HNHacker News
TopNewBestAskShowJobs

dgreensp

3,476 karma · joined April 10, 2007

David Greenspan
submissionscomments
dgreensp··on Show HN: Open-source AI video editor
Thanks for posting it!

Remotion does look really cool for in-browser video editing. People should note that you can’t make free-to-use software with Remotion. It is a paid product, and you can download and run the code for free in very limited circumstances that let you experiment with it. You can’t use the output of your own software built on Remotion for any for-profit purpose, nor can anyone else, without a license.

dgreensp··on Things people get wrong about Electron
While it’s just one small section of the article, I notice there is the sort of industry-standard erasure of the entire concept of good vs bad UI or engineering, in the “Web apps are bad” section.

A “good” app, in this article, is one that is “successful, versatile, [or] capable.” Any McDonald’s kiosk is an example of a “good” app, because it’s so “successful.” If you think a much-used app has UI that isn’t “good” and would be improved by using native technologies, well, first of all, “the market” is the arbiter of what’s good, not you, and second of all, even if you are right, your criticism lacks empathy for the “requirements and constraints” developers face, you are ignorant about the fact that things have tradeoffs, or you are some perfectionist who thinks everything needs to be a work of art.

Software engineers are so preemptively defensive about quality.

dgreensp··on Populism and the World of Oz (2016)
I can’t say it’s not possible there is symbolism such as “the scarecrow represents farmers,” so when the scarecrow is elevated to the position of Ruler of the Emerald City, that symbolizes something.

At the literal level, though, the government in these books is an imperialist monarchy. We read all the books to my ten-year-old, and he had criticisms of this without us saying anything, pointing out it didn’t seem right for entire races/cultures that had never even heard of Princess Ozma to be considered her subjects and subject to her rule.

dgreensp··on Most people don't care about quality
By this article’s own standards, I can’t really fault it for jumping to half-assed conclusions, as long as a casual reader might not notice or care. I’m not being glib. Devaluing quality is a choice, and if it’s fashionable to say it’s fine that everything’s being enshittified (because then we can feel smart, rather than disappointed), I’d rather it wasn’t.

The flaw is in extrapolating from the fact that different situations require different levels of craftsmanship or attention to aesthetics, standards, etc—and something fancier or more meticulously or expertly made, like a meal from a Michelin star restaurant, might not even be more enjoyed by the particular person who experiences it—to make the point that the things only professionals and those versed in their field know to do, which a layperson might not even notice, don’t really matter.

The reality is, whether we enjoy the products and services and experiences that come our way or not is largely due to design decisions beyond what we can consciously attend to and appreciate. It’s all those little details.

There is a limit to how low-quality content can be, and it’s actively being explored with the help of AI.

One of the effects of capitalism in the US used to be that companies were trying to make higher-quality things for less, because of competition. The strategy now, by the big players, is to just sort of swallow everything up, and then as quality gets lower and prices get higher, consumers feel compelled to just roll with it, because the shift is happening everywhere all at once. In corporations’ ideal world, there is no social contract, no real market forces, the consumer has no leverage, the corporation puts in as little effort as possible, and the consumer pays as much as possible, and still buys the thing, because what else are they going to do.

You can put ice cream in front of someone and they’ll eat it, so let’s just all eat ice cream for breakfast, lunch, and dinner.

dgreensp··on Show HN: Experiments in AI-generation of crosswords
I found this article a bit disappointing.

The link at the bottom doesn’t work.

The grids shown do not follow the well-known rules of (American) crosswords: every square is part of two words of three or more letters each.

Coming up with a pattern of black squares, and writing good clues, are two parts of making a crossword puzzle that are IMO fun and benefit from a human touch, and are not overly difficult. There are also databases of past clues used in crossword puzzles (eg every NY Times clue ever, and various crossword dictionaries) for reference and possible training. If you don’t care about originality (or copyright) and want quality clues, you can just pull clues from these. If you do care about all those things, you can surface the list of clues used in the past to the human constructor and let them write the final clue. Or you can try to perfect LLM clue-writing. In my experience, LLMs are terrible at clues. Like sometimes if I try to give it feedback about a clue, it will just work the feedback into the clue… it’s a little hard to describe without an example, but basically it doesn’t seem to understand the requirements of a clue and the process of a solver looking at a clue and trying to come up with an answer.

