Hacker Newsnew | past | comments | ask | show | jobs | submit | sxp's commentslogin

Why use a new format instead of SVG or JS+canvas? The latter would deviate from "plain text" but it's still pretty close. It avoids reinventing the wheel and a webbrower is installed on almost every computer with a display.

Another format to look at is Markdeep which has support for rendering ASCII diagrams and is very close to plain text: https://casual-effects.com/markdeep/features.md.html

Personally, I just wrote a thin wrapper around the somewhat verbose JS Canvas API that lets me write code such as

  ctx4b.save();
    ctx4b.translate(x+=10, 10);
    new Line(ctx4b).b().m(36,50).lAD(180 + 50, 56).lXY(0, -7.1).lXY(7, 0).lAD(50, 55).f();
    ctx4b.fillRect(36, 0, 10, 50);
  ctx4b.restore();
when I want to draw diagrams in my Markdeep notes.

I wasn't aware of Markdeep when I built this, but looking at it now, it's 6k+ lines of code: https://casual-effects.com/markdeep/latest/markdeep.js And then you need a browser underneath that weighs in at multiple GB (a scale that transcends lines of code metrics) By comparison, lines.love contains 2.6k LoC, and underneath it is LÖVE containing 100k LoC. This sort of analysis of the total size of the software supply chain and total hackability of the entire stack is what motivates me to look past web browsers. Once they grow past a certain size there is zero value to being open source for 99.99% of people.

(I do love that Markdeep is a single file of js without the ubiquitous pox of 1k-line package-lock.json!)


> I just wrote a thin wrapper around the somewhat verbose JS Canvas API that lets me write code such as

As I said at the start of OP, I want to draw by doodling. So writing code would seem to be disqualified.

> Why use a new format instead of SVG or JS+canvas? It avoids reinventing the wheel and a webbrower is installed on almost every computer with a display.

My goal is to move off browsers. Reinventing the wheel can also often be a good thing.

I do have tools to export documents to html/markdown/SVG: https://git.sr.ht/~akkartik/lines2.love#associated-tools But I myself am trying to live outside the browser as much as possible.


> Why use a new format instead of SVG or JS+canvas?

Because those are both hellaciously complex, and implementing a usable fraction of the functionality of either is not a job for a single hobbyist programmer. At least, I don’t think they are — I could be wrong. Even if I am, the number of programmers who can write code to draw a line given a JSON description has to be orders of magnitude of orders of magnitude larger than the number who can write an XML parser, an SVG implementation, a Javascript parser and runtime or a canvas implementation.


What about TinyVG? Seems fit for this purpose.

https://tinyvg.tech/


Oh this is awesome! I might try my hand at implementing it, though maybe not in lines.love..

It looks pretty neat. I like that it has a textual representation!

It seems to be in a half-baked stage. The Quick Start guide has broken links: https://m-creativelab.github.io/jsar-runtime/manual/quick-st...

Oops, Thank you for catching this, I will fix it.

> I doubt my students will get their way. Perhaps A.I. companions will plateau, the way self-driving cars seem to have done.

What world is he living in where self-driving cars have plateaued? https://www.thedriverlessdigest.com/p/latest-waymo-californi...


Related news:

- OpenAI claims gold-medal performance at IMO 2025 https://news.ycombinator.com/item?id=44613840

- "According to a friend, the IMO asked AI companies not to steal the spotlight from kids and to wait a week after the closing ceremony to announce results. OpenAI announced the results BEFORE the closing ceremony.

According to a Coordinator on Problem 6, the one problem OpenAI couldn't solve, "the general sense of the IMO Jury and Coordinators is that it was rude and inappropriate" for OpenAI to do this.

OpenAI wasn't one of the AI companies that cooperated with the IMO on testing their models, so unlike the likely upcoming Google DeepMind results, we can't even be sure OpenAI's "gold medal" is legit. Still, the IMO organizers directly asked OpenAI not to announce their results immediately after the olympiad.

Sadly, OpenAI desires hype and clout a lot more than it cares about letting these incredibly smart kids celebrate their achievement, and so they announced the results yesterday." https://x.com/mihonarium/status/1946880931723194389


What a great metaphor for AI. Taking an event that is a celebration of high school kids' knowledge and abilities and turning it into a marketing stunt for their frankenstein monster that they are building to make all the kids' hard work worth nothing.


Not only, by not officially entering they had no obligation to announce their result so if they didn't achieve a gold medal score they presumably wouldn't have made any announcement and no-one would have been the wiser.

This cowardly bullshit followed by the grandstanding on Twitter is high-school bully behaviour.


If they failed and remained quiet, then everyone would know that the other companies performed well and they didn't even qualify.


If they failed while not participating officially, they would have never competed at the eyes public if they didnt disclose it (doubtful, given prior decisions to prioritise hype vs transparency)


also did they self-rate themselves?


Well they keep doing it with programming, art and video, and soon probably most white collar jobs , terrifying and hurting hundreds of millions of workers in the process. I think we are way passed the point of tender sensibilities.


Google did the correct and respectful thing.

