Second hand NUCs cost 70 bucks, sometimes even lower and offer way more bang for buck than a RPi. Use an adafruit breakout for usb -> GPIO and you're good.
Yeh, just load the RPI software image on and follow all the guides and.. oh.. yeh it's slightly different in small ways that could trip you up. Great, I love making myself more work.
For tinkering and GPIO stuff I still think the Pi is unmatched, but with the specs of these newer models and the prices, they seem to be entering more of the homelab/desktop/server segment, at which point these mini PCs we're talking about are a far better value proposition.
If you want to connect breakouts and do some electronics work etc, a Pi 4 or a lower-RAM/lower-cost Pi 5 is an excellent choice, for those who were using them to host software like I've been doing mini-PCs destroy Pis these days at the price/feature and price/performance level.
If the Pis weren't looking more like these mini PCs in features and cost, they wouldn't be coming up as comparisons.
Newsflash, Raspbian is just another Linux distro, which means you can still follow the same tutorials on other distributions, Debian is the closest. Unless you play with drivers, bootloaders, Kernel hacking, or try to integrate with another embedded platform, you wouldn't see a much difference.
More news just in: you and I won't see much/any difference because we've done this (playing with computers) before, many times I expect.
But I appreciate some of you left your Pis in the sock drawer and are unaware of the seemingly obvious (to you) things that learners trip up on. I spent years maintaining a Pi-based audio player aimed at beginners. I even tried to support some other platforms, should have been "simple" but the little differences matter to those that don't know what's little and what's big.
I should probably give some examples but I guess this is a case of what's obvious to me, isn't to you.
Of course I use Raspberry, as I already mentioned - it's great for learning typical embedded stuff. But if the use-case is different than embedded, and more like a server or PC, then just go with PC. I too frequently see people putting too much effort into turning a small SBC into a fully-fledged PC, with a poor effect of course.
That is if the connection isn't downgraded by some other mechanism. Lots of corporate clients downgrade connections to HTTP/1.1 due to how their network is setup.
There is also an issue with SSE I've noticed that websockets doesn't seem to have as much of a prevalent issue with, and that is being blocked by corporate firewalls.
Most websites use TLS, HSTS and the like nowadays. It’s hard for a corporate proxy to downgrade the connection.
Also, the limitation is only applied by browsers, not by (reverse) proxies and other non-browsers clients.
That being said, Mercure allow to subscribe to many "topics" using a single HTTP connection. This helps mitigating the limitation with HTTP/1 (which is very uncommon these days).
a child is not a goddamn pet or a commodity that you can just give away
> In some countries you can even sell your kids, but it is frowned upon in many Western societies
A common country where that happens is Pakistan/Afghanistan where young boys are sold to be prostitutes. I am GLAD it's frowned upon as it should be everywhere in the modern world.
> fostering is the trial run you are looking for
Again, children are not pets, fostering is the same as raising a biological child, if you're doing it any different and treating the child differently, in my opinion, it's tantamount to child abuse.
Frankly this comment lacks any humanity, and hopefully you choose not to have or foster children.
Surprisingly I saw this on the front page mere minutes after deciding to use protobufs in my new project.
Currently I'm not quite sold on RPC since the performance benefits seem to show up on a much larger scale than what I am aiming for, so I'm using a proto schema to define my types and using protoc codegen to generate only JSON marshaling/unmarshaling + types for my golang backed and typescript frontend, with JSON transferred between the two using REST endpoints.
Seems to give me good typesafety along with 0 headache in serializing/deserializing after transport.
One thing I also wanted to do was generate SQL schemas from my proto definitions or SQL migrations but haven't found a tool to do so yet, might end up making one.
Would love to know if any HN folk have ideas/critique regarding this approach.
Seeing almost feral iPad kids has put me off giving my children a portable screen, especially one with unrestricted access to the internet.
The surprising part is I grew up around computers and the internet (got my first email when i was 4) and have always had a desktop PC around, but somehow gravitated towards building websites, small C games, and even though YouTube, Metacafe, Facebook, Miniclip were easily accessible, me and a lot of my peers never got addicted to them.
Maybe it's the way games and apps are designed these days and attempt to hijack your attention, nearly all of them utilising a similar UX pattern (infinite scroll videos, stories, for example) and the effects it has on developing children could very well be magnitudes higher than how it affects adults severely stunting their intellectual growth. If adults are developing attention issues due to such patterns, can't imagine what it must do to children.