Coming up with an interlocking set of fun, high-quality words and phrases is the hard part. I agree that LLM wordlist curation is a great idea, and I started playing around with that once.

Beyond that, I don’t think LLMs can help with grid construction, which is a more classic combinatorial problem.

dgreensp··on Grammarly acquires Coda
I had that impression when I joined Coda; not so much by the time I left.
dgreensp··on Grayjay Desktop App
The site CSS is a little broken on iPhone, causing elements to hang off the screen or overlap. FAQ link gives a 404.

Technically, I think this is against YouTube (for example) TOS, though I don’t expect that would be enforced against end users.

dgreensp··on Float Self-Tagging
It’s also worth noting that engines with JIT, like V8, don’t box intermediate floating-point numbers in calculations after optimization has been performed. Arrays of (only) numbers also don’t box the numbers (though all numeric object property values that aren’t small integers are now boxed in V8). This means you can read floats from arrays, do math on them, and write them to arrays, without boxing or unboxing. This wouldn’t necessarily be true in a less sophisticated runtime that lacks an optimizing compiler (of which V8 has two or three IIRC).
dgreensp··on Show HN: Erin – Open-source and self-hosted TikTok feed for your own videos
Yup, it's a sketchy-sounding concept. Is it for "your own videos," or TikTok videos? Why does it have the TikTok logo in it??

It looks like it's a fork or repackaging of someone else's TikTok "clone."

The author's writing is fluent, but the whole thing is kind of odd. This is the Internet; the developer could be 17. Who knows.

dgreensp··on Accountability sinks
Came here to say this.

And even in the case where the company's decision is arguably just "bad," it still might not be a problem from the company's point of view.

Companies (including start-ups) create buggy products all the time and don't care, and aren't very responsive to requests for support, as long as money is coming in. I don't think they are using special accountability-flushing techniques. It takes real work, intention, experience, and power in a company to create feedback channels, and use them, and ensure that the customer has an experience of quality. It doesn't happen by magic or by default.

dgreensp··on Y Combinator Traded Prestige for Growth
Full disclosure, I'm a YC alum whose last start-up was acquired by Google, who applied to this batch and didn't get an interview.

YC is not the stamp of quality it once was, to be sure, though it still works as social proof, because investors (especially VCs) want to invest in companies other investors like (or failing that, companies they imagine other investors would like). YC would say that they aren't trying to be a stamp of quality or social proof, they are just trying to help start-ups.

If I had to take a stab at articulating how YC has changed, it's that it's become a VC, picking ideas to generate returns.

This rejection email from 2022 that someone posted online cements the idea for me (excerpted):

"Unfortunately, we've decided not to fund {Company} this batch. We enjoyed our conversation today and were impressed by you as founders building something they are passionate about bringing into being. However we weren't convinced that this product strategy is going to yield a big company in its current form. ...

Of course, things are very early and you are still figuring out the right way to build and structure your business. If you're able to make significant progress with it, we'd be very interested in hearing about it for a future batch."

Original YC would not reject founders over their current product strategy and because they haven't figured out the right way to structure their business--haven't put together the "proof" that they could be on track to be a unicorn, etc. That's a VC rejection. Of course, YC gets many thousands of applications and has to reject most of them. You could say they shouldn't be criticized for trying to give a little feedback.

It's just hard to convey the sense from the early days of YC that they really didn't care about the return, or the progress so far, or VCs, or fads, or anything. That said, I really was at the right place at the right time and got very lucky.

Finally, I want to call out the phenomenon in the world of VC where the ability to generate hype alone is enough to make you and your investors a lot of money, even if there isn't a lot of substance in your company (and even if things are morally or ethically questionable), through the mechanism of greater fools. Cryptocurrency is a whole exploration of this effect, turning hype into money by building a streamlined mechanism for bringing in greater fools, but we can also look at examples like Theranos. (There's actually a ton of money at the top of society looking for somewhere to go, which ideally would be routed to more worthy ventures than it is, but that's a whole topic in itself.) The point is, the greater fool "strat" works. In a moral vacuum, if you are trying to maximize returns, you lean into it.

