Now that the iPad kid generation is growing up, I fear most don't even know what a file system is. The post desktop world is real.
Abstractions are comfortable, but they might reduce the number of techie teenagers like OP in the future, and that worries me a lot
NAT doesn't get the hate it deserves for stopping people from running their own servers.
I remember being being 18, making my first websites using the LAMP stack and not understanding why my friends were unable to access them from their homes. I knew the basic workings of the IP protocol, but hadn't read about NAT, and the responses online talked about "public and private IP addresses", which made no sense to me -- "Aren't all IP addresses supposed to be public? Isn't that how the protocol works?"
Even after learning about how NAT works, I had no way to work around it, as my ISP blocked the router's web interface and I was unable to do port forwarding.
I would bet that there are at least dozens of millions of people that didn't pursue a career in IT because of NAT. We should ask ourselves if collectively allowing NAT to be what it is was a good decision in the first place.
I’m confused, are you complaining about ISP CGNAT or NAT in general? Did your ISP assign you a public IPv4 or not? If the former then you are lucky, that is a relative luxury.
Even after learning about how NAT works, I had no way to work around it, as my ISP blocked the router's web interface and I was unable to do port forwarding.
Well this is Hacker News, you replace the router with your own.
I have yet to encounter a residential wireline ISP in the US or Europe where using your own router is not a possibility. Even those that implemented 802.1x or similar I have been able to either deal with directly or work around. Not saying there don’t exist counterexamples but I would assume using an alternative router +/- MITM the old one would have been viable.
Of course if the ISP was firewalling (likely) then you have a different problem.
This isn’t really a NAT issue. More of a specific ISP policy issue. (NAT and ipv4 have nothing to do with it, they can just as easily port block inbound 80/443 on ipv6.)
The argument boils down to: My ISP locked down the admin page on my rented router ergo NAT is evil. It’s not terribly sensible.
Did your ISP assign you a public IPv4 or not? If the former then you are lucky, that is a relative luxury.
At least until the 10G fiber upgrade happens, Sonic has given me a static IPv4 address for a decade.
Do they guarantee it is static?
My current IP hasn't changed in years but I need to pony up if I want them to guarantee it forever.
Are you running any servers? I can't find it now, but I remember reading somewhere once that ISPs force IP changes more often on residential users that they detect are running servers.
I've been running a web server for the last year or so, and I still have the same IP from Comcast that I've had for years. When I was a teenager I also ran servers and had a stable IP for years. Ostensibly they say you're not allowed to run servers, but I'm guessing they'd only care if you're constantly using a significant chunk of your bandwidth.
Do they guarantee it is static?
Yes, but it's only for grandfathered DSL.
https://help.sonic.com/hc/en-us/articles/360009880293-Static...
As a data point, the ISP I'm personally using here in Australia (Launtel, highly recommended) has static IPs as an option.
It's not the default choice (most people don't need one), but people that do need a static IP can get one.
Sure. I still remember when I could order additional static IPs on residential cable internet.
That doesn’t mean most getting online today in Asia or Africa basically have a choice other than CGNAT IPv4 and IPv6 from their ISP. Your own dynamically assigned IPv4 is a relative luxury.
This is my take as well, we would not have any ip space if every machine had a public ipv4 address - this is kind of a silly argument to make, nat’s provide a ton of usefulness.
As a kid I didn’t get it either but would port forward, run a vpn server, reverse proxy, and so on. It was a good learning experience trying to get my friends to install vpn clients!
nat’s provide a ton of usefulness
... to companies like IPXO that are trying to acquire, rent, and treat IPv4 addresses like speculative real estate.
we would not have any ip space if every machine had a public ipv4 address
Right, but then ISPs and hosting providers would have to have supported ipv6 10 years ago (or even earlier) when we ran out. Customers were already demanding multiple devices on their wifi 20 years ago, and without NAT, ipv6 would have been necessary.
With NAT, you can't do p2p or self hosting. Without it, you couldn't use centralized services either.
NAT's usefulness is that it helps with IP address shortage. If there wasn't a shortage, NAT wouldn't be useful anymore.
And for those that say "but security!", firewalls are a thing, and need be no more complicated than NAT.
NAT, by virtue of also being a firewall, has actually prevented millions of home machines from being powned. The world is better with NAT. And I wouldn't want any modern IoT (including the IoT I am not aware of in all sort of kitchen appliances) to be on the WAN directly.
Your router comes with a firewall for this very reason. IPv6 is enabled on over 40% the internet connections to Google, surely you don't think all of those people don't have any kind of firewall on their home network?
Then if you have a firewall you have the same port forwarding problems than with a NAT.
Not really? No need to mess with destination ports, no need to visit whatismyipaddress.com, no need to risk exposing your entire home network through SIP ALGs, and your Xbox/PlayStation doesn't tell you that you have a bad "NAT type". Plus, when you forward port 1024-65535 to your Nintendo Switch, as Nintendo says you need to do in its documentation, the rest of your devices don't have their P2P network traffic impeded.
when you forward port 1024-65535 to your Nintendo Switch, as Nintendo says you need to do in its documentation,
Nintendo is a shit company that treats their customer base like bitches yet gets a tremendous steady amount of good will solely due to nostalgia. I'm the last to want to defend the challenges that NAT brings, but blaming it due to some rando corporation's asshattery is silly. Luckily, or not, their documentation is wrong, you almost never have to blanket port forward all UDP ports for the Switch.
The GP point still stands as well. If you have a firewall set to block inbound by default (typical) you have the same basic port-forwarding bootstrapping/configuration problem. Also, the reasons you would absolutely need whatismyipaddress.com or similar apply to both v4 and v6. If you're configuring your router anyway you can get the v4 address.
Without NAT, you can do hole punching. A third server can coordinate two peers sending packets to each other to make their firewalls each see the connection as outgoing and allow traffic. With NAT, the peers don't know which port they're sending on, so they can't relay that info through the third party (though maybe you could have one port scan the other).
Also, an application can use upnp to configure your router/firewall to allow traffic. You could do that with NAT too I suppose if ISPs allowed that, but you'd be contending with other users if you wanted a specific port like 80.
Firewalls on home routers are often of the "shit" level of quality / capability though. :(
Don't worry, the NAT is no better than those firewalls.
Can’t upvote this enough. It breaks the original spirit of the internet and is lazy “security” with a huge negative externality
How exactly does it break the spirit of the Internet? The spirit of the Internet is that we run out of IPv4 addresses and everyone who uses the Internet has to learn IPv6 and be an expert on firewalls?
I can't tell if you're being elitist because your defacto position is that everyone who uses the Internet has to be a cyber security expert, or if you're just being obtuse because a couple of Google searches will usually set you straight on setting up port forwarding through your router.
You don't need to be a firewall expert to use a firewall. Stateful firewalls have been around for longer than I can remember. The default for any consumer firewall is "disable by default", as they should be.
Now with NAT we need to be disable UPnP, and deal with "NAT types" on gaming consoles. Or Nintendo Switches breaking your LAN because Nintendo tells you to put your switch into the DMZ: https://en-americas-support.nintendo.com/app/answers/detail/...
Because of NAT, we now need to deal with firewalls _and_ port forwarding. Plus, because of https://www.armis.com/research/nat-slipstreaming-v2-0/, your IPv4 firewall is practically disabled _because_ of NAT workarounds embedded into your router. You can pick between having a firewall on IPv4 or allowing WebRTC on any of your devices.
I don't understand your first point about firewalls, but port forwarding on a consumer router is no more difficult than opening a port on a firewall. Either one can be overcomplicated by a badly designed interface. See Uniquiti Edgerouter for an example.
I know gaming consoles can be a pain with NAT, but is that NAT's fault? That link you pasted about the Nintendo Switch is literally just an article about setting up port forwarding.
And that NAT slipstreaming issue is just a vulnerability caused by complicated protocols. Saying that a fundamental network technology is bad because its implementation is flawed doesn't make sense. I guess we should throw away x86-64 because Intel Skylake processors had side channel vulnerabilities.
Yes, firewalls are easy, that's my point. Any consumer router will cone with apps strong a firewall as NAT us able to provide, and an even stronger one if ALGs are enabled.
The consoles are a common and obvious downside of using NAT. The Nintendo Switch article is an example of the stupid workarounds vendors will require because of NAT. None of these issues existed if we used IPv4 as it was designed.
The issue behind NAT slipstreaming isn't that the protocols are too complex. They work fine on IPv6 and they worked fine on IPv4 without NAT. The issue is that NAT requires hacky workarounds to do normal networking. The ALG vulnerabilities can be fixed, but fixing them wouldn't be necessary if NAT wasn't such a hack.
Port forwarding doesn’t help when your public facing ip could change at anytime
That isn't NATs fault. That's a dynamically assigned IP address. You can pay for a static, or you can use Dynamic DNS to work around it. If you didn't have NAT, you'd have a static IP by default, but again, you'd be paying for all those extra IP addresses you need.
I don't understand -- you can purchase your own router instead of renting one from your ISP, which usually works out cheaper in the end too, and port forward.
NAT was a necessity when it was being introduced, so I don't understand the complaints. Sure it's annoying, but what was the alternative?
I don't think anyone was dissuaded from IT because of NAT. Heck, if anything, learning how to configure it might have been some people's first step into learning about networking and then getting into it as a career!
And let's not forget that NAT served as a default firewall that did more than any anti-virus to protect PC's from malware.
NAT was a necessity when it was being introduced, so I don't understand the complaints. Sure it's annoying, but what was the alternative?
IPv6
Nope. IPv6 still isn't supported everywhere.
When you're running out of addresses, you need a workable solution at the moment.
Not something that requires all networking equipment worldwide to be replaced with a new standard all at once, with is a practical impossibility.
It doesn't have to be replaced all at once. Most of it already has been replaced.
But it has to be entirely replaced in order to turn off the last NAT, essentially.
Being half-replaced or even mostly-replaced gets you nowhere in terms of NAT removal.
IPv6 would have been great if it was just bigger addresses but the actual system we got suffers heavily from the second-system effect.
