268 karma · joined March 9, 2025
■ True fan of HN since ~2010
■ Interested in programming, business, design/architect/engineering, philosophy, music, gym, cooking... anything of beauty, utility, and inspiration/relax!
Also, critical business processes are great expressed as checklists.
That said, I'd still listen these two guys (+ Schmidhuber) more than any other AI-guy.
Oh, and if you have opportunity to start the codebase from scratch, making one semantic change per version control commit is insanely powerful explainer of how/why your code became as it is. They are like chapters/paragraphs in a story.
Also, separation of concerns: having clear separation of libraries and your "own" code. Including not mixing "business logic" (changing human rules/parameters/requirements) with technical non-human machinery.
Then came emojis, and now the Unicode Consortium's efforts for Unicode version updates seems to be about adding more different kinds of poop emojis and shades of skin colors. Well, maybe it projects accurately the language and culture of this modern time.
UTF-8 is great because it is a superset of ASCII, but because its byte-width varies, it has more complexity for decoding/encoding it (similar to constant/variable width ISA's in CPUs).
Different languages have different concepts, e.g. text direction==flow (left/right, up/down, characters/logograms, different kind of visual cues etc.). Humans create problems when they want to combine different languages at the same time. E.g. mathematical notation is in my opinion 2D graphics, and it cannot be (usually/always) inlined with text glyphs (to be aesthetically pleasing). Same kind of problems may come when trying to inline e.g. languages with different flow directions. Its like trying to combine native GUI widgets in Win32 and Cocoa/SwiftUI and GTK/Qt/WXwidgets - the (visual) languages doesn't have the same concepts or they are conflicting.
I really wonder how all this AI stuff feels/seems to a management person who is not a programmer/low-level-minded person. I remember the time when large portion of average workers could barely understand how to "use computer", let alone the wonder of some simple Excel-VBA/Python magic, and as of now, humanlike magical text prompts. Now after decade(s), some of these persons are in a management position dealing with "AI".
- https://www.apple.com/newsroom/2025/06/apple-introduces-a-de...
But here's the thing with "GUIs/UIs/UXs whatever":
The best UI/UX is created by a domain professional, who knows why and how it serves as the best designed tool for that domain - a tool made by a professional for himself and/or for other professionals in the same domain.
This is why Bloomberg terminal UI/UX is like it is for finance professionals, as are DAWs for music professionals, as are CAD tools for EE/architects etc. They act as the right tool for the right job.
Coders, (figma) designers, and other "implementers" (including management and "product owners"!) has to understand the business domain in order to fully manifest their craftsmanship talent. It is very hard to start and/or iterate UI/UX design if the implementers are not personally using the tool in some professional domain, and therefore know what is right and cool design and what is not.
- Specs (be it for hardware or software) are too complicated. The complexity can be controlled only by large companies (e.g. HTML5 == Chrome == new IE). And software/hardware companies fail to implement specs securely.
- "Self-hosting" is becoming harder. Microsoft & Google controls (blocks) the email traffic for own gains. Public static IP address from ISP costs more than the whole thing (IP, virtual machine/hardware, electricity, etc.) from cloud. Additionally, Cloudflare etc. can take control of internet traffic.
- "Cloud-hosting" has uncontrollable pricing. DDOS attack can make $100k bill in a few hours.
- There only couple of "OKish" non-scamming/no-dark-patterns domain name registrars.
- Google Search is dead (results are only AI-generated stuff + ads + scams). LLMs are the greatest manipulation machine ever invented.
async (() => (e) { try { await doSomething(); while (!done) { ({ done, value } = await reader.read()); } promise .then(goodA, badA) .then(goodB, badB) .catch((err) => { console.error(err); } catch { } finally { using stack = new DisposableStack(); stack.defer(() => console.log("done.")); } });
As it is easier and cheaper to do anything, the result is low-quality products. This of course serves well for the real professionals, who can now sell premium quality products at premium pricing for those who know and want the quality.
If the actual page content is loaded and the cookie is just a modal box slapped on top of it, right click the cookie modal box > open inspector > remove the child element of body that contains the cookie box. If cookie box disabled scrolling, select body tag at the inspector and tick off the overflow and position stylings to default.
- "Firefox uses AI that runs locally on your device to read the linked page and create the key points." --> Last time I played with local LLMs they were gigabyte sized models. So this has to mean Firefox is shipping with a very small model, not gigabyte-size?
- And so does this mean Firefox visits these links beforehand, or when you hover link or what? How many unnecessary requests are made? Since this is "local", I assume you reveal your IP/browser fingerprints to unnecessary sites?
- If the content behind the link is already LLM generated, it gets double LLM generated.
- And what about webservices (shopping carts, management dashboards, banking..), if it is hover based, it might easily and accidentally request url that cause some action, right?
Timedates are hard, and units may require even harder historical/present "political" tracking as how they are defined, and I would never want to maintain this kind of dependency: https://github.com/ryantenney/gnu-units/blob/master/units.da...
And what comes to timedate problems, I try to keep it simple if the project allows: store and operate with UTC timestamps everywhere and only temporarily convert to to local time (DSTs applied, if such) when displaying it in user-facing UI. This functionality/understanding can be locked into own 20-line microlibrary-dependency, which forces its responsible person to understand country's timezone and when e.g. DST changes and where/how/who decides DST changes and what APIs is used to get UTC time (and of course, its dependencies, e.g. NTP, and its dependencies. e.g. unperturbed ground-state hyperfine transition frequency of the caesium-133 atom, which result is then combined with the Pope Gregory XIII's Gregorian calendar, which is a type of solar calendar mixed with religious event for fixing the time, which which is then finally corrected by rewinding/forwarding "the clock" because its too bright or dark for the politicians).