If I were running a fund like YC with basically unlimited money at my disposal, balancing the goals (strategies?) of 1) make money for money's sake, 2) advance technology for technology's sake (let's go to Mars, etc), and 3) make the world a better place, I would focus on (3). Like can we at least try to build a unicorn that feeds people, and maybe fail, rather than trying to build FooBarBazCoin that eats the word, and failing? We mostly see a mix of (1) and (2).

dgreensp··on Visual Studio Code is designed to fracture (2022)
It takes a LOT of reading to find out what the actual point is here, and the concept of a “fractured” “ecosystem” is brought up a dozen times without being explained. A “venus fly trap that is designed to fracture”? What in the heck.

VS Code is an IDE you can download and use for free from Microsoft. It’s not some magical open-source platform/ecosystem thing that anyone can use for anything, which Microsoft has no control over. It’s a product.

It seems like everyone wants to make “universal” developer _services_, but no one wants to build or fund an IDE, or it’s just too hard or something. That’s not Microsoft’s fault.

dgreensp··on Why Apple Uses JPEG XL in the iPhone 16 and What It Means for Your Photos
Ok wow, that was not clear from the article. It also seems kind of confusing and unfortunate that Apple is using so many different image formats.
dgreensp··on Design Patterns Are Temporary, Language Features Are Forever
You don’t come across visitors every day, and they aren’t something to reach for instead of a simple switch, but if you are writing a Webpack or ESLint plugin or something like that, you might encounter them for traversing an AST. It’s not just an OO thing, either.

I learned the visitor pattern in pretty ideal circumstances: from a classmate, back in college, when we were writing a compiler together for a class project, and it was just what we needed.

The right conditions for a visitor are when you have a data structure with many different “node” types, and the traversal logic is not simple. In addition, you will be doing many different traversals. You might also be doing transformations on the structure. You might have a library running the traversals/transformations provided by the code that uses the library. There could be correctness or memory management considerations involved that you don’t want the client to have to worry about. You might want to “pipeline” multiple transformations, doing several manipulations in one traversal, in an efficient way. Common traversals might care about only certain types of nodes, like a particular lint rule might just be looking at string literals or references to global variables, but it needs to walk the entire syntax tree of a file of source code.

dgreensp··on LM Studio 0.3 – Discover, download, and run local LLMs
I filed a GitHub issue two weeks ago about a bug that was enough for me to put it down for a bit, and there’s been not even a response. Their development velocity seems incredible, though. I’m not sure what to make of it.
dgreensp··on State of Text Rendering 2024
Wait, so the idea is in the future, fonts will contain arbitrary full-color vector and bitmap images, they will contain Web Assembly code that needs to be run, and they will be streamed over the network? I’m probably missing a few other proposals, as I only skimmed.

Did anyone stop to consider if this is really necessary? The author makes it sound like he has used his influence steadily over the years to make fonts more complicated. “In year X, I proposed that fonts be able to do Y, because why not?” I get that text shaping is so complex, that in terms of open source, there is just Harfbuzz. I’m not an expert in this area. But I don’t think it’s a good thing if “font standards” are constantly getting new features, like web standards, and font renderers are like mini browser engines, where the sheer scope and number of features and rate of new features keeps everyone using the same codebases.

dgreensp··on Ladybird Web Browser becomes a non-profit with $1M from GitHub Founder
Where on the roadmap is GPU compositing? In modern browser programming, I kind of take for granted that I can control the rendering “layers” and certain CSS properties, like “transform,” will be accelerated.

Edit: In Blink, the layer/compositing system extends to SVG elements inside SVG tags, as well, and in WebKit, it doesn’t yet, but there is an active years-long effort going back to 2019 that will eventually land: https://youtu.be/WxqJFxiprrU?si=dhQIgW1V4yS_Ca4s Compositing and using the GPU seems like a complex but important part of rendering in a browser, and a case where it could be good to implement the kind of system that other browsers have arrived at after years of iteration, when it comes time to do so.