This is a terribly uninformed take.
ISPs charge for public IP addresses. Do you want to be charged for a large enough block of public IP addresses to cover every networked device in your house? Because that's what you get when you don't use NAT. And if it weren't for NAT, the Internet would be running out of IPv4 addresses (it already is, sort of), so you'd either deal with much higher prices for IPv4 addresses, or you'd have to learn IPv6, which is WAY harder than dealing with port forwarding through an IPv4 NAT.
Not to mention the premium you'd pay for a router with an actual firewall, and you had better make sure you understand the firewall. Port forwarding on a working router is generally a lot easier than a firewall.
If it was really impossible to set up inbound connections to your server, it was either a problem with the router or a problem with your ISP.
Arguably if people had somehow rejected NAT more violently, IPv6 might’ve been popular much sooner cause you gotta solve the “woops there are now more devices than addresses” problem somehow.
For the sake of running a home server from a network not behind a NAT router, you don’t need to understand much more about IPv6 than how to copy&paste an address.
The fact that my laptop right now has 10 different ipv6 addresses assigned to it when my ISP doesn't even support ipv6, (and two ipv4 addresses) says users need to understand more than copy and paste.
ISPs charge for public IP addresses
Because NAT allows them to.
Do you want to be charged for a large enough block of public IP addresses to cover every networked device in your house?
Yes.
or you'd have to learn IPv6, which is WAY harder than dealing with port forwarding through an IPv4 NAT.
Back in 2005, maybe. Now, it works out of the box, either through SLAAC or DHCPv6.
Not to mention the premium you'd pay for a router with an actual firewall
In 2005, maybe. Today, $10 routers off AliExpress have a firewall, probably set to "deny incoming" by default.
Port forwarding on a working router is generally a lot easier than a firewall.
I know for a fact that the routers of at least two ISPs use the exact same page for port forwarding and the firewall. You don't enter iptables commands, you specify a list of ports that you want to be open for a specific address, except now you don't need to differentiate between "source port" and "destination port".
> ISPs charge for public IP addresses
Because NAT allows them to.
NAT increases the available supply of IPv4 addresses, it makes them cheaper.
> Do you want to be charged for a large enough block of public IP addresses to cover every networked device in your house?
Yes.
You can actually just go out and buy a block of IPv4. https://auctions.ipv4.global/
Tailscale solves this problem as is beginner-friendly, highly recommended.
How would you set this up? You'd still need a box with a public IP to proxy the traffic, no? (Unless you get all your friends to install Tailscale so that they can access your private network).
it/s called exit nodes in Tailscale
I don't think that would work? Exit nodes proxy traffic from the Tailscale network to the internet, not the other way around. Kind of like a consumer VPN.
The other reply mentions Funnel, which I think fits the use case better.
If you want to expose a single web service you can do it via tailscale funnel[1] easily.
Most ISPs I've encountered (various places in the USA) do allow port forwarding or even replacing the ISPs router with one of your own. I'm not sure NAT has been a massive obstacle for most people interested in networking. And without NAT, we would have had to use IPv6 or something even worse. That would probably have turned off a lot of aspiring network engineers too.
In some countries, NAT was implemented at the ISP level and you had no public IP at all. I remember it being a consistent practical barrier back in the 90s for stuff like file transfers. I had a public IP but the other person didn't and had no way of getting one or setting up port forwarding. Back when file transfers were often done with direct connection.
We had Comcast for 17 years, and they always allowed us to bring not only our own router, but our own modem. Even when we rented a combined modem+firewall+AP from them, it was trivial (and documented) to put it into bridge mode.
We just switched to Metronet last month (5gbps symmetrical!), and while we have to pay $10/month to get a public static IP (CGNAT otherwise), they didn't care what was connected to the ONT, which operates below the IP layer.
Most ISPs I've encountered (various places in the USA) do allow port forwarding or even replacing the ISPs router with one of your own.
In the United States my entire life I have always been able to port forward when necessary, though I've only been on two or maybe three providers. I also would help friends in middle and high school port forward so we could connect to their game servers. Many providers seem to try to make it as hard as possible, attempting to lock down the interface behind a web panel. This can typically be bypassed by finding the real panel address locally or by getting a new router that isn't locked down. Growing up it was extremely easy, my provider didn't try to make it difficult, but now there is often a barrier to entry for people who desire to casually tinker with their network setup, which is unfortunate.
Nat provides safer networking for the same masses who can’t understand it
Depending on the year this was it was a pretty easy Google to learn to traverse nat, and use dynamic dns, or do port forwarding, even out to a Linode
If we didn’t have NAT ISPs would provide a router with a firewall that was just as safe.
ISPs do provide routers with firewalls, for this very reason of course. You still need to open ports to actually reach the service you're hosting, only now you don't need to figure out what your actual IP is.
There was a time the modem was commonly separate from the router.
Many ISPs do include a router in their modem with some wifi.
It’s often limited in some capacity in real worth use which lets customers over pay and undercover speed and be open to upgrading to a faster plan.
I had the advantage of being 18 when the internet was still pretty open and I had all the access I needed through my university. I even remember finding a list of root passwords to various machines at Standford lying in one of the computer labs. No idea how it got there. But I tested one with a quick rlogin and it worked fine. Oops. I emailed them to let them know and got basically a <shrug> in reply. Those days certainly were easier for a kid who was learning networking, but they couldn't last with more and more people getting connected and making it more tempting for bad actors to fool around.
I had the advantage of being 18 when the internet was still pretty open
There is no doubt that is an extreme advantage. When I started college in the late 90s an upperclassman was in my dorm for a party and saw a trophy i had from a USFirst robotics competition in HS, he invited me to join an organization he was in that built autonomous robots (we called them "vehicles") for a competition. In that lab, we had network drops and static IPs on the schools inet connection. The lab's admin knew the DNS guys and so we could run whatever we wanted on a fast internet connection with access to our own hostnames. At this time Linux, Apache, PHP and MySQL were just starting to take off. That lab and inet opportunity convinced me to change my major to CS and kickstarted my whole career.
edit: that "lab" was really just an unused room in a random building with some desks and hand tools like screwdrivers and stuff to work on our robots. ..but it had rj45 plugs :)
It wasn’t as hard back then to sift the chaff from the grain. You were either a computer nerd or you were part of the unwashed masses.
If you showed any interest in computing, other geeks would take you with open arms (mostly) versus now where everyone and their dog wants in on the ground floor.
NAT was/is a great for blocking incoming connections by default. A lesson I learnt when I found in early Windows XP days that my files are gone and a .txt note was left on my C: disk...
Remember when connecting an unpatched system to the internet would get it infected by the Blaster worm within the day?
I ‘member.
A non-NAT firewall can have the same default rules.
NAT is an accidental consequence of IPv4 address shortage, but it turned out to a blessing in disguise. NAT gives every home a basic hardware firewall for free.
Everything, for better and worse (mostly worse) is connected to the internet nowadays. Your watch. Your dishwasher. Your TV. Your light switches. These devices are all comically insecure but because they can't accept outside connections because of NAT it's not such a big deal. IPv4 limitations make outbound connectivity easy but incoming connectivity hard. This is what you want 99% of the time.
NAT's firewall effect is purely coincidental. In fact, because people don't run firewalls because of NAT, attacks like NAT slipstreaming can open any port to any internal IP in your network by abusing SIP ALGs and fragmentation.
Wherever NAT is enabled now, a statefull firewall is a better solution. Thankfully, this is the default wherever IPv6 is rolled out to consumers.
Please see this video: https://www.youtube.com/watch?v=GLrfqtf4txw
In order to get some context and knowledge of History.
I've seen this video, and it's an interesting deep dive into the history of NAT, but it brushes aside IPv6 as an "alternative solution" and doesn't pay any more attention to it. Of course, that's not necessarily something to dig into when the topic is NAT and not IPv6, but it's hardly a defence for NAT, merely an explanation on how we got to this point.
Agreed, but does the blame lie in NAT, or the endlessly rainchecked IPv6 transition?
I remember joining an IRC FTP share channel when I was a kid, and wanting to share some of my files to get ratio to download some music and anime they had, and it was my first time attempting to setup an FTP and I thought i was all good to go and dropped my info in the channel and about 3 seconds later a bunch of people were lol'ing at my internal IP.
They were cool though and everybody was helpful, learned a lot that night. IRC was so awesome in the 90s.
my ISP blocked the router's web interface
This is the weirdest thing in your story. Your ISP disallowed you from configuring your router? I've never heard of such a thing, and it certainly sounds ... suspicious. You should be able to get to your own router without even connecting to the internet. Your ISP should never even know when you're connecting to your router. I commonly hit my router's admin page when my internet is down to figure out whether it's my computer or my external connection.
Why would your ISP want to hire so many extra people to nanny all their customer's routers for them? Are you sure this wasn't an excuse from your parents who installed parental controls on it or something?
It is shitty is that your ISP blocked NAT, that is certainly a crime against humanity, as running your own server is such a pure expression of freedom.
Some might find your take bizarre, but the early architects of the internet were against NAT as well. Although, in their idealistic scenario IPv4 would not have held up as long as it did, and probably would have hastened the introduction of IPv6 - not to mention the security concerns like all devices needing competent firewalls. Some people just didn't like that computers could be locked behind a network veil/wall.
When I was a young teenager in the 90s, the thing that blew my mind when I got my first computer was that it turned out that _all computers were servers_. The computer didn't need to be in a rack form factor, or expensive, or any of that. It just needed Winsock, TCP/IP, and that was it! My computer was the same as any other computer on the Internet.
So, while this post is cool, I wish there was less emphasis on the hardware, which is pretty irrelevant for running a Node-base app and serving it to the Internet. You also don't need a static IP—you just need to figure out dynamic DNS :).
I prefer non-server-class machines for purposes like this. Modern single-board computers make very capable home servers, and are far smaller, quieter, cheaper, and less power-hungry.
Servers piss away power at idle because they're not designed for low idle power consumption.
