts by telling others "hey, look, i found this thing and you should think I'm cool because I'm telling you about it."

Another really important point: open source companies aren't proven. The current crop of open source startup VC funded happy towns are all experiments. It's very possible you can't create sustainable open source software companies (support contracts aren't zero margin software products, and nobody buys software anymore because lol "it's free!!!"). There's no natural "supernormal profit" engine if you make a good, clean, understandable open source product. In fact, it's actually the opposite: open source companies get big when the product is garbage (or too complex) because then everybody needs support to CYA against higher management.

(I've been thinking about writing a few articles about all this in January. Stay tuned.)

seiji··on Why it’s so hard to innovate in the e-mail space (2014)
crm114 isn't actively maintained, but it still works great too.
seiji··on Google Patents Needle-Free Blood Draw
Google also has a whole bio lab developing artificial human skin. just completely normal and innocuous coder edm bro stuff. nbd.

https://www.google.com/search?q=google+human+skin

seiji··on We need less powerful languages
Comment of the year right there.

But, great writing/analysis like this about two levels higher than the HN hivemind can comprehend. X = Y (lisp = productive), sure. But X = Y until Z then maybe Q = F is too nuanced for mass understanding around here. If it can't fit into a pmarca simultaneously smug and condescending tweet, nobody will believe you these days.

seiji··on The War ISIS Wants
There are only so many young people willing to sacrifice everything for fanatic dreams of afterlife fulfillment as payment for promises of improving the here-now. Most youth, in other words, aren't that stupid.

Startups seem to be pretty good at convincing a never-ending stream of youth to surrender 500% productivity now in exchange for (usually lies about) riches in the future.

"Yeah, it's linkedfacebox, you know, a super cool startup and everybody there just works through the night because we're all so dedicated to the cause. Gotta keep crushing it so the founders can get a few billion dollars richer otherwise they won't let me cash out my 0.0001% for $3 million."

humans have limited psychological reactions to things. The "believe in a cause" bucket uses the same manipulation tools no matter if you're a startup, a cult, a thousand year old religion, or a radicalized intentional hate group with a persecution complex. You set up an enemy ("the others" / "the adversary", "disruption"), you rally your troops to beat the enemy at all costs (sacrifice yourself/your time/your health for the greater good and you will be rewarded), then perpetuate cycles of wins/losses to reinforce belief and dedication to externalized, and personally abdicated, leadership.

seiji··on The Culling of the Herd
Don't get caught up in working for one of the startup darlings just for the name recognition.

I've avoided working for trendy startups, but I know a dozen people in my same age cohort who did go for the trendy startups. Now they are out buying teslas and giving away $10 million to charity just for fun.

Can't be successful from the sidelines.

seiji··on Deep Neural Decision Forests [pdf]
Time for a fact update! My, my, how time flies.

Deep learning has really only recently become successful

hinton coined the term "deep learning" around 2006/2007 (more around deep belief nets/RBMs, but still, same thing), if that's considered "recently."

constrastive divergence and convolutional neural networks.

CD was also ~10 years ago. CNNs were reading your checks and postal zipcodes in the mid 90s.

successfully train the network.

In the early 90s, RNNs were driving cars on highways using only webcams under basically VFR. No giant sensors, no LIDAR, no GPS, no mesh networks, just camera input.

seiji··on We need less powerful languages
The other side of massive isolated productivity is: nobody except them knew it or could maintain it. Yahoo ended up doing the whole "rewrite lisp in C++" thing because you can't hire "experts in paul graham's custom lisp macros" at scale.

But, it worked, didn't it? Grow, scale, cash out, dump code on non-founders, become startup messiah.

seiji··on Why New York Subway Lines Are Missing Countdown Clocks
count-up timers like "you missed the last train by 30 seconds! haha?" Seems like a very tauntingly russian thing to do.
seiji··on Startup Playbook
Replacing your water jugs with Kool-Aid at the office just seems evil to me.

Founder/VC/moneyhat friendly isn't always employee friendly. Employees exist to be exploited so the founders and VCs become extremely wealthy.

The best way to drive productivity is to convince your employees there's a vague "external enemy" they've gotta beat. Turn it into mental cult gymnastics. Bind their self worth to the success of the company. Grow at all costs.

There are circles of belief though. You should always know the truth yourself (founder), then you decide which lies to tell the board, then which lies to tell your management, then which lies to tell your employees, then which lies to tell the media/public/world. (where "lies" also equals "putting a spin on the truth" or selectively withholding information (hiding faults in the hopes of improving them before failing completely, etc).)

seiji··on Startup Playbook
It's pretty, but it also warms up my laptop with 20% CPU usage. Everything is tradeoffs.
seiji··on 8-bit number to binary string
as scott adams says: I agree with your analysis of your hallucination.

"Professional Engineers" (i.e. certification required) to write code?

Do we let self taught basement dwellers design bridges crossed by millions of people and 100 story buildings?

seiji··on 8-bit number to binary string
there would be massive outcry and we'd put a stop to it.

Massive security branches? Massive fraud? Ineffective nationalized cybercommands?

Just because we can't see the world burning with our eyes doesn't mean lead in the air isn't giving us brain damage.

seiji··on 8-bit number to binary string
My job is basically fixing the code and processes of corporate programmers who copy/paste examples from websites into production systems. There's little understanding of how the code works and almost no reading of related documentation. If I compiles, I ships.

It's like if bridge designers didn't really know mechanics and just copy/pasted minecraft designs into the real world. Sure, it's not the fault of minecraft designers, but we can help kill fewer people by making better examples.

seiji··on Googling stuff can cause us to overestimate our own knowledge
This is also why interviews are a joke when they expect candidates to have full recall without Internet assistance, documentation, syntax checking, or compiler feedback.
seiji··on Why Neurons Have Thousands of Synapses: a Theory of Sequence Memory in Neocortex
Biology component: http://www.nature.com/nrn/journal/v7/n9/images/nrn1971-f2.jp...

ANN component: http://i.stack.imgur.com/KUvpQ.png

seiji··on 8-bit number to binary string
Towards the bottom of the article it says:

the last thing I did was changing the constants to decimal - they look way more scary this way ;>

because, you know, programming should be innately difficult and confusing so we can laugh at others who understand less than we do.

Never underestimate how many times code (any code) posted online will be copy/pasted throughout the world everywhere from school projects to production systems.

seiji··on 8-bit number to binary string
Here's a fancier, slightly less-magic, version. How to compile is left as an exercise for the reader (hint: Haswell or newer, requires BMI2 instruction set):

    void toBinary(uint8_t byte) {
        uint64_t zeroes = *(uint64_t *)"00000000";
        uint64_t resultStr =
            zeroes | __builtin_bswap64(_pdep_u64(byte, 0x0101010101010101ULL));
    
        printf("hello: %.*s\n", 8, (char *)&resultStr);
    }
seiji··on Ask HN: Verizon Wireless can't stop spoofing. Really?
Depends on what you mean by "spoof."

Anybody can generate custom outgoing caller ID numbers. That's not really a "spoof," that's just lying about who you are when you make a call.

Most phone "spoofing" would be if someone cloned your phone number and the phone was registered multiple times on the same network, so when you got a call, one or more of the phones would ring at the same time.

seiji··on YouTube Red, a $9.99 Site-Wide Ad-Free Subscription with Play Music
http://money.cnn.com/2010/08/26/technology/teachbook/ is one case from the *Book part.
← PreviousPage 4 of 34Next →