Looking back on the history of Windows, one of the things that most stands out to me as a user is the utterly insane UX around the C++ redistributable. They really should've figured out a better solution than having every single application ship their own copy along with a separate installer, and then having the installer run always because the developer cannot be sure if it's installed correctly. As I understand it, this might still be a problem with the latest version? If someone like Steve Jobs had been responsible for Windows he would've fought with managers, lawyers, and engineers until the problem was resolved.
This is absolutely baffling that you can't just make an api request and windows deals with installing those redistributables, must have seen that installer pop up close to 80 times in my life. I understand why maybe it was that way pre-internet but really they should have rolled it into windows and made it automatic by now if their dev tools team and windows team can't figure out how to just compile software on their own OS without shipping another bundle beside it.
The newer .msix apps are supposed to be able to handle this, and WinUI3 is supposed to come with a package-dependency that uses the package system to automatically make sure you have the 100+MB library up to date. However, this is the new cool system that nobody uses and doesn't help ordinary devs. They should just ship it in the OS.
And has lots of interesting bugs to explore, depending if the application uses only C++, only C#, a mixture of both, is packaged or expanded, ships the language runtimes along the MSIX, or depends on the store for dependencies,...
Yet another opportunity to submit a few issues on the repos.
> the utterly insane UX around the C++ redistributable
It could be worse—it could be Linux, with no forward compatibility. The VC/C++ redistributable means that applications can be written on modern Windows that targets much older Windows, by simply providing a sidecar installer that brings with it a newer C/C++ standard library. This situation is basically impossible on Linux without messing with Docker.
> means that applications can be written on modern Windows that targets much older Windows
What older windows? Everyone consumer (who are all on forced updates) has to deal with this for the benefit of the small minority on older versions of windows which are probably systems you're not installing new software on anyway.
The part that often gets left unsaid or glossed over is what the transition period looks like. At most we get some Underpants Gnomes claim about unlimited abundance without actually engaging with the substance of what happens if this technology gets built and deployed. What do you imagine the political and economic impact will be if a huge portion of the population is left without jobs and the political reality hasn't caught up to the speed with which the technology gets deployed?
Oh no, but Elon Musk tells us that out of the kindness of his heart we're going to have unlimited abundance. The same man responsible for taking away aid from thousands of the poorest people in the world through DOGE's interruption of PEPFAR and USAID.
With a single sentence from him, he could start saving thousands of lives without impacting his wealth in the slightest. He could do that right now.
Do any of you deliberately integrate the golden ratio into anything you create or do? For me it always seems more like an intellectual curiosity rather than an item in my regular toolkit for design, creative exploration, or problem solving. If I end up with a golden ratio in something I create it's more likely to be by accident or instinct rather than a deliberate choice. I keep thinking I must be missing out.
The closest thing I do related to the golden ratio is using the harmonic armature as a grid for my paintings.
The golden ratio is very mathematically interesting and shows up in many places. Not as prolific as pi or e, but it gets around.
I find the aesthetic arguments for it very overrated, though. A clear case of a guy says a thing, and some other people say it too, and before you know it it's "received wisdom" even though it really isn't particularly true. Many examples of how important the "golden ratio" are are often simply wrong; it's not actually a golden ratio when actually measured, or it's nowhere near as important as presented. You can also squeeze more things into being a "golden ratio" if you are willing to let it be off by, say, 15%. That creates an awfully wide band.
Personally I think it's more a matter of, there is a range of useful and aesthetic ratios, and the "golden ratio" happens to fall in that range, but whether it's the "optimum" just because it's the golden ratio is often more an imposition on the data than something that comes from it.
It definitely does show up in nature, though. There are solid mathematical and engineering reasons why it is the optimal angle for growing leafs and other patterns, for instance. But there are other cases where people "find" it in nature where it clearly isn't there... one of my favorites is the sheer number of diagrams of the Nautilus shell, which allegedly is following the "golden ratio", where the diagram itself disproves the claim by clearly being nowhere near an optimal fit to the shell.
