HNHacker News
TopNewBestAskShowJobs

toomim

3,057 karma · joined May 5, 2012

Independent Scientist in Computer Science and Psychology

https://invisible.college/@toomim

https://braid.org

submissionscomments
toomim··on Why Jony Ive left Apple to the ‘accountants’
You must mean software vs. hardware, not software vs. design.

Software is designed. Hardware is designed. Design is not how something looks on the outside. Design is how something works.

toomim··on Twitter re-examines Elon Musk’s bid, may be more receptive to a deal
No, the banks forced him to settle. It wasn't the SEC. Watch the video.

The banks didn't want a liability. The SEC disagreement was going to take too much time to resolve legitimately. If he didn't please the banks they wouldn't give him funding.

toomim··on New on Reddit: Comment search, improved search results relevance, search design
And you can install the "old reddit redirect" extension in chrome to make this automatic.
toomim··on Why the WHO took two years to say Covid is airborne
Having no consensus is more accurate than false consensus.
toomim··on HTTP Feeds
Yes, the limit is implemented in browsers, but it's for the protection of servers. The idea is to prevent a web app from DDOSSing a server somewhere.

This is especially a problem for servers implemented with one thread or process per connection, rather than an async event queue like nodejs, because each thread takes up a few megabytes of memory, and if a browser opens 1000 connections, that's 2 gigs of memory right there.

But if a server is implemented properly to support long-lived connections (by using node, or greenlets, or other async system) then it should be able to opt out of that browser protection. A standardized header would allow a server to opt out of it.

toomim··on HTTP Feeds
> Thus, I think a "standard" should encourage a cursor concept that can use something that may reliably provide _all_ changes.

I think the Braid standard might support what you want: https://datatracker.ietf.org/doc/html/draft-toomim-httpbis-b...

We don't call it a "cursor", but the idea is to encode whatever clock or versioning information you want as an arbitrary string, with the only general requirement that the string is unique to that version.

You can use this with Lamport clocks, Vector Clocks, Version Vectors, Wall Clocks, (Agentid+Sequence) pairs, SHA hashes, or even blockchain transactions.

toomim··on HTTP Feeds
That'd be great, but is a lot harder than adding a header. HTTP/2 changes the entire wire format into a binary protocol that multiplexes multiple streams over a single connection. It's a lot harder to implement than HTTP/1, which you can type out manually over telnet:

    telnet news.ycombinator.com 80

    GET / HTTP/1.1
    Host: news.ycombinator.com
    \n
    \n
Adding a header could be as simple as:

    HTTP/1.1 200 OK
    Content-Type: text/html; charset=utf-8
    Multiple-Subscriptions: OK
toomim··on HTTP Feeds
Perhaps we should define a new HTTP header for servers to say: "It's ok for clients to initiate N connections with me" so that we can write web apps without this restriction.
toomim··on HTTP Feeds
Thank you for writing this great spec up for others to use!

I think totally right that back-pressure and plain GETs are an important use-case to support, and am really happy to see a beautiful spec written up to articulate concretely how to support them.

It is also great to be able to switch amongst these methods of subscription, for instance, if your server can keep a persistent connection open, it's nice to be able to get realtime updates over a single channel, but to still be able to fall back to polling or long-polling if you can't. And if you switch between a polling and a subscription, it's nice if you don't have to change the entire protocol — but can just change the subscription method.

There's a growing effort to integrate all our various subscription-over-http protocols into an IETF HTTP standard, that we can use for all these use-cases: https://lists.w3.org/Archives/Public/ietf-http-wg/2022JanMar...

Maybe you'd be interested in incorporating your experience, use-cases, and design decisions into that effort? We have been talking about starting at this November's IETF. [1]

For instance, you can do polling over the Braid protocol [2] with a sequence of GETs, where you specify the version you are coming from in the Parents: header:

    GET /foo
    Parents: "1"

    GET /foo
    Parents: "2"

    GET /foo
    Parents: "3"
Each response would include the next version.

And you can get back-pressure over Braid by disconnecting a subscription when your client gets overwhelmed, and then reconnecting again later on with a new Parents: header:

    GET /foo
    Subscribe: true
    Parents: "3"

    ..25 of updates flow back to the client..