I think I'm just going to give them a dumbphone, like a cheap Nokia and computer access at home, but also something else to think about is bullying that is pervasive based on your status and wealth often displayed as the latest iPhone, Playstations, etc and the chance of them being outcasts for not conforming to such structures.
> Maybe it's the way games and apps are designed these days and attempt to hijack your attention, nearly all of them utilising a similar UX pattern (infinite scroll videos, stories, for example)
I'm pretty sure that that's exactly the problem: modern phones and tablets are interactive TVs first and foremost that happen to have the capability to work like computers if you put in a lot of work.
Even if all you did was play games in the pre-smartphone era, those games were not optimized for addiction and microtransactions.
> Maybe it's the way games and apps are designed these days and attempt to hijack your attention
Very much this. I think a social media ban would have most of the benefits of a mobile phone ban.
While you mention FB, you do not say when, and I do not think it was always as addictive.
The other problem with phones is that thy are very much consumption devices. You are not going to build websites, let along games, on one.
> bullying that is pervasive based on your status and wealth often displayed as the latest iPhone, Playstations, etc and the chance of them being outcasts for not conforming to such structures.
The real solution there is to find a different environment for them. Unless its a really toxic environment, its not a big issue. Do you really want your kids to grow up with having to conform to status and wealth displays?
People also tend to overestimate this. lots of people said my kids would be ostracised because we did not have a TV. it was not a problem.
> While you mention FB, you do not say when, and I do not think it was always as addictive.
You're correct, I haven't logged into my account since 2018, but I am talking about the period from 2007-2013, now I am assuming since Meta has Facebook & Instagram & Whatsapp the UX would be more or less the same across platforms like stories/reels or some form of infinite looped community engagement.
> Unless its a really toxic environment, its not a big issue.
I think bullying exists everywhere, it ranges from being explicit, like everybody knows what is going on to very implicit bullying which involves people being iced out from social circles slowly but surely, but I agree with your point, I'll do my best to equip my children with tools to navigate such scenarios, since they exist in adulthood as well. Also these experiences are imo essential for character development ie; being sure about who you are and where you come from without being swept up in the beliefs of your peer group.
I tried writing C games when I was 12 and never got anywhere. I personally feel like I've cultivated quite the overgrowth of dopamine receptors now due to the access of the internet I had as a kid. Curiously, my parents tried to limit access for a long while, and for a long while, the desktop PC only had games and no internet. I do not think giving my child free access to the internet via a desktop would have a similar outcome to yours - not without replicating a whole lot of other environmental factors.
I vividly remember conflicts with my parents over sitting in my room in front of a computer. Potential benefits are never obvious, and certainly not to everyone. "Small C games" seem innocuous in hindsight (specially through the hn lense), but could easily have been center of debate and concern of well meaning parents back then.
You're absolutely right, thankfully my father was in the tech field back then and could very easily discern if I was wasting my time or actually learning/doing something.
Even then I was scolded quite a number of times to go do my homework ^_^
Creating a framework of communicating and channeling children's energy into a path that aids learning in today's attention/dopamine hungry world is something I'm still struggling with.
My neighbours have a way out of this for their 8yo, even though their intention was to just stay in touch as he plays outside unsupervised: smartwatch.
Functions perfectly well as a phone with some additional utilities, but doesn't draw the sort of attention an equivalent phone would if that's what you're going for.
Happily using Windows 11 LTSC IoT edition without copilot and any bloat.
Don't think remapping a keyboard shortcut would be that hard? Also if you're using alt + space for a third party spotlight alternative, wouldn't it automatically override the copilot shortcut?
I recently just started using psst which is a Spotify GUI that's much lighter. When you right click a song and go to show similar tracks u get an array of sliders corresponding to the audio analysis/features like valence, danceability, energy, etc to tweak the recommendations.
It made a light and day difference for music discoverability for me, while the default spotify radio keeps giving me songs i skip instantly multiple times along with songs I've listened to a hundred times, doing this through the API, is 100x better. I've discovered 30 new songs that I love this past week while that number has been steadily dwindling for the past 6 months using Spotify.