At least by analogy with sound, it doesn’t make sense to me to use the golden ratio. If you consider the tonic, the octave, the major fifth, you have 1:1, 2:1, and 3:2. It seems to me that the earliest ratios in the fibonacci sequence are more aesthetically pleasing, symmetry, 1/3s, etc. but maybe there is something “organically” pleasing about the Fibonacci sequence. But Fibonacci spirals in nature are really just general logarithmic spirals as I understand it. Would be interested to hear counterpoints.
When I'm working out where to place hardware or otherwise proportion a woodworking project, if there isn't an obvious mechanical/physical aspect driving the placement, then I always turn to the Golden Ratio --- annoyingly, I don't get to hear the music or bell ring from
I agree with you. The harmonics/diagonals of the notional rectangle(s) of the piece are more important than any one particular ratio. Phi is no more special than any other self-similar relationship in terms of composition. The root rectangle series offers more than enough for a good layout even without phi.
And yes, for the people who get hung up on what the Old Masters did, it’s mostly armature grids and not the golden ratio!
It can be useful in a "primitive" environment: with the metric or even the imperial system, you need to multiply the length of your measurement unit by a certain factor in order to build the next unit (10x1cm = 1dm for instance).
But if your units follow a golden ratio progression, you just need to "concatenate" 2 consecutive units (2 measuring sticks) in order to find the third. And so on.
This was a long time ago, so we didn't have GPUs or fancy rendering h/ware. We addressed every pixel individually.
So a radar image was painted to the screen, and then the next update was painted on top of that. But that just gives the live radar image ... we wanted moving objects to leave "snail trails".
So what you do for each update is:
* Decrement the existing pixel;
* Update the pixel with the max of the incoming value and the decremented value.
This then leaves stationary targets in place, and anything that's moving leaves a trail behind it so when you look at the screen it's instantly obvious where everything is, and how fast they're moving.
Ideally you'd want to decrement every pixel by one every tenth of a second or so, but that wasn't possible with the h/ware speed we had. So instead we decremented every Nth pixel by D and cycled through the pixels.
But that created stripes, so we needed to access the pixels in a pseudo-random fashion without leaving stripes. The area we were painting was 1024x1024, so what we did was start at the zeroth pixel and step by a prime number size, wrapping around. But what prime number?
We chose a prime close to (2^20)/phi. (Actually we didn't, but that was the starting point for a more complex calculation)
Since phi has no good rational approximation, this didn't leave stripes. It created an evenly spread speckle pattern. The rate of fade was controlled by changing D, and it was very effective.
Worked a treat on our limited hardware (ARM7 on a RiscPC) and easy enough to program directly in ARM assembler.
For most of my music listening needs, I self-host SwingMusic and keep it pinned in Firefox. Occasionally I'll open the music files directly with MPV or VLC.
The automatic lyrics fetching and playback sync in SwingMusic is pretty nice. My only complaint is that it doesn't let me do full-collection shuffle. Ideally it would also allow me to do something like "full collection shuffle but only of songs that I have never heard". Sometimes I'll pick up an album because it seems interesting but things happen and I forget that I added it and it might languish without listening to it for months or years.
I'm waiting a bit for this to mature before I try it out, but I've seen that there's a few ongoing projects to analyze your full music collection to do feature extraction and generate smart playlists using AI tools. I'm not sure if it'll pan out but it seems like a fun tool for exploring large music collections and possibly making unexpected connections.
I think protests are good since it requires you to go outside and interact with other people, it requires a higher level of commitment than the slacktivism of the 2010s that was so prominent in online spaces. Polls are gamed all the time and social media is dominated by bots, but you cannot fake a large crowd in a protest. If a protest is large enough it creates a force that cannot be easily ignored.
Agreed on the slacktivism point. Physical presence means something that bots and polls can't fake. My issue isn't with protesting itself, it's that the assumed impact often seems out of proportion with what's actually being achieved. A crowd showing up doesn't automatically translate to minds changed or policy moved. And crowd sizes can be just as ambiguous as poll numbers when it comes to representing broader sentiment. If the tactics alienate more people than they persuade, visibility alone isn't doing much.
> A crowd showing up doesn't automatically translate to minds changed or policy moved.
Strawman much?
> If the tactics alienate more people than they persuade, visibility alone isn't doing much.