Will the JS engine still be LibJS?

dgreensp··on What UI density means and how to design for it
The author doesn’t deliver on the promise made early on the essay to examine the question: Why have UIs gotten so sparse?

It’s like entire web design world decided more whitespace is better, and won’t hear anything about it. And now some desktop apps are being designed like web apps. Or take Hulu on my Roku, it will literally only show the first three lines of a four-line movie synopsis, surrounded by tons of space, and make you expand it, via multiple button presses on the remote.

I once implemented a file list view for a start-up I was working for, off of a mock from the designer, similar to what you see when you are browsing Google Drive or Dropbox in a web browser, but with only one view, a list view with very large icons. The amount of whitespace was massive; use of screen real estate extremely poor. But then, these web UIs never look like the Finder, or Outlook, do they? They could. They could feel almost as snappy, too, with some allowance for network roundtrips. The Finder is actually pretty slow these days, even on a high-end MacBook browsing local files. There are lots of pauses and stutters.

There’s an unspoken rule against labeling things, too, if you can use a row of inscrutable icons instead.

There will always be designers experimenting with taking things away. Apple has done it plenty. What if there were no scrollbars, no ports, no home button or menu bar (just swipe from the edge of the screen), no keyboard, no headphone jack. Sometimes it’s a bold direction, occasionally a clear net negative. But minimalism is a thing, it just has to be tempered.

Often, design “trends” are just trends, in my (admittedly cynical) view; there isn’t necessarily any merit at the core, or the people propagating them seem more interested in conforming to trends than asking what is good. The dynamics of fashion are easy to underestimate as an engineer.

People love to copy things. People who grow up watching action movies and become action movie directors just want to make an action movie with all the action movie tropes. That’s the main thing, not necessarily picking the things that work well in action movies and bringing in some things that just work well, period, that are original or timeless, like good directors or writers do.

Besides people loving to copy, people sometimes think following trends is why something will sell. If you’re making clothing and you aren’t hip to this year’s styles and colors, no one is going to buy your stuff, is the sentiment I presume. It can easily become overwhelmingly about conforming to trends. Designers also tend to put famous people and sources of influence on pedestals and think they will never be 1/10 of the genius of, say, the person who decided that some famous building should look like a pile of mashed potatoes, or that an Apple billboard should just be black text centered on a plain white background. In art, as in philosophy, there is so much pressure to agree that certain people are good, regardless of any objectivity or lay opinions—so much focus on status—that to even think of what you are doing as potentially-good-in-others’-eyes, you need to copy someone or some brand with high status, or somehow attain your own status, is how I think people sometimes feel.

In other words, designers of UIs might falsely think the customer cares about trends and fads, and that their work will be evaluated through a system of reference points and status divorced from actual merit, as can happen in the design world and adjacent spheres (art, fashion, etc).

dgreensp··on Figma’s Journey to TypeScript
They don’t really mention the ongoing developer experience impact (even if it is outweighed by the popularity of TypeScript) of losing Skew’s niceties, they just talk about the one-off transpilation of them when migrating the codebase. For example, the fact that it’s easy to end up with files that need to be imported in the right order in TypeScript, or things will break; or the fact that destructuring is slow and so should not be used (when performance is at all important). I know from using TypeScript for years that there are dozens of these gotchas (some inherited from JavaScript, some not), requiring an extensive style guide at the very least, especially if you have a lot of engineers.

I wonder if some engineers were sad to see Skew go.

dgreensp··on For twenty years, PostSecret has broadcast suburban America’s hidden truths
I’ve been a fan of PostSecret since the beginning and bought the (first?) book. Many years later, Facebook sent me some PostSecret posts. Do not read the Facebook comments. There isn’t the “art therapy” safe space that you usually feel witnessing Frank’s exhibit, just a bunch of moral judgment. “This was written by a bad person.” “They shouldn’t be glorifying this.” “This person needs help.” (Ya think?) It’s a really interesting contrast, people being vulnerable and people feeling morally obligated to condemn or silence them.
dgreensp··on 4 Internal Apple Emails That Helped the DOJ Build Its Case
What I meant was, the case will be decided by applying the law to the situation, so in order to discuss from the sidelines what could, should, or might happen, we need to look at what it says and how it might be interpreted.
dgreensp··on 4 Internal Apple Emails That Helped the DOJ Build Its Case
I want the government to have a case, here. I want the DOJ to be able to push back against what companies like Apple are doing.