It appears that OpenAI didn't officially enter (whereas Google did), that they knew Google was going to gold medal, and that they released their news ahead of time (disrespecting the kids and organizers) so they could scoop Google.

Really scummy on OpenAI's part.

The IMO closing ceremony was July 19. OpenAI announced on the same day.

IMO requested the tech companies to wait until the following week.


OpenAI announced their results after the closing ceremony as was requested. https://x.com/polynoamial/status/1947024171860476264?s=46


> as was requested.

They requested week after?


He claims nobody made that request to OpenAI. It was a request made to Google and others who were being judged by the actual contest judges, which OpenAI was not.


IMO agreed to cancel the embargo after OpenAI announced.


OT, but please consider posting links like this using xcancel, so that non-X users can read them.

https://xcancel.com/polynoamial/status/1947024171860476264?s...


This is weasily bullshit from Brown.


Proposed question for next IMO: "Show a proof that 'after the closing ceremony' and 'one week later' are not the same unit of time"


> He also proposes that USCIS should replace the random selection of H-1B visas with a system that prioritizes highest wages....

This change is so obvious that's it's incredible that it hasn't been implemented yet. Auction off the H1-B visas based on how much federal income tax the role will pay and let the market sort it out.


The problem with this argument is that most non-spouse H1-Bs are granted to new college grads who just graduated from a US college. Brand new engineers do not have very high salaries, so if you insisted on a 'highest wages' system you'd be sending virtually every new grad from say Stanford back to their home country the second they graduate. US public policy since the 60s has been to welcome foreigners who graduate from our elite colleges.

As I say above- just get rid of visas for consulting companies instead, and everything else fixes itself


I'm not sure what you mean by "spouse H-1B" as that doesn't really exist.


>new college grads

So? The point of h1-b isn't to hire new grads. It's to find the already educated, industry proven, experienced top talent that exists in the world. How would a new grad fit that description?


H1-Bs are overwhelmingly new grads, and have been since the inception of the visa. The purpose of a system is what it does- H1-Bs are issued to new grads, hence the point of the visa is to hire new grads. You may want a different system, but that's not how it works now.

Experienced career professionals don't generally move to the US as adults, and anyways if your company really needed one you could just hire them remotely



Can't this statement be generalized from "conspiracy theorists" to "most people"? E.g, "Religious people unaware their their religion is in the minority"? Unless a person has subjected their belief to rigorous scrutiny and debate, they will probably be unaware about how many people in the world agree or disagree with it.


I think the more interesting one is "religious people unaware their religion is in the majority".


I don't believe rigorous scrutiny and debate would have this effect at all. Most people would just get defensive and entrench. There has to be another way to reveal reality to them that doesn't result in their entrenchment.


> There has to be another way to reveal reality to them that doesn't result in their entrenchment.

Before you reveal anything to them, make sure you are actually revealing reality and not just your version of the story.


There are degrees of things. It's true that many people have some beliefs that are unconventional, irrational, or even radical, but many also realise at least to some level. There are many religious people who will admit they can't prove any of it.

Similarly, in general people tend to overestimate how much people agree with them, but you can overestimate by a little or by a whole lot.

It's my general observation that many conspiracy theorists are far more sure of "the truth" as they see it than the average person. This is also why so many keep banging on about it in any venue, whether appropriate or not.


> And one way to help make that happen is to substantially increase the guest worker fees large corporations pay to fund scholarships, apprenticeships, and job training opportunities for American workers. This is something that I have advocated from my first days as a U.S. senator.

This is the proper solution. Ideally, the H1-B program would remove quotas and lotteries and switch to a pure auction program for the fees. E.g, a Dutch auction for the X thousand available slots where the fee can be deducted from federal income taxes. If companies really want the best talent in the world, they should be willing to pay for it.


That’s a great way to screw over every industry that isn’t tech.

Hospitals and labs can’t afford to match tech spending to obtain H1Bs.

If there’s going to be an auction then there needs to be an industry or job code cap too.


> Hospitals and labs can’t afford to match tech spending to obtain H1Bs.

That's the point! You can train Americans to do those jobs. Visas should be reserved for truly exceptional workers that will create enormous value in the U.S., as opposed to those doing mundane skilled work who wills imply drive down wages for Americans.


> Hospitals and labs can’t afford to match tech spending to obtain H1Bs.

This is just basic demand-supply. I don't see a problem here.

If hospitals and labs can't afford to match the salaries of other occupations, maybe you have too many American candidates or the end result isn't producing enough value.

Most of the occupations doesn't really need H1B. And that's okay.


Are there ways these other industries could increase interest from citizen workers, in the scenario that they can’t obtain H1Bs?


They obtain interest from citizen workers by increasing wages. H1Bs artificially suppress wages and make the market inefficient by removing those price signals. This creates a self-perpetuating problem that can only be "resolved" by bringing in more H1Bs, as the H1B impact on wages disincentivizes citizens from entering the field and pursuing the requisite education and training.


And why not? Check Kaisers bottom line, they’re not exactly hurting. Same for many hospitals. Also tech is not yet hiring doctors anyway and if there’s a true shortage of medical pros then by all means there will be cash to pay them.