    **client disconnects**
Now the client can reconnect whenever it's ready for new data:

    GET /foo
    Subscribe: true
    Parents: "28"
Or if it wants to re-fetch an old version, it can simply ask for it via the version ID it got:

    GET /foo
    Version: "14"
And if the source for these updates is a git repository, we could use a SHA hash for the version instead of an integer:

    GET /foo
    Version: "eac8eb8cb2f21c5e79c305c738aa8a8171391b36"
    Parents: "8f8bfc8ea356d929135d5c3f8cb891031d1539bd"
There's a magical universality to the basic concepts at play here!

[1] https://www.ietf.org/how/meetings/115/

[2] https://datatracker.ietf.org/doc/html/draft-toomim-httpbis-b...

toomim··on “This shouldn’t happen”: Inside the virus-hunting nonprofit EcoHealth Alliance
This comment is so good that I'm replying to it just so that it appears in my /comments history.
toomim··on Reinstating our SAT/ACT requirement for future admissions cycles
> the idea that my success could be predicted by my knowledge of geometry and vocabulary words is laughable.

You really should look into the g-factor research. This isn't about knowledge, but rather performance. It turns out that your performance on geometry and vocabulary tests is highly correlated with your performance on tests in such disparate fields as Classics and Music:

https://en.wikipedia.org/wiki/G_factor_(psychometrics)#Cogni...

The crazy result in general intelligence is that your performance in all these areas is highly correlated, and incredibly correlated with career success:

> "Research indicates that tests of g are the best single predictors of job performance, with an average validity coefficient of .55 across several meta-analyses of studies based on supervisor ratings and job samples."

toomim··on San Francisco VC firm required job applicants to take online IQ test
The SAT does.
toomim··on Apple’s Universal Control
It used to be so great... :-(
toomim··on A Bitcoin Mine at a Natural Gas Well in Texas
Often because there is no way to get the gas to a power plant. There might not be a local power plant that uses natural gas, or there might not be pipelines, or the local plant might not need the gas.

Methane pops up in many places, and in order to use it for the power grid, you need it to pop up in a convenient place for the power grid.

The only other options left are to flare it, or vent it into the atmosphere. Methane is 40x worse for the environment in its uncombusted form than after burning it, so it's better to burn (flare) it. But sometimes the extractors don't bother. Bitcoin mining provides an incentive for entrepreneurs to seek out methane that isn't being burned, and burn it.

toomim··on San Francisco VC firm required job applicants to take online IQ test
There are many structures for IQ tests. Some use analogical reasoning, some use arithmetic pattern matching, some use vocabulary and language skills. You are perhaps thinking of the style of IQ tests that require you to fill in analogies and rotate shapes together.

But what makes an IQ test is the theory that all performance on each of these skills correlates together strongly, indicating a single "general intelligence" factor (aka "IQ"). [1] Reading comprehension turns out to be highly correlated with mathematical aptitude, and musical aptitude, etc.

So the main criteria for an IQ test is not how it looks, or how it's structured, but whether it measures knowledge that has been acquired (aka "nurture") or ability to reason about new things. Since this test is measuring reading comprehension on topics that are unfamiliar to the reader, it is measuring the latter, and is likely to have a strong correlation with g, and thus looks like an IQ test.

[1] https://en.wikipedia.org/wiki/G_factor_(psychometrics)

toomim··on San Francisco VC firm required job applicants to take online IQ test
Perhaps you've never interviewed someone.

I am an employer. When I interview job candidates, I ask them about their lives, about their day, where they grew up, and lots of questions because I am starting a relationship with them.

The interview is not just a place to filter employees. It's also where you begin a relationship with them. You are envisioning how you might employ them. What kind of work could they do? Could they work well with John and Alice? How does it feel to talk to them? Do they like it when I'm direct, or when I speak softly?

I'm sorry that you've had such a cynical view on hiring, but the reality is that, yes, employers do use MBTI to get a richer understanding of their candidates. Here are some additional examples for you: https://www.talentinsights.com/blog/using-myers-briggs-in-hi...

toomim··on San Francisco VC firm required job applicants to take online IQ test
You might want to read up on general intelligence, which is the well-documented phenomena that your performance on tasks like rotating shapes in your mind correlates very strongly with other tasks:

https://en.wikipedia.org/wiki/G_factor_(psychometrics)

Look at the figures in the "Cognitive ability testing" section -- see how strong those correlations are across the disparate fields of Classics, Math, and Music!

