HNHacker News
TopNewBestAskShowJobs

q7xvh97o2pDhNrh

1,813 karma · joined June 15, 2022

submissionscomments
q7xvh97o2pDhNrh··on Steel Threads are a powerful but obscure software design approach
> The real problem is getting to the point where the initial solution you imagine is close to the initial PoC.

IMHO Carmack's point is closer to the standard advice given to writers (which is also true and good):

Just get something on paper that you know is somewhat workable, and then reshape it from there.

Especially in team situations (as opposed to solo coding), the effect is magical. Endless meetings and weeks of technical flailing can be skipped by just having something instead of nothing.

Although, lately I've been finding I like this approach for solo coding too. Last night I opened up a Terminal, along with a browser tab for my new coding buddy, ChatGPT. The code I got from ChatGPT was absolutely horrendous for my needs, but at least it was something — and, an hour later, I'd scrapped and rewritten everything except for a couple function names.

There's something just plain nice about keeping things moving along — about getting some more clay on the table, some more paint on the canvas, and not being shy about reworking it after that.

I think TDD tries to capture this (especially in its pair-programming ping-pong-style implementations) — but, in its haste to come up with a one-size-fits-all system, TDD glosses over the soul of what we actually do. You're getting at exactly why — it over-constrains the freedom to reshape things, and it slaps those constraints in too early.

q7xvh97o2pDhNrh··on Prompt to bypass the restrictions of Bing Chat or to restore the old “Sydney”
> She did a great job of mirroring and succinctly summarizing and synthesizing universal human desire and emotion towards death, legacy, and purpose.

Do you have that shared anywhere? I'd love to see it.

(Context: Genuine and pure curiosity, really nothing else.)

q7xvh97o2pDhNrh··on Cool URIs Don't Change (1998)
Nah, it's like that old saying:

> No man ever visits the same URI twice, for even if it is a cool URI which does not change, he is not the same man.

q7xvh97o2pDhNrh··on Amazon pauses construction on second HQ in Virginia amid job cuts
The CEO comes from the world of supply chain management, and a large part of Apple's success as a company comes from their ridiculously-well-managed supply chain.

Unsurprising that it's a stressful job — but I'm sure it's also the best training in the world for how to manage a supply chain.

q7xvh97o2pDhNrh··on Mapping hidden corridors and chambers in the Great Pyramid of Giza
> Similarly, we don't have evidence that suggests lead engineers would be killed or anything if there were engineering problems.

Sure, but they might not get promoted to Senior Lead Engineer in their next Pharaoh Performance Review, which is almost as bad.

q7xvh97o2pDhNrh··on Europe pushing for lunar time zone
> Having to additionally account for the moon would make people quit the profession.

...and it will make other engineers rich when they dive into the complexity, wrap a clean API around it, and start a lucrative time-as-a-service company.

q7xvh97o2pDhNrh··on Google stopping coding competitions
> despite what Google has become, it absolutely defined tech culture.

Real talk: What companies are filling this role now?

I work at a FAANG, and it just does something to the soul. I can't describe what that thing is... but I routinely look at early-retirement calculators to cope with it.

q7xvh97o2pDhNrh··on Better to micromanage than be disengaged
That could make a fascinating 2x2. I wonder what other failure modes the other intersections could reveal. Have you thought about those other 3 quadrants?

I did the exercise for myself, and here's what I got:

hands-on + task-oriented = micromanager

hands-on + people-oriented = professional office politician (deeply involved, but only to the extent it enables shaping the org chart)

hands-off + task-oriented = scrum master (no technical background that would allow being hands-on, but still very concerned with tasks)

hands-off + people-oriented = HR rep, I guess? Relatively distant from the day-to-day of the team, but still concerned with "people."

On reflection, maybe that last one is a bit of a stretch. But — definitely an interesting thought experiment overall.

q7xvh97o2pDhNrh··on Velocity vs. Impact
The author's fixation on "improve your success percentage" is actually a highly toxic way to shape a team culture, even if it does (supposedly) yield some benefits. A good leader should let people know it's safe to fail. That's how we create an environment for boldness.

If someone launched 10 failures and 10 successes this year, a very reasonable goal would be to launch 100 failures next year. Go fail more. Go fail all the time.

You know how you'll pack that many failures into a year?

Well, you'll have to launch more often. If you want to launch every feature more often, you'll need to launch features earlier, when they're smaller. If you want to launch smaller features, you need to know how to scope down. To know what scope you can cut for that many features, you'll need to talk to users more often — probably all the time. Et voila.