What tactics? What evidence is there that people are being alienated by the peaceful protests, rather than by the murders and other violence and lying of administration officials?
Yes, protests are fertile recruitment grounds. I have inducted many a liberal into leftist thinking after they experience the shocking violence the State is willing to deploy against them for executing what they thought was a guaranteed right.
don't forget the shocking violence leftist have inflicted in autonomous zones, riots, not to mention arson, assault, and in a couple cases, murder. 70 million votes said no and accepted the baggage that came with that no vote.
Of course you can fake a small/large crowd in a protest.
From the top of my head I can think of news reporting both "few (tens of) thousands" vs "hundreds of thousands" (different news reporting different numbers/estimates/etc) in 2025 protests in Serbia/Belgrade, as well as those comparisons of Obama vs Trump inauguration news/photos.
Meanwhile to you as an individual there on the spot - both crowds of say 50K-100K and 1M+ look basically the same = "huge amounts of people in every direction that you look".
Counting large crowds is hard, but the tools continue to improve: we have increasingly advanced drone photography and access to better AI tools to generate more reliable estimates.
If crowd sizes become a significant point of contention it'll become increasingly commonplace for multiple parties to take lots of aerial video and photos that serve as independent verification. You could probably get a pretty accurate estimate of how many people show up to an event by sending drones to take photos every 15 minutes.
In any case, I think the problem you highlight is more focused towards the upper-end, while I was thinking about the lower end of the spectrum. Where some people might be very vocal online, but they're unable to gather more than a dozen or two people for any given protest. If a protest is gathering an unknown number of people that ranges between 100k and 1 million that sounds like a really good problem to have.
Your criticism of inconsistent people estimates are valid, I'm not sure if newspapers have published the set of tools and criteria that they use when generating these estimates, so that's an area where it would be great to see increased transparency.
While 100K itself is indeed impressive - the order of magnitude difference between 100K and 1M makes a lot of room for interpretations, rationalizations, spins ...etc.
The "publishing the set of tools and criteria used to generate estimates" is happening, and so far it seems that usually doesn't matter.
It doesn't matter because of course those sources/news that report wildly wrong (be it larger or smaller numbers) are usually (not always, but very commonly) controlled by the governments.
So despite students that organized the biggest protests in Belgrade giving their estimates (based on combo of RSVP and how many people accommodated people from other cities). And those being close to independant research (using drone footage, VR/AR crowd simulations, AI) with loads of posts/videos providing detailed explanations ...
Most "ordinary people" saw (and keep seeing) just the "official version".
I need to share a video [0] which helped contextualize Alex Honnold for me by contrasting him with another climber I've watched for years: Magnus Midtbo. In this video they're solo climbing a fairly simple and safe mountain, and Magnus is visibly stressed out while Alex calmly shouts encouragement all while recording.
When watching Alex Honnold in Free Solo, I understood there was a exceptional aspect to him, but it took me seeing him climb with other people to really grasp the magnitude.
Reminds me of that time I was taking climbing lessons in the Belgian Ardennes. Helmet on, in harness, hanging in the ropes, holding tight to not fall, we where climbing half way up the mountain, when a person out of nowhere ask if he can pass and just flew up the key section of the route. It was just a local, casual clothes, no harness, no helmet, no rope, maybe not even proper climbing shoes but I can't recall that. Just casually climbing the mountain like he was on a lunch stroll. Even now with years of experience I still don't have that confidence.
I suspect a lot of this is habituation due to repeated practise. As long as one climbs well within one's abilities, the actual level of danger is comparatively low. But the fear is still there and needs to be trained away.
They also did an MRI scan on Honnold and found that he doesn't have the usual fear response. It's not clear if this was trained away, or if it's something innate.
I recall reading about a certain species of birds where, to impress the females, the males dives to the ground. The closer to the ground before they pull out of the dive the more impressive.
The scientist found there was a gene encoding how daring a bird would be, mostly clustered in two groups IIRC. But there was a rare variant which made them much more fearless, causing them to go much lower than the others.
However they only found birds with one copy of that variation. Turned out if a bird inherited the variant from both parents, they never pulled out of the dive and smacked into the ground, killing the bird.