In other words, there is a mountain of evidence that IQ tests correlate with an extraordinarily broad range of aptitudes in daily life, and especially with job performance. IQ is the single most important psychometric correlate with job performance and income of any out there; it's a much stronger signal than race, gender, birthplace, or personality for instance:

"Research indicates that tests of g are the best single predictors of job performance, with an average validity coefficient of .55 across several meta-analyses of studies based on supervisor ratings and job samples."

toomim··on San Francisco VC firm required job applicants to take online IQ test
That sure looks like an IQ test to me.
toomim··on San Francisco VC firm required job applicants to take online IQ test
They weren't filtering people based on MBTI.

MBTI is quite commonly used to characterize what types of jobs an individual might take on. If you understand a person's personality, it gives you a way of imagining how you might work with them. It's also popularly used in romantic dating, for the same reason.

toomim··on San Francisco VC firm required job applicants to take online IQ test
On the contrary, there are literally thousands of formal studies of MBTI. Here, look: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=mbti...

It is very well-studied in psychology. To say that most academic researchers don't "endorse" it is silly— it's not the job of a researcher to endorse a particular survey instrument, but rather to dispassionately study the mind using the best instruments available. That said, the Big-5 personality scale is certainly a more standard survey instrument in psychology, and it really doesn't matter too much which survey instrument you use, because they all correlate together. They just differ upon where you draw the coordinate axes in the space of personality; but they're all measuring essentially the same space of personality.

toomim··on Server-sent events, WebSockets, and HTTP
Your question is the topic of the article [1] Mnot was replying to.

[1] https://news.ycombinator.com/item?id=30312897

toomim··on Server-sent events, WebSockets, and HTTP
Your use-case sounds perfect for Braid: https://braid.org

This works like SSE, but is designed specifically to articulate changes to the state of HTTP/REST resources.

If you're in Javascript, you can use the braidify library: https://www.npmjs.com/package/braidify

toomim··on Server-sent events, WebSockets, and HTTP
These are great insights by both Mnot and Josephg. I've followed up on this conversation on the HTTP working group mailing list:

https://lists.w3.org/Archives/Public/ietf-http-wg/2022JanMar...

toomim··on Server-Sent Events: an alternative to WebSockets
And the Braid protocol, using a variation on SSE: https://braid.org
toomim··on I went back to using a ThinkPad from 2012
Nah, the upgraded 4-core T430 is plenty fast and very productive. You'd be surprised.

I used a dual-core upgraded T430 for quite a while. The problem was actually the battery life. 2-3 hours just doesn't compare to a M1 Macbook's 13 hours.

toomim··on Best Paper Awards in Computer Science over the past 25 years
True. Whereas most Computer Science communities gather to solve a particular problem (like comprehending natural language), HCI contributions tend to be the problem definitions themselves.

A common CHI paper would read "we studied a set of users X and Y and found that they tend to run into problems Z. We designed a solution W to address Z, and suggest some other work that could be done to address Z better."

The conference thus contains many papers proposing radically different ways of seeing the issues that users face in computing along with radically different ways of addressing them.

If these new-found problem definitions happen to gather enough attention for a sustained period of time, then it warrants a new conference of researchers ready to address those problems. But I don't think this happens often enough in CHI for it to break into sub-conferences very often.

(Also, hi Jason!)

toomim··on Google no longer producing high quality search results in significant categories
Wow. This should be the top comment.
toomim··on You don't need that CORS request
CORS is such an ugly web standard. It's not only a pain in you butt, but it also makes your requests slow through these extra pre-flight round-trips.

I have hope that we can remove it, though, in new versions of HTTP.

CORS only exists because XMLHTTPRequest broke the assumptions of web 1.0 servers. Suddenly any web browser loading any page anywhere could make a request to your server without the user's explicit permission, and a ton of web servers had already been built and were running under the assumption that users would only hit their endpoints by explicitly typing in a URL or clicking on a link.

But each time we design a new version of HTTP, we get an opportunity to remove CORS restrictions for it, because there aren't any servers running the new version yet.

And with Braid (https://braid.org), we're making changes that are big enough that I think it really warrants taking a fresh look at all of this stuff. So I have hope that we can eliminate CORS and all this ugliness.

toomim··on Ask HN: Who wants to collaborate?
Or indieweb. Or making improvements somehow?
toomim··on Ask HN: What is your spiritual practice?
This is the philosophy that Steve Jobs followed, if I'm not mistaken.
← PreviousPage 14 of 34Next →