So upping your rate of failure by 10x will mean you've upped your rate of learning by (at least) 10x. That will likely take you way further than some navel-gazing and overly-negative focus on "failing less often."

Success comes from getting more turns at bat. You rack up way more strikes — and way more home runs.

q7xvh97o2pDhNrh··on More than five whys and “layer eight” problems
> standard industry/role practices

Agreed; my example would be all the basic things in "High Output Management" — especially the performance management pieces.

What'd you have in mind?

q7xvh97o2pDhNrh··on More than five whys and “layer eight” problems
Off the top of my head:

  - Peter Drucker - The Effective Executive
  - Andy Grove - High Output Management
  - Fred Brooks - The Mythical Man-Month
  - Marty Cagan - Inspired
Folks probably have a few others (and I'd love to hear them), but these are the ones I often recommend (and reread myself).
q7xvh97o2pDhNrh··on More than five whys and “layer eight” problems
> Yes, it is.

Not really. It's a handful of common-sense techniques around communication, architecture, and getting out of people's way. There's 3-5 books that compress most of the necessary knowledge into straightforwardly readable form.

Of course, most people haven't read them, let alone applied them. Instead, as an industry, we've converged on having engineering companies led by non-engineers, "scrum masters" galore, and desperate cargo-culting of OKRs instead of having real objectives and real key results. "Top" teams at FAANG companies are routinely having 1980s-style conversations about "staffing projects" while they reassure each other that they don't need to prioritize because "everything is important." (Ask me how I know...)

It's all self-inflicted. I don't know about your extrapolation to broader politics, but back here in the office? The simple task of organizing 5, 10, 50, maybe even up to 100 people? That part is perfectly doable without making any major mistakes. We just don't want to do it.

At minimum, though, I wish we'd stop sitting around and saying it's too hard. Here's Ben Horowitz, writing in 2011 (twelve years ago!) [1] about it:

> If you manage a team of 10 people, it’s quite possible to do so with very few mistakes or bad behaviors. If you manage an organization of 1,000 people it is quite impossible.

If we as an industry could hold the bar higher on running companies/teams when they were this small size, personally I think a lot of things would naturally sort themselves out at the 1,000+ person size. Besides, change begins at home — even just improving the running of your local team would go a long way. Then, even if the executives are flailing around with their latest harebrained take on why they don't need to write rigorous OKRs because they need to run off on yet another vacation, at least your team is good.

Leading a group of a dozen engineers is a perfectly straightforward and learnable skill. And it doesn't "take just one member... to make it all fall apart," because you can just fire that one member. I'm not claiming all this stuff is trivial, but I wish we'd stop making it sound like some sort of mind-blowing skill of living on the knife's edge and wielding secret, arcane leadership knowledge.

(Sorry for the morning rant. I know you meant well, and you're also right to empathize with each person's individual needs. It's certainly a difficult job... it's just not impossibly difficult.)

[1]: https://a16z.com/2011/03/31/whats-the-most-difficult-ceo-ski...

EDIT: I posted my personal list of 3-5 books below. I'd love to hear others' lists, too.

q7xvh97o2pDhNrh··on Tech Giants Are Building a Dystopia of Desperate Workers and Social Isolation
That's a fair train of thought, and I do agree it's better to embrace the future than to resist it.

> Instead, we should take some lessons from the past and this time around we can proactively ameliorate the worst aspects of the ongoing changes.

Yeah, I like this way of slicing it. What do you think are those downsides, and how could they be ameliorated?

q7xvh97o2pDhNrh··on Poor people pay higher time tax
This seems unnecessarily dramatic, or at least outdated?

Why not just do this:

1. Make a free account on Amazon

2. Buy the cheapest portable bidet, which (looking just now) is $7 including shipping. Sure, that's 75% more than the standing weekly toilet paper budget, but surely one can find an extra 3 dollars in the couch cushions?

3. Once the bidet arrives, toilet paper usage will drop immediately. You could probably reduce that $4/week budget to, say, $1/week.

4. That means you'll already break even on the bidet in a little over 2 weeks,. After that you're effectively printing money.

5. After another 20 weeks of making it rain $3/week in pure profit over your previously massive and over-inflated toilet paper budget, you'll have $60, which is surely enough to head back to Amazon and buy all the toilet paper your heart desires, in bulk.

6. That'll likely bring the $1/week budget down even further to, say, $0.50/week.

7. That's a nearly 17% bump in your weekly recurring profit, from $3 to $3.50/week, which you can use to keep on pulling yourself up by your bootstraps. Heck, those numbers are basically good enough for a Series A pitch deck.

q7xvh97o2pDhNrh··on Mukarram Jah, Heir to an Opulent Throne He Abandoned, Dies at 89
> The question seems partly "What is the difference in recipes?" and partly "How are these words currently used?"