These crazy free solo climbs and similar reminds me of those birds.
This. Watching Honnold makes your palms go clamy and makes you uncomfortable because you imagine how terified you'd be in that position. But for an athlete like Honnold, the experience is more similar to just a "hard hike". Strenuous, but just work. It's just normalized because he does it so damn much. He really seriously is not gonna fall off that building, just like you're not gonna get seriously injured on a class 3 hike.
(Source: I'm also a climber. Not remotely close to Alex's level. But frequent exposure significantly changes how your brain processes these situations)
That’s not true, he has debunked that. It’s due to repeated practice and his confidence in his skill set that he doesn’t feel fear under those conditions
When I go out running a might feel a twinge in Achilles. Or a pain in my knee. Maybe I'm just tired.
When that happens I have to do a mental inventory and ask myself, "Am I better off finishing the run or should I just bag it and take tomorrow off?" Two things; firstly every run hurts a little bit - especially the first mile. I usually get into a groove and sometimes, very rarely, I really have gotten to that place where I'm feeling no pain and a run seems like less effort then a walk - mostly though the nice part of a run has an undeniable unpleasantness bound up with it. I like being able to go out for a run though so I put up with the bad. Second thing. I'm an unreliable source. For all the reasons I just talked about I don't trust my ability to take stock of my physical state. I do occasionally take off or skip for days at a stretch but it's like candy - I don't trust it because I like it so.
Here's the thing. If I get that balance wrong I end up walking in the middle of my run.
I imagine Honnold has to do that same self assessment. If he gets it wrong he plunges to his death. Which - to my mind - is totally crazy. Takes all kinds though.
How do I square "he has debunked that" with the article about his brain fMRI and the results about his amygdala, linked above in this subthread? It's full of direct quotes from both Honnold and the doctors. Where did he debunk it... and how? He's got a more accurate analysis than the fMRI? Do you have a link?
This post just makes me reflect on the sad state of the web and how it continues growing its own little silos that don't integrate with native applications. Actually, I already have the perfect video player (mpv) and I should be able to use that for everything. The dream of the user-agent continues dying: just to show you a few more ads, just for a designer to pad their portfolio with another video player design, just to create redundant work so programmers can keep their six figure jobs.
Note how the author in detailing their 6-year journey only focuses on their customers without any care for the actual end-users that have to engage with these tools in their final form.
This problem needs to be addressed at an operating system level. Websites should be reduced to (paid-for) content providers.
Current situation around ad-blocking is just symptom of the problem. Users try to extract value from the website they are presented with and websites try to hook users into their platform without chance to leave.
All the major newspapers (e.g., WSJ) and magazines (e.g., NatGeo) of the world have already transitioned to online subscription model.
Guess whether their reader base has increased or decreased from their heydays.
Most newspapers or magazines have reduced or stopped their print editions.
Subscription model works only for niche audience, willing to pay for the premium content and premium experience. Rest of the audience will not pay a penny - they are okay to use the site if it is free but with ads, and many users will use some adblocker, but if site refuses to show content if it detects adblocker, they will simply go elsewhere rather than paying for a subscription.
I don’t think your newspaper analogy works very well here.
Newspapers had willing subscribers especially since they had few alternatives. Their subscriber base was not niche at all.
These newspapers were also heavily subsidized by not only advertisements but paid classified ads.
You either caught the local/evening news on TV or subscribed to a small handful of locally delivered papers. Maybe they also deliver major national newspapers like the New York Times and Wall Street Journal.
The willingness to subscribe didn’t go away, the problem is that alternatives quickly arrived that didn’t cost any money.
Free online classifieds ate the newspaper cash cow. People are often still willing to pay for the services that newspapers provided: just see eBay, Autotrader, and Craigslist. The problem is that your local paper can’t match the product that the internet makes available.
>The problem is that your local paper can’t match the product that the internet makes available.
The reverse is true actually.
Print-edition newspaper or magazines had better quality of content than online sites or blogs.
But they eventually deteriorated as the consumer market shifted to online.
Most newspaper subscribers canceled their subscription, but it was not because newspaper couldn't match some quality of content. It was because the way we consumed news changed.