Thanks for this suggestion, I’ll try it. I have also been frustrated with Spotify for replaying the same stuff over and over. I tried their new AI playlist feature and prompted it for “tracks with fewer than 10,000 plays”; it correctly named the playlist something with “underground” in the title but failed at delivering on the premise. That’s really what I want - stuff that is similar to what I’m into in terms of genre, but obscure and fresh. There have to be thousands, maybe even tens of thousands, of tracks it could try out on me, but it seems to recycle the same 100 or so tracks over and over again.
I live in Spain, so it suggests Spanish podcasts to me, even though I only listen to podcasts in English or Catalan.
Music the same, I mostly listen to music in either Catalan or English, with a couple of Spanish songs in my lists. But lots of his suggestions are for music in Spanish. Heck, I just see that one of his recommendations is new things in Flamenco, even though it's a musical genre I haven't listened a single song of (nothing against it per se, only that I don't like it).
As I'm using it more to listen to podcasts now I find it hard to listen to music, because if I leave a podcast mid reproduction and play some music I have to remember which podcast it was, search for it, and then I can listen to the rest. Two separate modes, one for music, one for podcasts, would be good. Maybe a mixed one for people who do both mixed.
I will not add on how most of the music it suggests (when not suggesting things that I like) are things I already have in my lists, not good things that I could add to them.
It always seems to be the large tech companies that think they are so clever tying the language to your IP location rather than your language preferences.
Google, Spotify, PayPal and others all seem convinced, no matter what I do, that because my IP address is in Japan that I should be shown Japanese language.
On the web you can for directions at least, select the "Options" button next to the "Leave Now" time selection, it has a distance units to switch between Automatic / Miles / km. To switch the little scale in the bottom right, just click it.
On the iOS app it's "User"->Settings->Distance Units (I assume Android is similar).
(And if you want to adjust Apple Maps in metric, it is a bit more cryptic, it follows the general system settings app, in General->Language & Region->Measurement System, with a choice of Metric / US / UK).
I find it so frustrating when software and web pages don’t default to using the locale info I’ve set in my system settings. Add an app-specific option if you must, but I’d love if these things just worked by default.
The primary issue is that Spotify can only group things into very coarse categories.
If you have a Spanish emo rock band and want similar Spanish emo rock bands then you are mostly out of luck unless there are so many of those that Spotify considers them their own category. Most likely you are going to get just Spanish popular artists if the category is small enough.
I think their embedding vector for musical styles is just way, way too small.
I've started to wonder how much Spotify is gaming the system. For example, by playing tracks lower royalties over tracks with higher ones. There could be some backdoor payola and we would never know.
One would think Spotify’s potential is capped by their expenses (in the hands of the music owners - Universal/Warner/Sony), and the pricing of their competitors, with the deepest pockets in the world (Apple/Alphabet/Amazon).
That chart shows a profit in 2024, and new investors don’t necessarily care about long term history. Tech monopolies have history been so valuable it’s not an unreasonable bet.
A biased algorithm gives them increasing power over time.
Music streaming is not and will not be a monopoly. Spotify has near zero pricing power, with a floor at what their vendors charge them and a ceiling at what their multiple competitors charge. It’s a pretty fungible service.
They have a profit for first three quarters of 2024, and maybe they will continue to eke out a small profit margin, but I’m not seeing what the play is for its current market cap.
It’s based on the potential for significant growth in streaming, and price increases. The major labels and Spotify agree that the price should be increased regularly - after no uplift for a decade - and as more countries begin to adopt streaming the labels and Spotify see significant potential for growth.
The CEO of Warner Music Group - the smallest of the three majors - Robert Kyncl, ex-YouTube exec, said on the company’s most recent earnings call that he believes the penetration of cable TV and SVOD is a good indicator of streaming’s potential, and currently subscription music streaming is lagging behind.
There is explosive growth in some countries that were until recently delivering little: for example something like 95% of Brazil’s recorded music revenue comes from streaming. That’s happened in a pretty short period of time.
There are currently fewer than 1 billion paying streaming subscribers across all platforms globally, but 1 billion is close. It will have taken around 18 years to reach 1 billion paying subscribers; I wouldn’t be surprised if we hit 2 billion in a third of that time. So by 2030 or shortly afterward there may be 2 billion paying subscribers and it’s likely that Spotify will continue to have the lion’s share of those.
It has deep relationships with the major labels and can use high value subscribers in developed territories to subsidise adoption in developing makers.