I personally just a regular PC desktop because I need the expansion slots. But SFF PCs are amazing home servers, and have more than enough power for what you actually need.
Also don't forget laptops. They're designed from the group up to be efficient AND have a built in UPS. My home server for the longest time was an old laptop with 32 gigs of ram (back when that was a huge amount).
Laptops make great home servers. I have a Thinkpad (broken display) with i7 8th gen, 16gig ram running quite a few workloads for me 24/7. Due to exorbitan cloud costs, I use it as a sort of staging area for apps from my side projects. Tunnel it to a $4 digital ocean vps and voila you have a serious cloud instance at a fraction of the cost.
Laptops are also not really designed to be always on. I've used some like this in the past, but there are things to watch for:
- batteries might not like being constantly charged. You might kill them, hopefully they don't swell. It might be interesting to remove the battery if you can, but then you indeed lose the feature they provide to survive power cuts.
- the disk drives are constantly parking their heads, make sure that the parking policy is ok (won't wear them to fast). Maybe less of an issue with many laptops coming with SSDs.
- some laptops just heat too much and power off too often, especially in summer.
- their hardware might end up dying quite fast.
I think I would still use an old laptop as a home server if I had suitable one, especially since it helps making such a machine useful for a longer time even when they are not quite usable for the web of today anymore (I like the reuse spirit it has), but I've found my rockpro64 (a nice single board computer) way more reliable in this role.
I guess YMMV depending on the laptop models. I can definitely imagine that a solid laptop with its battery removed can work well as a server.
I notice that macOS optimises charging for this scenario. Keeping the battery under 80% charge if you rarely disconnect it.
I’m surprised that the Linux distributions I use don’t do this automatically but I bet it could be configured.
You can hack that on any system that lets you read the battery status. All you need is a smart plug. You power the laptop from the smart plug. You switch off the plug when the battery is at 80% and you switch it on again when it's at 20%.
This has a major drawback: it's constantly using battery cycles.
Does a computer with charging throttle like described draw power from battery too, or does it use the power from the plug directly?
Good point, and good feature.
If I'm to believe this bug report [1], KDE actually offers throttling charge, but I suspect the Linux kernel lacks support for this kind of stuff for many laptops, maybe also because sometimes the laptop itself does not support this.
Apparently, one should be able to see these files on a supported laptop:
/sys/class/power_supply/BAT0/charge_control_start_threshold
/sys/class/power_supply/BAT0/charge_start_threshold
/sys/class/power_supply/BAT0/charge_stop_threshold
But thanks for reminding me this. I was considering using an original PinePhone as a Bluetooth receiver. Unfortunately, Wifi and Bluetooth require a battery to be present in this phone and I was a bit concerned leaving it plugged in at all times. I'll need to check if the kernel the hardware and support this. Still a bit concerned that the battery will be wearing out fast, I'm afraid Bluetooth will be using the battery power even if the phone is plugged, which is a bummer.
I don't have those on my Dell XPS13 Developer Edition 9370 either.
Would be interesting to know which chipsets support it.
I've used several laptops as file servers in hot apartments without AC and haven't had issues. The trick is to make sure you're not running a bunch of bloat you don't need (which is the root cause of 99% of "my old computer became slow") Slap something like debian on the laptop and don't even bother installing a desktop environment. The load of an NFS server sitting there idle most of the time and a few other system services doing basically nothing is negligible, any laptop should be able to handle it in 40 C weather. If not then you probably have some dust issues, so blow it out and see if the issue resolves itself. Keep the screen partially open (you can turn it off though) to help with ventilation.
I've never had any laptop die while operating like this. I only replace them when I replace my primary use laptop with something new, demoting my last primary use laptop to the server role.
Except for network latency I presume?
Laptops rely on batteries so... no. Thats a fire hazard to have them connected 24/7
I keep a Dell laptop as a server. I used to run stuff on my desktop, but an electricity outage screwed me one time I was away for 2 weeks and the fist night the building admin decided to do some works after midnight (I know because I stopped getting alerts on my phone). That 2 week downtime cost me money (thanks guys!).
So I bought a 2018 Dell laptop with a messed-up screen. Now when there is downtime, the laptop stays on, connectivity goes away for a few mins (which doesn't bother me) and everything is resolved fast.
So, I suggest either your desktop+UPS, or a 5yo laptop (just make sure it's 'high-end' - for its age). I then upgrade RAMs & SSD/M2 from Crucial (I like the convenience of their app) so I end up making decent "server laptops" for under €300.
Yeah! Back during lockdown was running an instance Boinc, folding protein using an allegedly obsolete HP SOHO server. The power consumption was around 300 watts and it kicked out heat like there was no tomorrow. Bearing in mind it was a hot summer, made worse by my attic office/oven I suffered greatly for the cause. In the cold winter months though it was bliss...
Why not just turn it off in the summer? That seems like needless suffering.
It was doing "protein folding at home" for Covid that summer and the work modules would take 7-8 hours so it wasn't an option. Next summer it stayed off!
I have an oldish (2018?) i3 Intel NUC that's absolutely fantastic as a server.
Ditto. Downsized from a larger xenon box into a nuc. Instead of RAID I just have a good backup approach. It doesn't matter if my home server suffers a day of downtime (which it did from a failed SSD).
Especially with the new Raspberry Pi 5 performance. Faster CPU but also much faster data access with M.2 SSD support. Get an 8GB model, add large CPU/GPU heatsink, and add extra heat sinks on the RAM and SSD. It silently run forever.
I'm curious if we'll see 16GB and 32GB RAM options to make it easier to run models like Mixtral 8x7b?
With the need for active cooling, I am no longer convinced Rpi is that great anymore. I loved mine 3b+ but at this point a cheap x86 sff provides much more utility at similar cost and power levels.
The Ryzen 7840 stuff is great but even a 5800 would be fantastic. They have been getting quite cheap lately.
I have a secondhand Dell Poweredge T320, with a single Xeon E5-2428L CPU, its server-class, but in a tower format (the T in T320), so it's the best of both worlds, and the CPU is low power (the L in the name). Cost me 400€ with 24GB RAM and some disks.
It's been running for over 6 years I think. It's only noisy for a couple minutes when starting up, then it's pretty quiet.
Now, dynamic DNS is not something to worry about but CGNAT
Thanks for sharing, I'd never heard of that.
That would be transparent for me as a subscriber, though? I imagine the packets are routed to my cable modem based on some other identified like a MAC?
CGNAT means you don't have an internet routable IP, so you have no way of hosting something anyone can access without some external proxy, such as cloudflare or tailscale
Oh so it’s just like home NAT, and they won’t let me port for ward
It's more complicated than that.
https://tailscale.com/blog/how-nat-traversal-works#concernin...
If I'm reading this right, just the fact you have a static IP does not mean you're not behind a CGNAT. It could be static at the carrier level only, right?
How could I reliably tell if I'm behind a CGNAT or not? If I poll one of those "what is my IP" websites, and the IP my router has assigned from the wan matches the what is my IP then I'm fine, I'm not behind one, right?
And if you can access the services using a dydns kind of thing, then you're safe, you're not behind a CGNAT as well. Only catch where you could be wrong is if you try to access from the same CGNAT network (is this how you call it?) But since the dydns is probably swing your wide internet IP that should still not work?
How could I reliably tell if I'm behind a CGNAT or not? If I poll one of those "what is my IP" websites, and the IP my router has assigned from the wan matches the what is my IP then I'm fine, I'm not behind one, right?
My layman understanding is that the WAN IP on your router will match the "what is my ip" website, but the problem is that it will also match many other people's routers.
Basically imagine your home router and NAT has been moved into the ISP and is serving thousands of end users as well as you... you can't just configure the ISP's CGNAT to forward port 80 to you alone.
I'm not sure if there is a definitive way to know if you are behind CGNAT, however its definitely the case on all mobile networks.
the WAN IP on your router will match the "what is my ip" website, but the problem is that it will also match many other people's routers
That's interesting. So the ISP is directing packets at the various customer routers using some other identifier than the WAN IP the router sees? MAC address or something?
I don't specifically know, but typical NAT can use stateful flow information, i.e. "a UDP packet originated from inside the NAT to external IP/port X, packets from that external IP/port should be routed to the originating NAT/customer" (see https://en.m.wikipedia.org/wiki/UDP_hole_punching)
Some ISPs use MAC address filtering.
HFC (Hybrid Fiber Coaxial) and FttN (Fiber to the Node) are such examples. The rest of them if you don't have a static IP or an internet routable address, use NAT (Network Address Translation) and in some even more exotic environments they use Port Address Translating (PAT), but it's been many years since I've seen PAT in any way at commercial, residential or enterprise setups in ISP space.
But PAT is quite common in the reverse and forward proxy space.
If I poll one of those "what is my IP" websites
You'll get your public IP, it won't signal whether you are behind CGNAT or not.
Check your WAN, if it's in a private range then you're behind CGNAT, likely going to be 10.0.0.0/8 for most ISP's but all the other private ranges count too. Otherwise it will show a public address that matches your "whats my IP" website.
CGNAT should be using 100.64.0.0/10 per RFC 6598.
It's a little different because all the ipv4 tunneling protocols were invented either for or to counter that. NAT64 is usually done at the ISP's gateway and the public IPv4 (at least) is shared with thousands of other customers.
(With such internet connections you are always classified as a bot/spam with captcha services that don't understand IPv6 btw)
As a customer, you cannot influence the IPs and neither the ports that get routed through, so you have to connect to a server that you host somewhere else, just for the purpose of tunneling the traffic. So the server at home is actually a client that reconnects to the reverse proxy as there's no other way to have the open connection.
I saw a big chance for WebRTC data channels to fix exactly that, but I've given up because Google and Mozilla both put a lot of restrictions in the protocol to make it unfeasible for this use case.
And this kinda applies for most cable internet providers in Europe because RIPE ran out of IPs a long time ago. On one hand we got Daimler blocking a whole /8 subnet, and on the other hand we have ISPs that have less than 16 IPv4s. Doesn't make sense to me how this is handled tbh.