However, in court, you have to compare what a company is doing to the letter of the law. It seems US anti-trust laws are vague enough that, if the way big tech companies operate today violates them, the enforcement over the years has been so little that it is hard to apply them now in the spirit they were intended. That’s my impression, anyway.

Apple will, of course, argue they are not a monopoly (all their products have competitors, like iOS has Android—even if they tried to kill it). They will argue that having an ecosystem of products in different categories that work together is not illegal, and that executives wanting to entice users to use their products, or deter them from switching, is all the same thing and just good business, not anti-competitive behavior.

If it’s already illegal for Apple to prefer and promote their own apps and products and extract exorbitant rent from everyone else, then please, DOJ, put a stop to it.

Maybe it’s just all the articles about this that are really poor. As it stands, not having dug into the matter deeply, I’m on the DOJ’s side, but I’d rather be a lawyer for Apple.

dgreensp··on Amazon Joins SpaceX in Mission to Destroy NLRB
It really doesn’t have to be this way.

Capitalism is like a game. For rich, greedy bastards, it’s literally a Monopoly game, while for the less fortunate, it can be life or death. But the point is, it has rules. Trying to change the rules and get rid of what feeble protections and regulations we have, in order to make more money, is just evil (in my personal opinion).

Corporations are not people and don’t have constitutional rights. I mean, apparently it is the case that courts have applied rights like the First Amendment to companies in court cases. But this should obviously be a very limited thing. We don’t need businesses to have a right to “pursue happiness.” It’s just a business.

dgreensp··on Why it's impossible to agree on what's allowed
> The idea behind the site is that it's very difficult to get people to agree on what moderation rules should apply to a platform

There's a big difference between "what moderation rules should apply" and "how do I interpret this one vague rule." Questions like "can I play with a toy car in this particular park" are answerable. If some people want the rule to be different, that's another story. People's ability to interpret a vague rule out of context doesn't really bear on the question of how hard or easy it is to agree on rules for a community, or to write good rules that are easy for moderators to apply.

Not every single person in a community needs to like a rule. And if rules are not specific enough, they can be made more specific.

dgreensp··on Spot Bitcoin ETF receives official approval from the SEC
Well the claim is that if HN isn’t bullish on crypto in 2024 it must be inside a strange “negativity bubble.”
dgreensp··on Spot Bitcoin ETF receives official approval from the SEC
People who aren’t in the tech/VC or crypto echo chambers can see that it’s (still) overhyped and mostly scams. They don’t really care or have more positive sentiment than HN.
dgreensp··on Ambient Co-Presence
I agree that seeing where someone is reading (with blobs moving around in your visual field while you are trying to read) is not good. It doesn't correspond to any physical experience of being in a room with someone, or any positive, helpful experience of collaboration. There's a reason we don't attend lectures with laser pointers strapped to our heads, to make visible where everyone's eyeballs are looking on the projector screen or whiteboard. When I'm reading a magazine in a waiting room, I'm not socially isolated by the fact that the reading experience is not a synchronous collaboration with other people around the world.

The fact that this idea was even proposed and explored makes me think there are some common-sense principles and ground rules that are not being followed, like a first filter after brainstorming new digital experiences. For example, hmm, brainstorming here, part of working in the same room as other people is you can hear their phones buzzing when they get texts. What if we made an app that gets notified when one of your friends' phones buzzes, for any reason, and it makes your phone buzz at a lower volume, so it's more like you're in the same room even if you are on the other side of the world? No, that is a bad idea because it would just be annoying and distracting.