Newspapers and Magazines went from content-heavy editions with ads sprinkled in, to ads-heavy editions with (mostly meaningless) content sprinkled in.
There were 3 reasons for this trend worldwide:
1. Google News proved that news can be consolidated from many sources and provided as easy headlines with short summaries linking to fast-loading Google cached (copy of) pages of actual news sites. So people bookmarked Google News as a favorite (and some intrepid power-users even built feeds that could filter Google News to fetch exact types of content they were interested in), as they found that Google News is timely and topical (whereas in the era of such internet-based news and satellite-TV-channel-based news, the newspaper news was already stale).
2. Google Adsense allowed anya website to show ads for some money in return. Advertisers worldwide realised it was more lucrative to integrate into this internet-driven ecosystem for page-views-based ads that could potentially reach millions of users (and every ad clicked could be tracked, hence ad analytics also came to the for3), than to rely on delayed revenue from ads in printed newspapers or magazines which reached only few thousand subscribers.
(Soon rivals to Google News and Google Adsense came to the market, and they too did their best to woo away consumers and content providers.)
3. Apple's pioneering iPhone and iPad ushered in the new era of smarter snazzier powerful smartphones which were better suited for content consumption.
To keep the advertisers somewhat happy, the newspapers and magazines reduced content and drastically increased advertisements in their print editions.
Online newspaper editions which were free or non-existent earlier, went to subscription model, citing server & infrastructure costs. But the real reason for such gatewalling by the news studios was to prevent the bots of Google, Microsoft, etc., from scraping their researched news and content from their sites.
Seeing the popularity of automated Google News and how it had been "borrowing" their content without paying, the news editors also started "borrowing" news and content from Google News and such automated online feeds.
So eventually, all the news headlines and their content online started looking similar (compounded by smaller news studios going out of business unable to meet rising costs due to reduced subscribers, and those smaller businesses were gobbled up by bigger conglomerates, which simply syndicated the popular news everywhere).
None of my friends or relatives get print newspapers daily. We all cancelled those subscription long ago, because those daily editions were no longer worth even the cheap money they used to cost.
One sad result of such online trends, is that previous famous good-quality and affordable magazines like Reader's Digest, have become scanty content interspersed with too many full page ads. It simply ain't worth it to buy RD anymore. I love reading high-quality magazines like NatGeo, it is just thrilling to hold them in my hands, fluck through their thick pages, see those beautiful/interesting photos and read the deeply researche content. And I used to get these high-quality magazines at low price by buying them from old books shops.
Such drastic shift from print-to-online have seemingly decimated the old-books shops and small-scale lending-libraries, that used to sell/lend old books and magazines at cheap rates. I hardly see such options in my city anymore (they have become few and far between).
Meanwhile, large-scale bookshops and government-funded libraries are doing decent business as parents and teachers are trying to coax kids into reading books and magazines.
As for classifieds, I feel that even that is a dying trend, even online. People no longer want to know or care who died in the community, as they don't even know who lives in the neighbourhood. They no longer buy used products anymore. They prefer it buy it brand new. Because China's mass-manufacturing made goods cheap (at a lower quality). So if someone wants a laptop, they just buy it new, they don't scour the classifieds to see who's selling their laptop at a bargain.
I think even the nice initiatives ike "yard sale", "flea market" or "farmer's market" are on the wane.
I think online time has made humanity to have become more closed and greedy & materialistic, rather than outgoing and ready to embrace new ideas or ready to reuse the old products.
I disagree that online classifieds have diminished in value and usage. Almost all of their functionality has been replaced by something else that is still heavily used.
Facebook Marketplace is huge. eBay is huge. Autotrader is huge. Poshmark is huge. Backmarket and Swappa are big too.
People definitely buy stuff second hand, especially automobiles, which were a huge classified and advertisement pillar of newspapers (local dealership ads took up whole pages with listings of specific cars in their inventory!)
A lot of what you’re saying about content being a race to the bottom is true, there was a paywall catch-22 where you needed site traffic from Google but you needed the paywall to get your subscription revenue.
Your three pillars of revenue were classified ads, advertisements, and subscriptions.