Now, dynamic DNS is not something to worry about but CGNAT
Thankfully, there are some pretty cheap VPS options out there. I actually have my homelab behind some, with the nodes running WireGuard to bypass CGNAT.
There's also a few proxy services mentioned in other comments that basically do the same.
Of course, you can also just run things "in the cloud" because the non-enterprise VPS providers come at many different price points, but at the same time it's also just really nice to put your own devices to a good use and have the data be on your own drives (hopefully with adequate backups)
Today, everyone has one computer, always connected to the network. So sad that we do nothing with it.
What do you mean with "nothing"?
I suspect we are relying on "doing nothing with it" for these computers to be usable the way we want them to. More in the following paragraphs.
always connected to the network
If you speak about smartphones, they are usable because they mostly sleep most of the time and thus can last enough without being plugged to be useful. They are always connected to the network, but also in a very specific way: they are mostly waiting for a specific wake up notification from the network, while most of their hardware is actually sleeping. Being actually always fully on might also hurt their battery.
Their connection is also often unreliable. As you move, their will be disconnections. That's good enough to casually browse the web. It's fine if some page does not load, or loads slowly, occasionally.
But I suspect it's not good enough for serving a website: making this randomly fail might be less than ideal.
Many people also light putting their phones in airplane mode, or turning them off during the night.
Now, maybe we could still figure out something useful from this "always" "connected" network, with some sort of paradigm change, maybe with more asynchronicity.
Given the average person's approach to InfoSec (what's wrong with "password" as a password????) this is probably no bad thing
My version of this was the realization that Windows NT Server was not any more suited to being a server, at least not due to any inherent magic.
NT Server was something MSFT could market to enterprise customers for big dollars. It did have meaningful tools for Active Directory and administering a Domain... but if you're a person and not an IT department, it's a safe bet that you'd end up using exactly zero of those tools.
TL;DR: they could have named it Expensive Corporate Seat Licencing Edition.
There was some differences in the registry settings IIRC, the server version simply came with fewer restrictions on the IP stack (i.e. with a bit of regedit hacking, standard NT worked just as well as NT server). There might have been things like needing the server version to run SQL Server as well.
Minor nitpick...Active Directory didn't appear until Windows Server 2000.
Damn, now you've made me want to dig up the story of Germany's first online rail journey planner as written by its creator: it was in the 90's, before Deutsche Bahn had an online presence, but they had the journey information on CD-ROM. The guy took the data from the CD-ROM and created a tool that would return journeys between 2 stations if you queried it. And how does the query work? You'd e-mail his uni address, his "server" (in reality a PC in the corner of his room) would fetch emails every 5 minutes, process the queries and e-mailed back the results.
Sadly Google and DDG are useless now, the last time I wanted to find this story it took a lotttt of time.
Pasted your comment in ChatGPT Plus, it did a search and found the link
https://www.remote.org/frederik/projects/railserver/history....
I think it got relatively lucky with the query, since tried a few variations, you.com and Perplexity and they all failed to find it.
Frederik Ramm went on to found Geofabrik: http://www.geofabrik.de/geofabrik/geofabrik.html
When I was a teenager one of the coolest technology things I got was an old spare desktop. I installed FreeBSD at first and later Linux but it was amazing to have my own Unix machine at home.
It was not powerful nor did it have tons of memory or storage. But it was mine to do with as I pleased. I ran IRC bots and a web server hosting shitty Perl scripts I wrote. It was also the router (via dial-up) for the house once I got a network card for it.
I learned so much messing around on that machine. I credit my ability to exit vim to working on that machine.
One of the defining moments of my life was installing FreeBSD and figuring out how to get X11 working. Everyday after school I would try something new and one day I finally got it working. Loading an X window with xterm was so amazing.
I get the same kind of joy today when I build my own Linux distros. First boot always brings a smile to my face :)
My version of this epiphany was w/ BBSs in the late 80s. I just assumed the BBSs I called were running on "mainframes" or some such. It was shocking when I found out I could run a BBS on my PC.
Ran a Commodore 64 BBS with multiple modems/phone lines, and many floppy disk drives for storage. It wasn't fast, but it was a lot of fun.
<_all computers were servers_
All devices with CPUs can be servers. Your phone, your car...
Modern hard drives / ssds have quite powerful ARM cores. They could, in theory, be used as a server :)
This is even more true now that SSDs have give any old laptop levels of IOPS that would've only been possible with a RAID 5 array of 10,000 RPM SCSI drives 20 years ago.
Put another way, at the height of P2P filesharing, most 16-year-olds did indeed have a server in their room.
When I released my first piece of open source software in 1999, I hosted it on my desktop Linux machine out of my apartment on a cable modem connection over dynamic DNS. I popped it up on Freshmeat[1] and then watched the Apache logs scroll by while eating dinner on the couch. One of the coolest feelings in my career.
Even without the problem of cgnat, as someone pointed out, some ISPs outright block outgoing http traffic from ports 80 and 443, so it really is quite inconvenient for homelabs these days.
Funny, I had this same epiphany!
I'm a 16 year old with a small homelab too :)
I used to have all of my services on cloud but since I got a 1G/1G home network and I found businesses decommissioning hardware and deals on local charity shops which source hardware from the landfill and give profits to charity missions, I decided to give it a go and try administering my own phisical servers. Currently running proxmox on 2 machines with one NAS and 14 spinning disks, with some Minecraft servers, personal programming projects, vulnerability scanners, telegram bots, VPSes for friends, android and MacOS building VMs, storage, some ML school projects with the recently added 1050ti, and hosting the infrastructure for my school CTF competition.
(Italy btw)
Who pays the electric bill? Sounds like it’s in the 50-100USD a month range for those two servers.
He’s a 16 years old doing a very productive activity, I’d expect the parents to happily pay for it, I’d argue that if someone’s son is interested in learn ing and the energy bill is obstacle to wait before to make kids, parents should feel an obligation to fuel their kids passions
As a former 16 year old, I distinctively remember having to negotiate to be allowed to run a computer 24/7 as a server. Part of my argument was calculating the cost to run the server ($10 / month for a 100 watt server at 0.10 / kwh running 24/7, a rough figure I use when referencing electricity costs to this day).
I had to submit an essay to my father, at 16, about why I should be allowed to use a computer more than an hour per day.
The essay won, but my father began talking about me while I was at the computer, saying I was a zombie who wasn't part of the family anymore.
I was learning to code. I built a bunch of websites, and some projects and schemes that earned me fairly serious money in high school. I learned VB6, html, a little perl, and php, on my own, with no mentor, and an active booing.
I'm not sure my dad has ever seen or visited a website I've built (dozens, maybe over 100). I even have bespoke code deployed right now, serving him, that he does not know exists.
I still don't know why he tried to stand in my way all the time. I often think about where I'd be if I had an active supporter, like many kids have.
"I still don't know why he tried to stand in my way all the time"
Maybe he did not try to stand in your way, but was simply worried:
"saying I was a zombie who wasn't part of the family anymore."
Yup I see my son enjoying time on the tablet and teaching himself scratch which is great... But it clearly has an effect on his emotional state and overall quality of life if he spends too much time on it. We try and keep things balanced - he has plenty of time to do more programming in his life!
I still don't know why he tried to stand in my way all the time. I often think about where I'd be if I had an active supporter, like many kids have.
He was probably concerned about your familiar relationships. All relationships have an element of reciprocity. Plus he probably just missed you.
In your situation it sounds like he could have benefited by being a bit more curious and accepting.
IMHO as a parent rejecting the authencity of a child cancels out all the loving things you do because then you are not loving the child. You are loving an imaginary version of the child that only exists in your own head.
IMHO as a parent rejecting the authencity of a child cancels out all the loving things you do because then you are not loving the child. You are loving an imaginary version of the child that only exists in your own head.
sooo many parents are guilty of this. Living through your child is unhealthy
I can't speak for your father of course, but I do think he likely had a valid concern that he may not have effectively communicated that all of this time with the computer was directly at the expense of time spent with the family. He may have felt hurt that you seemed to prefer the company of a machine to his and the rest of your family. He and you may also have very different conceptions of what a successful/happy life might look like. I'm not saying he's right, but from the outside I can imagine why he might not have felt supportive and also why you might feel resentful for the lack of support.
Have you directly asked him why he wasn't supportive? Have you been open to his perspective or just assume that you're right and he's wrong? It's easy for us programmer types to try and simulate the mind states of others to avoid difficult conversations (speaking for myself of course).
with no mentor
I don't think most programmers had any, in the age of 'everyone can code' it might be different but back then, you just read a book
At €60 a month, that is €720 a year.
OP indicated elsewhere here that they’re trying to cover some of the costs with side projects, but €720 anually is a big bill to foot for less affluent parents.
Money well invested considering the guy seems to know more than some IT professionals already.
Eh, power bill is the best one for an increase to appear on, way more wiggle room before anything important is shut off.
Plus if we're talking about being poor it has two important benefits 1) excellent hours occupied to dollar spent ratio 2) engrossing hobbies typically mean they aren't out there rattling around stirring up trouble (which can get expensive in a hurry)
I don’t want to get into whether the ability to afford these things for your children makes you more/less of a good parent…
But if you can afford it, you absolutely should support their passions. The only reason I’m a high salaried software engineer is because my parents generously bought me a computer, allowed me to attend classes (programming in BASIC), etc.
As a former 16-year old, very grateful for my parents for doing exactly this to support my passions :)
60€ a month (tplink smart plug, calculated on the cost per kwh and other taxes). I try to pay most of it myself with money from side projects
Good on you and all but I wanted to note that its weird how persistently people are focusing on the powerbill. It is a drop in the bucket for many households monthly bills and very cheap as hobbies go. I think my mother would have killed for that deal.
60EUR absolutely isn't a drop in the bucket for households in the EU, that can be a monthly bill for someone living in a smaller apartment.