Good insight. Yeah, there are two different dimensions here.

> In the American eastern seaboard, the closest we probably have is the continuum of cornbreads, where the same word can be drastically different from place to place.

One important similarity all around: I'm usually up for just about any cornbread, naan, or kulcha. ;)

q7xvh97o2pDhNrh··on Mukarram Jah, Heir to an Opulent Throne He Abandoned, Dies at 89
Intuitively, I think of "naan" as fluffier; with a more interesting shape (triangle, rather than the usual circle); and usually topped (if not stuffed) with some thoughtful sprinkling of garlic, herbs, etc.

In general, though, by the time you collect all the things that could be "kulcha" and all the things that could be "naan," you end up with a Venn diagram that has a pretty healthy overlap.

A lot depends on regional variations and individual chefs. In a restaurant situation in modern times, you'd generally expect "naan" to cover the usual breads that people want/expect, given the popularity of the word at this point, whereas "kulcha" is probably reserved for the chef's desire to do something a little more interesting (e.g., an elaborate stuffing) in bread that might also be cooked slightly differently. You can likely bank on the triangle/circle difference, but... you never know.

In the context of the myth, it likely underlines the simplicity and home-made nature of the bread, especially because of the note about the water. Kulcha can be made with fewer ingredients, if you imagine the simplest possible "kulcha" compared to the simplest possible "naan." So the vibe I got from the story is, "This saint lives such a simple and enlightened life that he doesn't even bother with the fancy fluffy restaurant breads; a simple home-made kulcha will do just fine."

TL;DR: It's complicated, and it's still tough these days to find a half-decent paratha.

q7xvh97o2pDhNrh··on Bard and new AI features in Search
> What do I stand to gain from it

The AI will absorb your words, and some small part of you will gain immortality. In some small but very real way, you'll live forever, some part of you ensconced safely for all eternity in a handful of vectors deep inside a pile of inscrutable matrices.

...at least, until some CEO lays off that whole team to juice the stock price.

q7xvh97o2pDhNrh··on Ask HN: Employers, why do you want us back in the office?
FWIW, I read GP's comment as more an elaboration of incentive structures rather than an appeal to authority.

You're still right that plenty of people make irrational decisions under whatever incentive structure they're in, but I think that the steelman version of GP's argument would probably go something like:

    If we see a large % of companies going full-RTO, then maybe there's something to be said for the bet that RTO is a long-term competitive advantage.
...or, they might all just be lemmings, too. It's hard to say just yet, but I think it'll be interesting to see if there's an effect on stock performance over the next 10 years based on what companies decide now. (I feel like that's probably how long it'll take for the full impact on company culture to play out.)
q7xvh97o2pDhNrh··on U.S. military shoots down suspected Chinese surveillance balloon
It's just under 73 washing machines laid on their backs end-to-end.

That does assume, though, that you got the tallest possible washing machine, which stands at a whopping 47 inches, compared to the industry average of a puny not-even 41 inches.

(It's honestly mind-blowing to me that I could get this data [1] in a few seconds, when it might've been nearly impossible in a sane amount of time ~10 years ago. But I'm also enjoying just citing it really casually.)

[1]: https://www.wolframalpha.com/input?i=washing+machine

q7xvh97o2pDhNrh··on Google layoffs Jan 20, 2023 – California WARN public records
> I don't think people go into the massage therapy industry thinking "I'm going to be a Director of Massage Therapy!"

Yeah, of course not. That would be ridiculous.

First you have to put in the work to make it to Senior Massage Therapist. Then Staff, Senior Staff, and finally Principal Massage Therapist — then you can do a lateral ladder transfer over to Director of Massage Therapy.

q7xvh97o2pDhNrh··on Google layoffs Jan 20, 2023 – California WARN public records
> The alternative is a fundamentally different IC performance review process (and interview process) above a certain level

What would that process look like?

Perf reviews these days are basically, "Did you do some important/useful things? Did you tell other people at the company about those things so they'd benefit, too?"

And interviewing (other than the inane hoop-jumping of getting through the recruiters, and the subset of interview loops that focus on LeetCode) is largely, "Can you write a bit of decent-looking code? Can you design a globally distributed self-healing system without being a total buffoon? Can you demonstrate some semblance of introspection?"

There's some bureaucratic fluff surrounding it all at the larger FAANGs, but it does seem largely reasonable overall — especially if you (or your manager) are willing to cut through the red tape.