One should ask, under what circumstances do people currently read the same thing at the same time, in each other's presence? What creates a feeling of presence in this case? What makes the experience feel better or worse? Do we even care what strangers are reading? In a meeting, people are usually reading on their own screens or pages, or one person reads to another or to the group. There are situations like reading the same sign at a museum exhibit or art gallery, where there is presence with a stranger (which may or may not even be a positive thing). I think the first thing my brain needs to know is how many people are standing around next to me. Reddit threads have a little indicator that says something like "17 people here." To have a blurry cursor blob suddenly moving across my screen without an indication of how many people are "here" is disorienting. That doesn't happen in real life unless someone is out of sight playing with a laser pointer or flashlight, or there's a sniper.

I really would hope this would be 101-level thinking for this field.

I've worked at workplaces that had a video "portal" between two office locations, and that worked fine, if you could choose whether you want to work in view of the camera.

dgreensp··on Interview about Austral, a systems programming language with linear types
Yes, this. The notable thing in Austral is not that linear types are used somewhere, it’s that it’s all linear, even when destruction is just deallocation and could easily be done by the compiler (even in the presence of exceptions).
dgreensp··on Interview about Austral, a systems programming language with linear types
As someone working on a language myself, I found this to be very high-quality material! It aligns with a lot of my thinking, and it’s educational.

A random part of interest: I followed the link about how linear types and exceptions don’t mix: https://borretti.me/article/linear-types-exceptions In it, the author explains how linear types always need to be explicitly destroyed, and if you end up in a “catch” block, you can’t easily go back and destroy things that are now out of scope, which control flow got interrupted while doing things with. “Why not just destroy things when they go out of scope?” I thought. The author addresses this, saying that types that you have to “consume” “at most once” are called affine types, and the compiler can clean them up for you, and it solves the problem with exceptions. Rust has affine types, and you don’t have to explicitly “destruct” every string/object/resource using the appropriate destructor for that type, manually, as you do with linear types (of course, the compiler will make sure you do it, but you have to do it).

So why does Austral use linear types, not affine types? The reasons given do not really resonate with me (but I’m not a systems programmer); it’s that the compiler would have to insert hidden function calls, and secondarily, for temporary/intermediate values, the order of invocation might not be obvious; plus, maybe the programmer not doing something with a value is a mistake. I’m really glad the reasons are written out, however.

In other areas I feel very aligned with the author, such as on the value of required type annotations and local inference rather than global inference, and I’ve saved the link to refer back to the way the opinion is stated.

dgreensp··on Memory safety is necessary, not sufficient
I think it’s worth emphasizing that the C spec’s love of undefined behavior—if you do X by accident, anything can happen—and the apparently massive amount of memory-unsafe software that has been written that will just allocate 16 bytes on the stack and then read from a file descriptor until it encounters a null byte… are examples of things that aren’t considered remotely sane or reasonable to a modern programmer or language designer. Any vague notion of “unsafe” in the context of a modern language—like saying well, if there’s a syscall, maybe something unexpected could happen—doesn’t compare to the bad decisions made by C and C programmers/culture that affect us today because we still use C code in our things.

The deep, idiosyncratic flaws of C trace back to “worse is better.” Few people remember or look up what “worse is better” actually meant. Wikipedia wrongly says it’s a “less is more” sort of thing, and some people think it’s about not being a perfectionist.

But actually… actually actually, if you read the essay, it says “worse is better” means (paraphrasing) it’s more important that C compilers be easy to implement than easy to use. Also, it is more important that the implementation—or the design of the implementation—of a piece of software be simple than that it be correct. It is more important that it be simple than that it be consistent. It is more important that it be simple than that it be “complete” (for example, handle edge cases; it just needs to work in “most situations”). This is not just anti-perfectionism, it is an objectively terrible set of engineering values. But there were so many different kinds of computers and operating systems back then—you didn’t even know if a byte was 8 bits—that it helped a lot that C compilers could do whatever they wanted in many situations. And making it easy for C compiler implementers enabled C to spread far and wide. It was a very different world, and just being able to write in a higher-level language than assembly on a particular computer was a big deal.

← PreviousPage 2 of 30Next →