According to Daniel Ek, Spotify’s CEO, in the US and some major European markets the company has significantly pulled back on marketing spend, but that hasn’t really harmed acquisition - plus a lot of consumers are going from account creation to paid subscription without converting through ad-supported.
Penetration in the biggest markets is still well under 50% of the population - around 30% to 35% - compared to 50% penetration for cable TV and SVOD.
I agree that the valuation seems detached from reality - but the last three quarters have helped build people’s expectations, and the share price has gone from around $170 a year ago to $450+ more recently.
Whether it can sustain that remains to be seen, it’s a crazy multiple, but the relationships Spotify has with the major labels give it a most that makes it hard for new regional competitors to launch, and its overall offering - and marketing clout - makes it hard for established regional competitors to compete effectively.
> Music streaming is not and will not be a monopoly.
People disagree. Spotify more than twice as many subscribers as the next most popular service. That’s not some stable equilibrium.
That “small” profit is already 1b/year and the quarterly profit has been increasing rapidly. Suggesting it’s only possible that that number will maintain exactly the same or less is just an assumption on your part. The market disagrees.
A lot of the profit has come via a price increase and radical downsizing plus big costs cuts. So whether it can be sustained long term will be interesting.
Apple does not need to lose money on anything. They might choose to sell music at 1% profit margin, because they're earning 40% profit margin on iCloud backup and iPhones.
I doubt Spotify investors will be happy with 1% profit margins.
Hasn't this been the main goal of their optimization of recent years, understandably to be honest because the percentage the major labels get of their income is massive, they're giving away what like 70% to labels/distributors?
Dunno how you can be expected to build a business that surfaces the objective best content when your hands are tied giving away that much.
Similarly I have fond memories of the last.fm era, discovering new music via people with similar tastes who I might often end up chatting to. Wasn't it also around this time that Pandora (using the Music Genome project) was also more readily available, seem to remember that was also a good way to discover new music.
But then I'm old enough, pre internet, to also remember the days when people used to send each other mixtapes or you'd go round someone's house with a small selection of vinyl you'd play to each other.
And then there was the more recent (pre lockdown) era of going to see live bands and discovering new bands via the bands that were supporting or via conversations with other people at those gigs, who if you saw them enough times became your mates etc.
Nowadays I can spin up a program to generate a playlist of everyone playing the Great Escape or Glastonbury (assuming they haven't boycotted spotify) and listen to all of them before I even get there and because of the number of acts, given slightly different tastes, unless I make an effort I might not even see my mates....
I'm old enough to remember 4 record stores in my provincial city and they were all selling the same CDs.
I think that secretly most people don't want to discover new music they want to listen to what other people are listening. That's why labels are still in business.
Most people subconsciously don't want to discover new music, they want to listen to songs that remind them of the world and feelings of freedom and discovery and experience they had in their late teens.
They want nostalgia. They don't want cognitive load, they want enjoyment with no strings and no effort.
This is partially a business issue; there’s a lot of money exchanging hands to push specific artists and the major labels are obviously the primary beneficiaries. This affects everything from playlist placement, recommendations, search results, etc.
Spotify has enabled fabulous and powerful music discovery experiences for me. For sure it's not perfect and there are many ways to discover music - I'm really glad that you and others are working on new ideas in this space. But I reckon it will be hard to compete with Spotify's algorithmic playlisting expertise and resources.
It would be nice if Spotify added some deeper playlisting functionality tools to give users a bit more control.
However, for me, algorithms are just tools and the best 'radio' will always come from DJs and other humans making mixes and music programs.
Thanks for the suggestion of trying psst. It is nice to see some innovation!
Can you explain how you use the recommendation feature? What does it mean to enable one of the checkboxes based on a certain song I like? Will it override the 'liveness' or 'danceability' of original song or emphasize that this aspect is particular important to match to the original song?
Two important things that psst needs to improve is reducing clicks when switching songs and showing album/song icons for each song.
Without having looked at the implementation, I am pretty sure the baseline song recommendations remain ie: if I change those sliders while on the recommended tracks page, it'll apply a filter over all the recommended tracks of the song depending on my input.
The results I get are similar in genres/feeling mostly to the root dong, just feeling more in line with my settings.
Again, this is just anecdotal, will have to look at how psst does it to be a 100% sure.
probably the best browser based note taking app for me, full markdown support + persistence.
reply