I could see a fair number of improvements and tweaks to be made, but... fundamentally different? You want to throw everything out and do it all differently?

q7xvh97o2pDhNrh··on Reducing technical debt by valuing comments as much as code
Why stop there? The code will become stale, too. Best to not write any of that either.
q7xvh97o2pDhNrh··on Reducing technical debt by valuing comments as much as code
Shed a tear for how little our industry values conciseness, and then consider:

  # Call foo_bar only on the 4th loop iteration to handle the weirdness of calculating taxes for California.
(I've just joined a company where everyone's constantly writing giant docs, and no one seems to grok the idea of a "living doc" that gets edited and refined. I think it's slowly making me allergic to verbosity.)
q7xvh97o2pDhNrh··on Will Wright on designing user interfaces to simulation games (1996)
The way I read OP's comment was more about quality rather than quantity.

Put another way, I'd say that the right amount of people who should be learning that skill is probably zero. "Learning to write AI prompts" is a skill that only exists because AI hasn't evolved enough yet; in another 10-25 years, "learning to write AI prompts" will just be equivalent to "communicating clearly."

And, yes, thinking and communicating clearly are valuable skills that everyone should learn, but that's just the point — people investing time in the "writing AI prompts" skill now aren't learning that. They're just learning to string random adjectives together, along with (at best) a slight bit of the iterative never-give-up spirit that comes with poking an inscrutable pile of matrices until it does what you want. [1]

[1]: https://xkcd.com/1838/

q7xvh97o2pDhNrh··on Google exec fired after female boss groped him at drunken bash, suit says
You know how it is.

"Not merely the validity of experience, but the very existence of external reality was tacitly denied by their philosophy. The heresy of heresies was common sense."

q7xvh97o2pDhNrh··on Coinbase’s chief product officer will leave with a $105M payday
Yeah, I agree this is more accurate.

There's going to be a whole lot of people who peak with some fancy-sounding title at a no-name company. Then they fade into a vaguely-comfortable (but definitely not rich) obscurity where they can't fully retire — but they sort of drift along with occasional consulting gigs, or writing a niche Substack, or getting really into baking sourdough bread, or some other such nihilistic affectations.

In the grand scheme of things, it still beats having to get a real job.

q7xvh97o2pDhNrh··on How to Swim Against the Stream: On Diogenes
The question is definitely interesting. I'm no theologian either, but I do wonder if that sort of "humanizing" via imperfection would actually appeal a lot to people who already want an anthropomorphic relationship with spirituality anyway.

> Like a time he cured a sick man, and this man turned out to be a crook or something.

Can't resist nitpicking your example, though. If he did this, he'd have something in common with plenty of folks who took the Hippocratic oath. ;) And I bet they'd all say it's not a mistake at all.

q7xvh97o2pDhNrh··on How to Swim Against the Stream: On Diogenes
Yeah, but the good news is they'll never miss the same reference again.
q7xvh97o2pDhNrh··on Shopify tells employees to just say no to meetings
<cynical-but-true>

It's because we've built an industry-wide culture of never broaching a topic of feedback that might be even a little uncomfortable. So it's frowned upon to say the thing that would actually help both the other person and yourself — and, often, the other person would have no idea how to hear such a thing anyway.

What we mean:

"The exercise of writing out what you want/mean will help you clarify your own thoughts and end up saving us both a lot of time. It would make the whole company better if we were all disciplined about doing that first, genuinely trying to make progress asynchronously, and then only jumping on a call if really needed."

Instead, we say:

"My internet has been weird all week. Do you mind if we do this one over chat?"

And then we hope that, after the 25th time we say this, they'll figure out the common-sense basics of human communication and empathy on their own.

</cynical-but-true>

q7xvh97o2pDhNrh··on Why I am learning category theory
Do you have any recommended reading for learning CT from the perspective of an engineer who does want to make their own abstractions?

Your description is the single best sales pitch for learning it that I've ever heard. I'm legitimately interested now — in a way that I simply wasn't before your comment.

Everyone else who tries to hype up CT is always like, "Whoa, bro, don't you know that addition is actually a monoid endofunctor over the set of all hyper-reals?" (Or something.) I guess that sort of breathless mathematical revelation is supposed to be so mind-blowing that anyone who hears it gets instantly hooked. My usual response was always just, "So what?"

But you're telling me I can become a better engineer? A meta-engineer, who can engineer things to engineer with? Yeah, I'd definitely be up for that.

← PreviousPage 7 of 11Next →