You could only choose one of two for advertisements and subscriptions thanks to the Google aggregator effect (either get clicks with no subscription revenue or get subscribers but no clicks), and classifieds were a lost cause.
Outlets like NYT figured out how to create new revenue pillars, but that doesn’t really work for small regional outlets.
You also had the problem where local newspapers used to re-print national news stories from other outlets but their role in doing so became worthless.
I'm bullish on LLMs being able to help with this kind of reverse engineering effort, if not current models then in a few more years. I've had conversations with people where they managed to get Claude to help reverse engineer old weird binaries with very little input. I wouldn't hype it up as being a magical tool that'll definitely work, but it can't hurt to try.
Interesting model, I've managed to get the 0.6B param model running on my old 1080 and I can generated 200 character chunks safely without going OOM, so I thought that making an audiobook of the Tao Te Ching would be a good test. Unfortunately each snippet varies drastically in quality: sometimes the speaker is clear and coherent, but other times it bursts out laughing or moaning. In a way it feels a bit like magical roulette, never being quite certain of what you're going to get. It does have a bit of charm, when you chain the various snippets together you really don't know what direction it's gonna go.
Using speaker Ryan seems to be the most consistent, I tried speaker Eric and it sounded like someone putting on a fake exaggerated Chinese accent to mock speakers.
If it wasn't for the unpredictable level of emotions from each chunk, I'd say this is easily the highest quality TTS model I've tried.
Have you tried specifying the emotion? There's an option to do so and if it's left empty it wouldn't surprise me if it defaulted to rng instead of bland.
Character Name: Marcus Cole
Voice Profile: A bright, agile male voice with a natural upward lift, delivering lines at a brisk, energetic pace. Pitch leans high with spark, volume projects clearly—near-shouting at peaks—to convey urgency and excitement. Speech flows seamlessly, fluently, each word sharply defined, riding a current of dynamic rhythm.
Background: Longtime broadcast booth announcer for national television, specializing in live interstitials and public engagement spots. His voice bridges segments, rallies action, and keeps momentum alive—from voter drives to entertainment news.
Presence: Late 50s, neatly groomed, dressed in a crisp shirt under studio lights. Moves with practiced ease, eyes locked on the script, energy coiled and ready.
Personality: Energetic, precise, inherently engaging. He doesn’t just read—he propels. Behind the speed is intent: to inform fast, to move people to act. Whether it’s “text VOTE to 5703” or a star-studded tease, he makes it feel immediate, vital.
Yeah, it's not great. I wrote a harness that calculates it as: 3.61s Load Time, 38.78s Gen Time, 18.38s Audio Len, RTF 2.111.
The Tao Te Ching audiobook came in at 62 mins in length and it ran for 102 mins, which gives an RTF of 1.645.
I do get a warning about flash-attn not being installed, which says that it'll slow down inference. I'm not sure if that feature can be supported on the 1080 and I wasn't up for tinkering to try.
An RTF above 1 for just 0.6B parameters suggests the bottleneck isn't the GPU, even on a 1080. The raw compute should be much faster. I'd bet it's mostly CPU overhead or an issue with the serving implementation.
you can install flash attention, et al, but if you're on windows, afaik, you can't use/run/install "triton kernels", which apparently make audio models scream. Whisper complains every time i start it, and it is pretty slow; so i just batch hundreds of audio files on a machine in the corner with a 3060 instead. technically i could batch them on a CPU, too, since i don't particularly care when they finish.
For me the real capability unlock from The Design of Everyday Things was that it made me start noticing and thinking deliberately about design decisions, which pushed me to begin evaluating everything through that lens. In general it comes down to looking at something and asking "what is good / effective and what is bad / annoying about this". If you keep doing that enough you develop your own taste and a greater appreciation of the world. Donald Norman isn't handing you a map, he's teaching you how to build your own.
- Photography, not to take better photos, but to see the world more continuously and deeply.
- Architecture, not to build anything, but to be more aware of urban context, visual design, practical use and emotional response in the built environment.
- Music, not to read a score or play an instrument, but to really hear the structure of music, appreciate melody, pattern, balance, pace, production, individual virtuosity and harmonious combination.
reply