As a 15 years old, I desperately wanted to keep my old desktop running 24/7 to run my own website, learn PHP, linux and do some selfhosting. I wasn't able to run it more than 2 days in a row as my parents would always shut it down to save in power
Because we've all felt the pain of power bills shooting up by 2-3x in a very short period of time, perhaps?
The cost of running a PC 24/7 used to feel insignificant, but not so much any more. I've been running an old PC as a home server for years, but should probably find a more power-efficient NAS box to replace it with.
I'm father of a 10 years old kid. If he wants to run a server at 16, I will give him the electricity bills, a 20U rack, a cisco switch and a bunch of cables. Congratulations to the OP and their family.
When I was a kid in the 90s my dad had raised money to buy a good car. When the time came, he bought a not so good car + a computer for me. I still remember installing windows 3.1 from 40 floppy disks
My memory says that win3.1 came on just 7 floppies. Did you install from 5.25 floppies?
40 is for Win95. Around 20+ for Warp 4.
Maybe 7 feels like 40 when you're 13.
It's quite normal for parents to pay for their children's hobbies.
At least for rich parents :)
Mine is in the 50-100EUR range per year, powered on all the time. Dell Poweredge T320 with Xeon-2428L and an SSD.
Your parents are probably footing a large electricity bill for this running hardware unless there is a revenue stream for your services? Either way, hats off to your folks for supporting your interest.
I know, my smart plug reports around 60€ a month of costs. Most of it comes from the massive amount of hard drives, I try to pay most of it myself with money from side projects. Anyway, I couldn't find rentable dedicated servers with this much storage for less than 50€ a month, the upfront costs of new, energy efficient and powerful hardware is too much, and I truly enjoy running these servers myself.
About the noise, my room has a small mechanical room and that's where the servers are located, the noise can't be heard from the rest of the room.
IMHO it's a good way to get your parents to pay for your hosting costs without having to beg :)
I was in a similar situation regarding storage. I ended up having to upgrade the drives for fewer bigger ones but there are other ways to reduce power consumption like putting your storage server on a timer and turning on WoL so that you can wake it up out of those hours if you need to. Obviously this is not possible if you are serving files to other users 24/7.
You’re not wrong, but as a parent I would be very pleased to pay that bill. Best bargain in education I could ask for.
You could be right — but it's hard to resist the cheap hardware you can find on eBay. I see numerous Hyve Zeus 1U servers on eBay right now for a couple hundred bucks — just add an SSD.
That is so cool! I've also been thinking of adding an old graphics card somewhere to support some ML stuff too - I just have to figure out how :)
I'm also an Italian who taught himself how to code at 16 (I'm now 34), although not even remotely at your level. Then I studied physics, did a PhD, a postdoc about neural networks, etc., but I felt that what mattered to me the most was to inspire young people like the one I was, and I decided to come back to Italy and become a high school teacher starting this year. I'm also organizing an extracurricular python course in my school. If you're not based far from Milan and are interested in talking to my students, we could consider that! Good luck with everything.
Grande!
This is the really cool thing about the internet. When otherwise would these 2 16 year olds have found each other?
They would otherwise have had to go to the same high school.
businesses decommissioning hardware been looking for this for years but never found them in my country (India). Been a 16 year old with a homelab once, now older but no business hardware yet
I was this kid in my day. I had an extremely high-end PC that I built myself (to date myself, it cost well over $6,000[0] and was a home-built 486DX50 (right when they came out) maxed out motherboard RAM (that might have been 16MB but memory escapes me), a Turtle Beach Wave Table Synth sound card, a video card capable of powering a display at 1152x1024 on a 16" CRT display. I ran a multi-node BBS on two 9600, then 16.8K modems.
Had it not been for the experience of "tinkering" with that PC and its predecessor (a ten year old 8088), I would likely have never become a software developer.
I had the honor of building a gaming PC for a neighbor kid and watched over four years as he (unbeknownst to him) became a very competent geek. A few years of tweaking settings in games to eek out a higher frame-rate, fighting with firmware updates, video card drivers, software updates and every other bit of mayhem that the words "I want to game on my PC" evoke, and you pick up a wealth of knowledge "by accident".
I hadn't talked to the kid in four years; met up with him, again, recently and I was blown away. First thing he does is pull a "Flipper Zero" out of his pocket and tells me about all of the crazy things he's done with it.
He is as capable as any systems guy I've ever talked to. He has no clue that he has this skill, either.
It was such a good experience that I opted to buy my son a decent gaming laptop to graduate him from consoles. Over the last year, the same thing has happened to him. This year, my daughter (she's two years younger) begged me to build her a desktop, so she's getting a really nice Christmas followed by a year of learning/frustration. :)
Oh, and the fam got a Flipper Zero this year, too. Can't have the neighbor kid have all the fun.
[0] I saved up the money to afford it by building computers for other people.
Had it not been for the experience of "tinkering" with that PC and its predecessor (a ten year old 8088), I would likely have never become a software developer.
Now that I'm 25 years into my software career, one thing that absolutely blows me away is just how few software engineers in my teams actually own their own computers. It's less than 50% for my current team. Meanwhile, I'm the weirdo with several (gaming PC, gaming laptop PC, and general purpose MBP) in addition to my work laptop.
It was such a good experience that I opted to buy my son a decent gaming laptop to graduate him from consoles. Over the last year, the same thing has happened to him
I went through that experience with my son as well. He's never shown great interest in being taught computer stuff, but did want to help with a PC build. Now, he's tech support for his friends.
The trend of people not even owning a laptop is fascinating. Smartphones are amazing devices, but how do people do that?!
Can't speak for OP, but we have a lot of people who request MacBooks at work and just do their personal activity on it. Saves them the $2k for the device and they don't have to concern themselves with it being lost, damaged or stolen.
This does lead to what can only be described as an acrimonious custody battle whenever we do layoffs. High rates of "lost" equipment returns led to us ransomwaring the things to hold their unsynced personal data hostage until they return the equipment.
High rates of "lost" equipment returns led to us ransomwaring the things to hold their unsynced personal data hostage until they return the equipment.
Is that really worth it to get back what I would assume is on average a 3 year old laptop? Do you then re-issue this old, used gear to your new hires?
That aside, using your work equipment for personal stuff (or vice-versa) is just a really dumb idea.
That aside, using your work equipment for personal stuff (or vice-versa) is just a really dumb idea.
I see people say this all the time, but so far I have never seen any convincing arguments for why I should stop using my work laptop for personal use hah.
If you work for a company that has any kind of audits or certifications or regulation compliance or such (so, basically any company other than a recent tiny startup), your company computer is riddled with spyware that tracks every site you visit, all data in your files and buffers, may have keystroke recording, remote screenshots, and on an on.
Even companies that don't necessarily want this stuff will have it forced on them by this or that auditor soon enough.
And that's just the privacy angle.
If you were to work on any personal projects on that company laptop now you have an intellectual property mess on your hands.
It's less about the hardware itself and more about recovering/accounting for the data on it.
If we can attest that no IP left with them and the device is out of its service life, sometimes we'll let them keep it out of pity. Ironically the most frequent justification is recognizance that they don't own a computer and can't easily look for another job.
Is that really worth it to get back what I would assume is on average a 3 year old laptop? Do you then re-issue this old, used gear to your new hires?
It depends on the device. A 3 year old M1 MBP is virtually identical to an almost brand new M2 MBP. There's no need to chuck an M1 MBP in the trash (metaphorically and literally) just because it's "3 years old".
Now if it's a 3 year old low end Intel machine then I'd be more than happy if it went away.
Yeah, even if it's not official policy, we just basically write-off all laptops given to staff and then give them if there's a layoff.
In the rare case there's still a depreciable value on the books there is an additional severance payment equivalent to the final "value".
High rates of "lost" equipment returns led to us ransomwaring the things to hold their unsynced personal data hostage until they return the equipment.
Sorry, but that's a pretty dick thing to do to someone who just got laid off, especially considering it could otherwise be a write-off.
I actually agree with you, and I feel terrible about it in every case (it sucks being forced to treat aggrieved colleagues as adversaries), but we've gotten shit from clients over the poor data stewardship on this issue. It's not just the employee's vacation photos and recipes on the device.
There are other factors too-- one new challenge terminations in a WFH world present is that after HR notifies them, the employee can simply not log out and still have access to the local filesystem...and external media. Or if they slammed the lid shut in anger, they'd later log back in to Windows using cached credentials in Airplane Mode.
So employees were regaining access and exfiltrating data indiscriminately. It's been much less of a clusterfuck if we just reset the FDE key once HR is done with them, then negotiate return of specific files they identify. (Sometimes we do let them keep the laptop too, after recovering their stuff and wiping the rest.)
It's not like most people are programming or writing Word docs on their phones. The most important thing they're doing is some form of banking. The rest of the time, it's just checking if things exist and sending messages to people. You don't need a laptop to do those things, though I personally prefer having the power of my Macbook.
Now that I'm 25 years into my software career, one thing that absolutely blows me away is just how few software engineers in my teams actually own their own computers.
I'm 15 years into my IT career and I'm amazed at how few of my colleagues seem to care about tech. They know their way around our company's instance of ServiceNow, or whatever their specific role demands of them, but there is very little curiosity, desire & wherewithal to branch out from their little space into something new.
People may have other things in their life that may mean they don't have the luxury of time to spare on these sort of things.
When I had kids all of my hobbies stopped. There is no time left for anything non-essential.
I simply no longer have the time - or frankly mental energy - for curiosity. Right now I am typing this while supervising bath time for a 3 year old.
If my experience is generalizable, it's going to get better when the kid turns 4. By 9 I was able to go back to spending real energy on hobbies, especially if I wake up early enough.
Hang in there!
Note: this seems like it only applies if you're putting your children into a school. If you're homeschooling that "free time" doesn't materialize. It also doesn't happen if you have two children (or more) a couple of years apart, as their school start and end times won't sync up except a few times over the course of > decade.
I had a friend gently remind me that the reason he can be prolific at writing music these days and i "can't seem to find the time" is because i've raised 2 and currently am raising a third, and if he had kids he wouldn't write music either. My youngest turned 7 this year. Older two are out of the house.