You could come up with some other code for industries where you actually need people.

You can let capitalism do its thing, and you can put your finger on the scale for things that society needs but that are not as profitable.


There’s already a cap-exemption for non-profits.

Most countries identify industry shortages and tailor their immigration needs to bring in people to fulfill those roles.


> Most countries identify industry shortages and tailor their immigration needs to bring in people to fulfill those roles.

This makes no sense economically. Auctioning immigrant slots to the highest bidder should allocate those spots to where they'll be most valued in the economy. What the countries you seem to be describing are doing is industry-selective wage suppression.


When you are a smaller country that's more exposed to the international markets than the US, wage suppression is often good for the economy. High wages indicate lack of competitiveness, as foreign businesses can generate the same outputs cheaper than your businesses.

If we could measure externalities accurately and tax/subsidize businesses accordingly, visa auctions might lead to even more accurate wage suppression than immigration policy. But because we can't, auctions optimize the wrong measure. Some industries create more value than they can capture, while others create less.


It's a short-term skills shortage that's being filled. Nurses and doctors take time to train.


Sounds short sighted to try to solve a supposed short-term skills shortage with a fix that’s permanent for people and their descendants.


If the US wants to keep visas temporary then it’s free to scrap AC21 and EB Green Cards.


Western countries don’t have the stomach to maintain “temporary” worker programs as such. Turkish workers in Germany were supposed to be temporary as well. In the U.S., H1B is still legally a “temporary (nonimmigrant)” but was turned into a gateway to permanent immigration by executive branch practice.


AC21 was passed by Congress


The H1 visa was created in 1952, and H1B was created in 1990. To this day, H1 visas are, in the actual statute, for temporary workers under 8 USC 1101: “H) an alien (i) [(a) Repealed. Pub. L. 106–95, § 2(c), Nov. 12, 1999, 113 Stat. 1316] (b) subject to section 1182(j)(2) of this title, who is coming temporarily to the United States to perform services…”

The INA has a provision that requires all temporary workers to have non-immigration intent. Moreover, it requires UCIS to presume immigrant intent if someone files an application for permanent residency, which would make the H1B deportable. The state department created “dual intent” decades ago as a legal fiction to allow H1 immigrants to skirt this provision of the law.

Congress accommodated that somewhat in 1990 and in 2000 (with AC21). But it’s all a hack on top of a kludge. The statue now exempts UCIS from being required to presume non-immigrant intent if someone files a permanent residency application. And AC21 allows extensions while a green card application is pending.

But nothing stops the administration from deciding a green card application constitutes immigration intent, or denying an extension. Congress has never changed the wording of the law—on paper it’s still a temporary worker program. A future administration could start treating it as such at any time.


I’d say add a 25% payroll tax that is earmarked only to train and support American citizens in the industry the H1Bs are brought into.


100% payroll tax would make more sense.


> Primarily out of frustration. The dominant players in XR keep promoting their hardware as “computers”, when really they’re an iPad for your face. The most you can do is browse the web, play games, and consume content. They’re overweight and over constrained.

I'm a fan of HMD programming in general so I love this project. But it should be noted that AOSP-based XR headsets can run pseudo Linux environments via termux + X: https://www.reddit.com/r/cyberDeck/comments/fc5sfr/oculus_qu...

Also if anyone is looking for a full-sized portable keyboard, I suggest https://www.protoarc.com/products/xk01-tri-fold-bluetooth-ke... which is my primary keyboard when I'm working remotely on my tablet. It is "pocket-sized" for certain definitions of "pocket".


So I’ve actually been in XR for nearly a decade now. So I’ve tried every version of “dev in XR” I even spent the last 2 years trying to build an XR native dev platform.

The feedback was always “this is too heavy” for all devices excluding XREAL.

Also I wanted a device that didn’t feel like a hack. Something tailored to the experience.


I am very interested in this setup as well. I feel Bigscreen Beyond 2[0] has a very good balance of size and weight. But its VR and not AR. I can't wait to create something that does computing in VR well. Something that Simlula VR[1] has been trying for a while now

[0] https://www.bigscreenvr.com [1] https://simulavr.com/blog/intermediate-products/


Big fan of big screen. Darshan is probably the most level headed founder in all of XR.


That makes sense. Your compute puck based design is much more ergonomic. And if you combine it with something like SCOTTeVEST , you have a pretty good mobile XR form factor. Keep up the good work :)


>excluding XREAL

You mean you haven't tested XREAL or you didn't find XREAL as too heavy?

I'm asking because I have Viture XR Pro which are basically a clone of XREAL and I find them too heavy after a few hours of use.

I would really like a solution that wouldn't be as heavy and that would have clear displays where text would be readable from edge to edge.


It might be that your pupillary distance is bigger than what the viture can handle. I have a pair of crap glasses I'm playing around with, and supposedly you can get custom lenses that account for your pupillary distance.


> termux

I was excited about termux until about 3 years ago, when I saw that the app needed to be compiled against an ancient Android API version. Google seems indifferent to bricking this project... Today it's still breaking Fdroid updates:

https://github.com/termux/termux-app/issues/4120#issuecommen...



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: