43 karma · joined November 14, 2017
This give me an idea. Here's my smudge pattern that works better: (shows a diagram with blotches in the shape of Rick Astley singing).
First reason, LLMs are modeled from what humans have been doing, and the have been writing software that way recently so it's easier to mimick that to get straight to results. This reason might fade away in the future.
Second reason, something related to impedance (mis)match, a signal processing notion (when the interface between two media is not well-suited, it is difficult to have a signal pass through).
Going through intermediate levels makes a structured workflow where each steps follows the previous one "cheaply". On the contrary, straight generating something many layers away requires juggling with all the levels at once, hence more costly. So "cheaply" above both means "better use of a LLM context" but also use regular tools where they are good instead of paying the high price (hardware+computation+environment) of doing it via LLM.
Interestingly, AIs are used to generate sample-level audio and some video, which may look like it contradicts the point. Still they are costly (especially video).
Submission title here is the HTML title. Visible title in the article "The Mother of All AI Supply Chains: Critical, Systemic Vulnerability at the Core of Anthropic’s MCP"
For eons the standard was: the only one with the title bar showing the theme accent color. That is consistent, predictable, keeps the user in the flow.
With CSD each app does whatever inconsistent thing they can fancy. You type and oops deleted something in the wrong window.
Alas now even many default SSD setups fail at this (selected and non-selected windows look pretty much the same) and keyboard-first workflow is much hindered.
(Nod to https://stackoverflow.com/questions/2497801/closures-are-poo... yet real question.)
In 2024, shouldn't hacker news visitors get that receiving a passive document to display locally is a thing, and having to run unaudited code from random strangers on your machine all the time is quite another? Then reopen the conversation about browser security?
Or, instead, get that simplicity is our ally, but unneeded complexity is the root of much evil?
Visit e.g. any stackexchange site with script disabled and be enlightened.
> It's minimal and most code is on the server. Also it's a small footprint and proven to be fast (Preact). This site has a 100 performance score on Google Pagespeed,
Yes, it's nicely minimal in a number of aspects. Good point.
> JS doesn't mean bad UX.
For those who run their browser with scripts off by default it is.
Disabling JS and cookies by default has another nice benefit: on a number of sites, the stupid content-obcsuring cookie banner just does not exist (as well as ads, of course). And pages load time is much better.
> Blocking third party scripts and cookies is one thing, but not wanting first party script just forces the developer to use server-side templating for their static content which means you either deliver everything from the same origin (bad for latency) or chain a CDN in front that picks out static content to cache (I don't think that's compatible with the web 2.0 ideology).
That doesn't hold.
If a server can format data into a Json, it can format the same data into HTML and serve it the same. In the specific case of this site, the generated HTML has the same lifetime as the generated Json.
From the site: "It is updated every 2 hours."
Ironically, because this site is nicely minimal in a number of aspects, one can play with statistics. Not counting cloudflare beacon js, I measure
1.16k HTML, 1.56k CSS, 18.44k JS, 1.05k favicon.svg, 1.61k JSON payload.
Therefore 77% of bandwidth used is Javascript:
p=(1.16+1.56+1.05+1.61) ; j=18.44 ; 100*j/(j+p)=77
By serving HTML instead of Json (at the same price), the bandwidth cost would be divised by 4.4: p=(1.16+1.56+1.05+1.61) ; j=18.44 ; (j+p)/p = 4.4
Think about it: running custom code on each browser of each visitor to perform exactly the same computation and yield the same DOM each and every time is totally useless.All in all, serving a full HTML page instead of a HTML shell + custom JS + Json totally makes sense.
The good news is, for the past year I have started to hear seasoned web developers complain more about the sad state web development has been for years, and longing for some simplicity. Wait and see.
Kudos to the author for embracing minimalism. You can go a step further.
Well, for one HN works without Javascript, while that URL yields just a blank page.
It started answering, then within seconds my question was replaced with "This content may violate our content policy. If you believe this to be in error, please submit your feedback — your input will aid our research in this area."
Then, seconds after, the still-appearing answer was replaced with the same message.
Doh! The content filter got tripped because "obviously" it's not a philosophical thought experiment about utilitarianism but an evil text about mugging someone, which is an illegal activity. What a time to be alive!
1. From the first page:
> please answer the question of whether the rule is violated (not whether the violation should be allowed).
In other words, if a question is about some kind of vehicle that you think should be allowed to go, overriding this rule, the answer is "yes, the rule is violated".
2. There is a twist in this game and it's best to not tell which it is before you try the game. (The twist is acknowledged in the result page, 4th paragraph from the bottom, not counting the P.S. line at the end).
But after checking out some Unreal Engine 5 demos and looking for glitches, I am now convinced that this is indeed a bona fide game.
The game plays off the fact that realistic flaws can trick our brains into thinking it's real-life, not CGI.
Right from the start, with the blown-out sky, wide-angle shot, and imperfections on the building, it's clear that this is a real-world setting. Yet it's not.
Around the 0:05-0:06 mark, you can see how a low dynamic range camera would react to range variation via exposure change. This is a departure from the usual method of processing high dynamic range in games, which involves pre-processing the world into a low dynamic range.
This news deserves more visibility on HN. I upvote!
Awesome, by the way.
Do you mean you have tested for usual https://en.wikipedia.org/wiki/Dichromacy and showed that you see all shades people usually see, or have you definitely tested for https://en.wikipedia.org/wiki/Tetrachromacy ?
Around 1994, I've seen and played a game on a 8086 machine with a VGA card and full-screen perfectly smooth scrolling. I don't remember the name but it was a 2D-from-above toy-car racing game where each level/race had a different theme.
* "app icon, document icon and welcome window",
* "ability to simultaneously edit multiple shapes of the same type",
* "minimum OS requirement has been raised to macOS 11 Big Sur".
For an overview of the software, see this URL: https://monodraw.helftone.com/