A _lot_ of IT these days, at least at software companies, is more akin to retail work (inventory, rotation, shipping/delivery logistics, compliance and permissions management) than tech support. This also makes it an attractive vector for tech/trade colleges because they can directly map such experience to higher-paying, less-seasonal work than retail.
It's still a good foot-in-the-door role for people who still can't access education or job experience reqs for tech, but it's also now one of the roles where you can just park, work 9-5, get a better wage than retail, and not have to corral shopping carts. The tech end of internal IT support has moved more to SRE/ops roles.
Hardware company IT is marginally more interesting but it's still more interested in tagging hardware than understanding it.
I felt that way when I was fresh out of school, now I find that my life is greatly improved by spending less time on (or thinking about) computers outside of work and more outdoors, doing some kind of physical activity.
When I'm in concentration mode for 8+ hours a day in tech, the last thing I want to do is fuck around with a misbehaving home lab. I used to have a server and switches and NAS and firewall etc. But got sick of troubleshooting unnecessary complexity. Now I've got a NAS and a desktop. My tinkering years are over and I just want things to work. That shit loses it's charm over time.
TBF working a full time job has sucked the soul out of me. The last thing I want to be doing on my free time is be on more computers. I used to be a kid eager about technology but now I am actually a boomer. I have a hard time using smart TV's now a days.
This is most people, at most jobs, TBH.
just how few software engineers in my teams actually own their own computers
That sounds fine, if there is no corporate spyware inside.
Meanwhile, I'm the weirdo with several (gaming PC, gaming laptop PC, and general purpose MBP) in addition to my work laptop.
That has more to do with gaming than with software engineering, doesn’t it.
That has more to do with gaming than with software engineering, doesn’t it.
I do use them for personal SW projects as well.
That sounds fine,
Certainly, there's nothing "wrong" with that, but it does lead to specific scenarios, especially in cases of layoffs, where one could be in a position to not have a job or a computer to use for maintaining resumes, taking virtual interviews, etc. When I was laid off in 2022, there were several people that were panicking about that very fact.
You can get a very cheap chromebook for all that if there’s a need. I own my laptop, but have had friends who went through layoffs and just grabbed a cheap one to find a new job.
It doesn’t make sense to own a laptop unless you will regularly use it. Quite a lot of people I know, just don’t really use one outside of their work commitments. So your home hardware sits and collects dust.
I built my first computer when the “turbo” button was still a thing, and have probably built dozens more (for family, friends, and myself) in the ensuing three decades. These days I do the bulk of my non-work computing on an iPad with a Magic Keyboard, only pulling out the laptop when I need Fusion 360. If I get back into gaming I can see picking up a nice laptop, but for me the need or desire to own a desktop has passed.
Pretty much the same. I know in these circles we hate the fact that supermajority of people do their daily stuff on their phones, but it is what it is.
So what, they “get off” using their employer’s laptop as well?
I have a similar story as many on HN do. I had been stringing 10BASE2 coax around my "home lab" for years before I realized people will actually pay me to play with my toys. Of course I then sunk all of that money into more toys for the home lab.
I now have a lot of friends with kids and some of them are teenagers. I've had some concern in my mind for a while now - their total knowledge of tech comes down to asking for a wifi password and watching prank videos (or whatever) on an app. When you ask them what they want to do when they grow up they say "influencer"... The most technical thing they know how to do is reboot "the router" when the internet is down. The second diagnostic step is to yell that the internet is down.
They're great kids and I'm not criticizing them but I've had some concerns that there might not be anyone around to keep the lights on when our generation is gone. In our day using a computer and getting on the internet had huge barriers to entry but that came with the benefit of an explosion in tech talent.
My point is it's great to hear there are still some nerdy kids out there getting under the covers!
They're great kids and I'm not criticizing them but I've had some concerns that there might not be anyone around to keep the lights on when our generation is gone. In our day using a computer and getting on the internet had huge barriers to entry but that came with the benefit of an explosion in tech talent.
I mean, thinking back to the '80s, I was probably one out of maybe 3 kids in my entire class who even had a home computer, let alone knew how to do anything more than play games.
Back to the '90s, many more classmates had computers, but only a tiny, tiny few of us knew how to program and edit CONFIG.SYS and AUTOEXEC.BAT.
Back to the early 2000s, I was the only person I knew with greater than a dial-up Internet connection or who knew the difference between a local and public IP address.
99.5% have always been laypeople, and that's OK. For every 200 kids today who only know how to ask their mom for the WiFi password, there's one who is setting up a home lab, and that's enough for the world's tech needs in twenty years.
Is this not how the generation before us felt about our inability to repair our cars and radios?
That's funny -- when my familiy was planning to move to a home that was being built, my Dad got out there before they put the drywall up and ran telephone and cable television to every bedroom. I followed in behind him with my punch down tool and put a shiny, new Ethernet port in every room, too.
This was 1993. There was no broadband. We almost went 10-BASE2 due to cost. I remember we had a Netware server and two computers. It was kind of overkill but it was fun!
People with the interest and skill to debug technical systems have always been a tiny minority
What cool things did he do with his Flipper Zero?
Mostly exploration and a little mischief -- cloning his garage door opener, turning the smart board off (and on) on the teacher in class (similar to those keychain devices that roll through the off codes on TV IR remotes), trying to break into the NFC tag the Laser Tag equipment uses at the place we play.
I've since sent him some suggestions :)
how old are your kids?
Same exact path here, just a few years later. Tried gaming on my mom's Gateway, which led to building my first PC with a Pentium D, which led to a hobby of constantly buying/selling hardware on ebay and swapping stuff out, which led to an EE degree, and here we are.
No video games, no PC tinkering and overlocking, no EE degree.
This reminds me, I lost my flipper zero and need to find it
When I was 16 (2001), I bought "Teach Yourself C" by Herbert Schildt, a video card and a Sony Trinitron with my summer job savings. I got to the end of chapter 5 where Herb starts talking crazy about multi-dimensional arrays, saw the next chapter was literal witchcraft and gibberish (pointers), got overwhelmed and decided I was on to plan B: become a rock star. Programming was just too hard. I used what I learned to program my TI-82 and cheat on math exams.
Fast forward a couple years, and I'm in my first semester in college. I had some advisor (no recollection whatsoever who it was) recommend that I avoid declaring a major for a semester and just take a few courses in a variety of my interests. What a life saver.
I bought my first Java book and started digging into it. It was a lot more approachable than the C book (sorry Herb!). I attended the first class and realized I'd already read past the end of what the class covered. That started a 15+ year journey (so far) on a path I thought I'd turned away from.
I still play and enjoy the guitar. But I know I'll never be better as a musician than I am as a technologist.
I got stuck in K&R as a high school student around the part where they implement malloc (I persisted through understanding pointers because they were necessary for me to understand linked lists, a data structure I used once in the 1980s and never since).
I learned C in the late 80's from K&R 2nd edition and an Amiga-specific book I no longer remember exactly, back in middle school. I learned sooo much from books back then and just playing around on my own. There's so much more to learn now, but it's also so much easier to find information...
I found the other book: https://www.amazon.com/Amiga-Advanced-Programmers-Abacus-Boo...
my college roommate freshman year (1991) had an amiga and we argued about whether amigas were better than PCs. At the time, his amiga definitely was better than my PC.
IIRC I bought a $25 C compiler, http://www.mixsoftware.com/product/powerc.htm but it only had a stack size of 6 function calls(!)
1991 was basically peak Amiga! Soon after that, 386's with SVGA and Sound Blaster cards started to become pretty cheap. I gave up on the Amiga around 1994, got a 486, and ran Slackware Linux.
Almost the same trajectory for me, except BASIC first then C and K&R on an Amiga. Also switched to PCs in 1994 and installed Linux.
Yes, I was also very into BASIC for a while! First AppleSoft on my Apple IIe, then AmigaBasic. I also played around with some basic compilers before moving to Lattice C (then it became SAS C.)
Yep. In 91 I had a 286 which couldn't do multitasking or interesting graphics. Compared to the amiga, which had multitasking, high quality graphics, and genlock...
By 94 I had a 486/66 with 4MB RAM running TAMU Linux (and Windows 3.1 I guess) basically so I had access to g++ and emacs (having no money for "Turbo C++" on Windows). It was too slow, so I upgraded the RAM (spending close to $1K upgrading to the max, 32MB). Much of my career rests on what I learned about Linux on that machine.
saw the next chapter was literal witchcraft and gibberish (pointers), got overwhelmed
I remember being very confused for a time by C pointers as an undergrad in the late 80s - a situation that none of my teachers really helped with. Eventually I realised that they were just memory addresses, and a whole lot of confusion collapsed into a simple concept. I still remember what that felt like in my head.
Turned out that many of my peers had a similar experience, and not all of them got through it.
So I'm wondering yet again: why do people find C pointers so difficult to grok? Is it conceptual (indirection), notational (& and *, arrays as pointers, etc), or something else?
(Pre-uni I'd devoted a lot of effort to getting VIC and SID chips to do stuff with peek and poke commands, and writing some assembly language, which I reckon did me no harm when I first encountered C - even though I probably didn't immediately recognise the connections. I do wonder what the contemporary equivalent is.)
When I first learned to program at John's Hopkins CTY class back in middle school (~2002), we learned on C++. We didn't wait long to get to pointers because we learned not just how to get the computer to do what we wanted but also about what was going on underneath. When you are in a mindset of thinking about how things are being placed into memory, pointers become a much more natural concept.
So I'm wondering yet again: why do people find C pointers so difficult to grok?
For me, understanding pointers to data types was not the hard bit(int* or some_struct*), but wrapping my head around *array being a pointer to the first part of a contiguous segment of memory took a lot of time. Once that idea(and what ++array_pointer really did) settled in my head, pointers felt fairly normal to deal with.
I suspect a lot is determined by presentation. Explaining arrays using physical dimensions breaks down after the 3rd dimension (and even the “cube of elements” examples tend to be a bit contrived). Expressing them as categories/levels of a nested loop would have been a lot more understandable for even my 16-year-old brain. Pointers weren’t completely inaccessible for me, but the syntax was so alien that nothing felt intuitive anymore after a certain point in the book. Most of the C operators I had learned to that point looked familiar to a lot of the math ones I already knew. Modulo made sense immediately. The Boolean ones were sensible enough. The bit-wise ones were okay once you at least know to convert to base 2 and back. Pointers just didn’t intuit the same way.
It was probably the teaching method. I had an easy time understanding Physics and Trigonometry class but during Algebra it's hard to digest and understand. Then I got older and found that there's a lot of teaching methods and if the teacher is good at adapting the lesson to the students lots of students will understand. Sadly most teachers are not flexible, it's just too hard for them to accommodate all students teaching needs.
The concept of memory address needs to make sense, first. The idea that memory is a big array of typeless bytes, and that all data exists in there somewhere. That an address is just an array index into memory, and you can use it to point to the first byte of an integer or struct or string or whatever.
I mean, it kind of makes sense but it's not obvious.
So I'm wondering yet again: why do people find C pointers so difficult to grok? Is it conceptual (indirection), notational (& and *, arrays as pointers, etc), or something else?
I think people need some background of computer architecture and operating systems before working with pointers, then instead of "what the heck are these addresses?" you get "oh that's how I can use all of these addresses, finally, yay, pointers" moment. If you just learn C/Cpp and related without the background, it's some magical and cumbersome feature.
Do you understand multidimensional arrays and pointers now?
Yeah, it ended up being one of those things you look at a second time and wonder why it seemed so hard the first time. It was a good lesson for me, though. Now when I hit something particularly dense I have a lot more patience with it.
I must have bounced off a half-dozen books for OOP and OO languages that were all like “Cat inherits from Animal…” before finally realizing (while reading OO code, not looking at a book) it’s just goddamn structs with some sugar and automagic for inheritance and auto-passing “self” and such (conceptually, at least) and a method is just a function with a different name for no good reason, et c et c.
If my intro had just been that one Pascal book appendix that describes OO by how it actually works (lookup vtables and such) it would have been so much easier. Instead this easy thing was “hard” for me because every book couldn’t stop introing it with stupid, abstract toy examples.
Yeah, that was the ‘aha’ moment for me, realising that “objects” were just a struct and a bunch of functions that work with that struct.
Pointers were a turning point in triggering my love of learning and improving myself. I remember learning C as a teenager, and all the resources warning me that once it got to pointers it’d get weird and hard.
I was mystified at first, but it was winter break and I decided I’d stick to it until made sense. I don’t remember if it took a few hours or a few days or the whole week, but when it clicked I felt so satisfied, and like I could take on anything from there. It was a really transformative moment.
I remember reading an interview with Hidetaka Miyazaki, the director of the notoriously hard Dark Souls series, where he mentions having a similar experience with playing guitar and learning bar chords - and infusing that notion of satisfaction from overcoming big difficulty humps in his games.
It’s a really powerful feeling.
It's funny. I learned 32bit RISC assembly before I learned C, and so I thought C was terribly hard or maybe tediously hard. Actually more like tediously hard to generate anything but (to my assembler eyes) goddamned terrible output.
Pointers, tho, pointers are just addresses in a register!
Barre chords being an inspiration for Dark Souls really resonates. I never connected that similarity between the satisfaction of learning a game and an instrument, but I love that feeling of overcoming a real challenge so much, it feels like it should have been obvious.
schildt's are widely recognised as being some of the very worst. perhaps if you had chosen better (anyone could have informed you at the time) you might have made a better start.
I wish that were true. As the only teenager in my rural school even remotely interested in programming, the closest thing I had to advice came from my dad, who didn’t finish college and liked to build computers as a hobby, but who didn’t know the first thing about programming. I spent a long time trying to find a good C or C++ book at the time, and the C++ book I ended up buying was so much worse than Schildt as to be unmentionable.
Having started my programming entering programs in raw hex (Z80) and Octal (PDP11), understanding memory indirection was second nature by the time I got to C - so there's something to be said for understanding how a CPU works from the ground up (though arguable whether that's necessary for most programming these days).
I remember crawling Freenode as a teenager with nobody to talk to about programming, and discovering ##C. At the time I had a basic idea of programming languages but was still pretty much a beginner. I started asking why my argument printer in C was printing a (null) and then unrelated stuff (envp :P), and some very helpful person in the channel patiently guided me to an initial understanding of pointers. This was about 16 years ago. I don't remember his name anymore but whoever you are, thanks for helping me out all that time ago.
i asked the highschool computers teacher if I could learn C as part of the program, they said no. So i never took it an an elective. It still strikes me how fkd it is for a professional teacher to say no like that to a budding student... Ended up learning it in college anyways.
I bought my first Java book and started digging into i
Herb also has a Java book ;) And it's fairly well written. Hope you give him another chance. Herbie is a nice fella
Please note that hosting a publicly accessible server in your room at home is specifically prohibited by some employers in heavily regulated industries, and if found, this may result in a large fine (enough to cover the company security re-audit cost) and termination. They even prohibit having a public IP on the router, having IPv6 support, having firmware other than the manufacturer-supplied one, and even having access to the router (i.e., it's the manufacturer and the ISP who should be responsible for security updates, for ensuring that there is absolutely no way for attackers to enter the home network by means of routing, and for the fact that you cannot screw that up).
Having a server for your own use from within the home network is fine.
I've never heard of this in financial services or network infrastructure. It doesn't even sound sensical; if your employer wants a secure LAN, they need to provide a device that treats your home network as a WAN (sort of like a travel router), and your work computer should be configured to only connect to that device. Otherwise you're probably on a network with e.g. TVs, which are definitely malicious.
I never said that this makes sense or that it was enforceable - yet this is what you get when your company is infected by checklist-based compliance people and if an important customer regularly sends these checklists as a condition of continuing the contract.
I am not permitted to name this former employer because this would allow hackers to create targeted phishing emails and social engineering attacks against other people working there.
Compliance with what? I've never heard of this as part of FedRAMP, PCI, SOC2, or FINRA requirements. Another user hasn't heard of it in an ITAR restricted context. I don't see how you could ever pass an audit if employees' personal devices are in-scope. You could never check all of the other boxes you need.
The router public IP requirement in particular is an impossible one. You don't control what address your ISP gives you. No checklist is going to have an item for some random unaffiliated third party. They might be stupid, but not that stupid.
I have never asked where these requirements come from. Possibly, they are just the whims of that customer. To me, they seem to be written as a windy way to require connecting through a mobile hotspot managed by the operator.
Won't someone think of the employers.
I don't know about the prohibited part but that's exactly how LinkedIn got hacked.
See https://darknetdiaries.com/transcript/86/ for details.
Others already expressed surprise on what you wrote and none has confirmed...
But here is some fun fact: when I was working at poultry plant (well, chickens laid eggs) I had a clause within contract that I cannot have my own chickens... I guess because of increased risk to introduce diseases for the plant chickens?!
1) I somehow doubt this is true. 2) If it was, it would not be possible to enforce. People are adding and removing stuff on their home network all the time. What if your roommate's machine gets compromised? 3) Wouldn't it be simpler for them to send out a VPN appliance and treat the home network as untrusted?
Which industries?
Could you give a hint to the employer or the industries you talk about? Surely if an employer wants to very very securely lock down the workspace they would just tell employees to come to office and design that like a fortress, right? I mean I know companies that have metal condensed onto the windows to prohibit mobile reception, but why regulate your home like that?
I work for an employer subject to ITAR and have never heard of this requirement.
Organizations should treat all physical networks, including the ones in their own offices, as untrusted. If they are trusting an internal network these days, they are doing something wrong.
What are you talking about
I wasn't 16 when I discovered used rack mounted servers, but I was pretty young (21 I think?).
I had a miserable job paying near-subsistance wages in 2012 (about $30,000/year, no health insurance, Dallas TX). I needed something a bit more powerful than my laptop to do some experiments with video encoding, but I couldn't afford a "fast" computer at the time, at least not fast enough to do what I wanted. It didn't help that I was particularly bad at saving money until I was about 25.
On a whim, I went on Craigslist and just looked up different ways of phrasing "fast cheap computer", and eventually stumbled upon "server". Upon doing so, I found that you could buy dated, but still useful rack mounted servers for basically nothing; in my case a seller was selling two rack mounted Dell servers, 16 cores each, 32GB of RAM, for about $250. I was just barely able to swing that, so I drove over, picked it up, and more or less defined the direction of my career for the next twelve years.
Rack mount servers have hidden costs. Sitting next to what sounds like a swarm of drones for one. Enterprise pricing for replacement parts is the other.
(Dells have worked out for me ok, but not without having to replace backplanes and other parts immediately.)
Workstations are (to me, magic) computers that use server processors and server RAM in kind of normal PC cases.
Check out HP z440 and z840 systems on eBay (fleabay?) and Amazon.
All my recent additions to my homelab have been Dell Poweredge SFP systems. Businesses use them for a few years and then upgrading, so they're always available. I do make sure to get ones that have a PCIe slot so I can add a 10gbps card.
Counting my blessings, but I never had to replace any parts on them. They used boring SATA drives, which I had a few spares of even at the time.
The bigger issues were the noise and the heat and power consumption. Not only did they take upwards of a kilowatt of power each, but being Texas I also had to take into the account the increased cost of air conditioning.
I was only bitten by that one month where my power bill was like $150 more than usual, and from that point on I was vigilant about turning the servers off between uses.
I stopped considering Dell servers when they started requiring Dell branded hard drives in their newer server models.
Old enterprise hardware was a godsend in college, where airconditioning and power was included in room and board.
And since we were in the old dorms, nobody cared, even when one room was running A/C in the middle of (admittedly mild) winter ...
Sitting next to what sounds like a swarm of drones for one.
When I was in high school I (for reasons I can't really remember) ended up being given a dual P3 rack server. I ran that as my primary machine for 2 years. We called it 'the Hurricane'.
Was there any reason you didn't use two 8-core desktops, for example?
1. Good luck buying 8-core desktops for $125 ea.
2. It's not easy to parallelise video rendering (not just encoding) across two machines.
How about $114 for a workstation?
https://www.ebay.com/sch/i.html?_from=R40&_nkw=z440+workstat... (sorry for the long URL, it was necessary)
https://www.amazon.com/s?k=z440+hp+workstation&s=price-asc-r...
Well they did mention this was in 2012
In that particular time, I couldn't find any kind of computers for cheap that had those parameters. Servers were cheap and easy as long as I wore headphones and was vigilant about turning them off when I wasn't using them.
If you don't want a super noisy server in your room you can rent one from Hetzner for cheap. (Cheapest one currently at 34 euro a month https://www.hetzner.com/sb)
What teenager has €400/year to spent on virtual servers?
One that has €400 a year or more to spend on powering a 120W server 24/7
Assuming it really was drawing 120W, that's about 1000kWh for the year, so €300 at this year's unusually high prices, €150 in a more typical year, and probably half of that will be offset by reduced heating costs.
Parents pay for that, perhaps without realising.
One who can afford to buy the one in the article.
Based on what? They bought a cheap used server off of eBay? That doesn’t disappear if you stop paying a monthly fee?
In May 2022, I finalized the crazy decision to get a real rack-mounted server - I got a great deal on used drives (3x3TB), RAM (128GB of DDR3), and relatively powerful (read: power-hungry) CPUs (2xE5-2690) in a Dell PowerEdge R720XD, which I paid for with some bug bounty money I saved up.
Plus electricity/internet costs, replacing broken drives, etc etc. You can get a lot of rental time out of that.
if all that you want is a "server", you might as well get a vServer for less than 5€
This is a real dedicated server though, it's just sitting in Germany instead of your house.
What $300+ worth difference does it being dedicated make to the 16-year old ?
I don't know, what difference is it to have one at home?
Everyone's doing when I was 16.
When I was 16 I read magazines about computers I couldn't afford, read Pascal code from a book that I wrote out in hand with pencil as I only had 2 hours time with a classroom computer each week, and almost died when getting a graphing calculator that could be programmed.
My dad worked for IBM in the software group but he was able to get decomissioned (usually broken) Thinkpads for me to tinker with.
This is actually how I started my MSP company in the early 2000s. I was selling webspace hosted on Thinkpads with broken displays running Debian Woody (and Sarge) in a ventilated book shelf in my room.
Good times
Woody was my first Debian, I think directly after I switched from Mandrake.
I was...17, I think.
Who's Pascal?
I had a Commodore 64 and could dial long distance for free using hacked Sprint codes. For a brief moment in time, my BBS had the best collection of cracked C64 games in town. But then I slacked for a few months and the game passed me by.
God that takes me back. Spent many hours writing programs on some cheap Casio and then for Christmas one year (15 or 16 years old) I received an HP 48s and it was a whole new world. Good times.
When I was 16 I wrote games on my TI-86 in TI Basic to play during school, and learned VBA to make silly desktop apps (one would open a bunch of internet browsers to a youtube video of mine- a montage of Call of Duty clips that I wanted more views on).
When I was 16 I was coasting to the State Science Fair on a fishing game I'd written two years earlier. Taught myself programming with the GW Basic language book. Totally worth it because I got to miss most of a day of class each for the county and regional fairs, and two for state.
When I was 16, I spent way too much time playing Leagues of Legends. My life would be so much easier now if I started practising what I do now that early!
If you ask your university nicely they may provide rack space for you.
Universities love when students take initiative and build/maintain servers. My university had a student apprentice program where students provided technical expertise to departments as a work study program. And my colleague in the program had a deal with the CS department where they hosted his game server for him. Since he was learning valuable server administration skills and working with other students they approved it.
It’s an easier sell if you ask to put it in the DMZ and if there’s some benefit for other students.
Few 16 year olds are at university?
He makes a comment about graduating and moving to university in the blog post.
I don’t know about Switzerland but that’s at least a few years away?
I'm in the graduating class of 2025, which is closer than it appears.
If you'd like to meet some likeminded people reasonably-nearby, feel invited to the Zurich Nix meetup [1] or the largest Haskell conference [2]!
Interesting, I'll take a look. Thanks!
Then it just shows they've got better planning timescales than most clients I deal with.. (so much last minute bullshit right now with Christmas next week...)
Ugh. Tell me about it.
In years past if you knew someone it might have been easy to get a server plugged in and on the network with a public IP at a university. Maybe in an office or lab somewhere on campus. Nobody would likely know, or care to ask any questions.
Today, they are more and more locked down, everything is run by "enterprise" IT, you need to submit forms and get approvals for every new device with a public IP address, etc. Anything unknown that pops up on the network will be automatically blocked until they know what it is and who is responsible for it.
Something that miss is, unless you need tons of IO (in the form of SAS/SATA storage, or old generation PCIe cards), avoiding these huge, noisy and power hungry servers are a lot simpler than people may think. Mini/Micro form factor OEM PCs on the same price point generally come with much newer generation of hardware (instead of a 3rd gen i7, you might get a 8th gen i5) and overall performance is so much better. It’s also so much easier to host and maintain (just plug it near your ISP modem, and forget about it).
Absolutely. I think that's the practical solution I'll do when I move out. Until then, it's definitely not nearly as fun though!!
When you do head to uni, maybe see if there's a nearby place which can co-locate your servers?
Stuff like this, though "near the uni" would make it easier to access if/when the hardware needs a bit of adjustment. :)
* https://www.hetzner.com/colocation
Note that co-location pricing can be all over the place (from low to incredibly high), so don't get discouraged too early.
This website might be useful too, as it seems to show data centers in various countries.
* https://www.datacentermap.com/switzerland/
Might be useful for putting together a list of places to investigate. :)
I was looking at running OpnSense for my router and almost used my old Skylake computer. Instead I bought a $100 fanless Intel N100 micro computer that sips 7w and has about the same performance. The electricity savings will pay for itself. And the i5-6600k wasn't even that power hungry.
You might be surprised about the old Skylake. If it's using a solid state drive and any unnecessary fans/peripherals are disabled, it's probably only going to use 10w or so at near idle (as a firewall would be with that chip.)
Even if it uses 17W on average, the 10W difference would be $14/year at $.16/kWh electricity. Unless your electricity is significantly more expensive, it probably will never pay for itself over its useful life.
But if you _do_ need a whole bunch of SATA disks, it's hard to beat a bunch of R720XDs.
I have a small fleet of them and they're actually not that noisy or power-hungry if you 1) disable cpus 2) undervolt remaining cpus 3) write your own fan controller off of raw ipmi byte sequences
By far one of the best cost/pb mass storage solutions. I have a bunch of NUCs and SBCs too, but they're definitely suited for different purposes.
One of these days I'll buy a surplus Isilon cluster or two and truly achieve bulk storage apotheosis. Cbf fighting Ceph in my spare time as well.
I've got an old Skylake HP elite SFF machine. Without modification it can hold two 3.5in drives and a single 2.5in drive. Also with great cooling design (cowling to direct heat from the cou straight out of the back) and an efficient PSU (80 plus platinum), it makes a great value always-on machine.
Had an Intel Motherboard + Atom CPU combo as a File Server once, but eventually had to upgrade it because Atom chips really suck. Eventually want to a Skylake Pentium, which performs much better.
and forget about it
This is exactly what I didn't want when I was 16. I did have a server too, but in the HP ML form factor (less noise than the pizza boxes). Also, ECC is kind of important, and a real HW raid is a nice learning expierience.
I love seeing the server on an Ikea table next to a paragraph about how expensive racks are. Did you know about Lack Rack? https://web.archive.org/web/20230131031301/https://wiki.eth0...
I do! I just haven't had the time to measure everything and find the right screws and table.
Careful, once you go down that IKEA hacking rabbit hole, it can be hard to stop buying cheap furniture.
:)
It is cool but Ikea used to be cheap but now they have significantly lowered the product quality and also increased the prices on top of that, atleast in Australia.
Yes the cheap stuff isn't good anymore and the good stuff isn't cheap
So that's what it's called! Thanks!
Key point from that link; IKEA's swapped the materials round to use a lot more of that honeycombed hollow stuff, which makes it less than ideal for drilling or holding heavy equipment.
This is annoyingly common in a lot of furniture, including desks. Anyone who's thinking of drilling their desk to put in a desk-clamp monitor mount: make sure the desk is solid, first!
Static IPv4 is very important for getting into networking. Out of my friends, one had a free semi-static IPv4 (rotated once a year iirc), now he has the best paying job in devops. Another one had a cheap surcharge to get it, also turned into a job after few years. My other friend had (and still has) a cg-Nat, so never got into any deep IT stuff, other then PC building.
You are very knowledgeable (for your coding and bug bounties) and very lucky (for the static ip, as well as understanding parents lol)!
100%. When I was a similar age to OP and getting into web development it was the coolest thing in the world that I could map a port on my router and type in my public IP address at a friend's house and the website that I wrote would appear.
If I did the same thing now there's a good chance I'm behind CG-NAT and it won't work at all, and a 100% chance that my public IP won't be the same in 2 months.
We've really broken the internet. Between this and the average kid using a locked-down tablet the barrier to entry is higher than ever.
I dunno, I don't really feel like having to use a service like dyndns or one of the dozens of clones is that big of a hurdle, is it? I recall using one of these services when I was in college (off-campus apartment with cable internet) back in the early 2000s.
For CG-NAT you don't actually get any inbound ports, so dynamic DNS won't help you.
That being said, I know some ISPs over here that will block standard ports inbound beyond SMTP when you're not behind CG-NAT. I can imagine that leading to some who are just starting giving up when they've followed all the instructions to get HTTP on port 80 working and it still doesn't work without explanation.
OK, fair point. Forgot about CG-NAT, but dyndns still meant I was able to access stuff remotely despite not having a static IP.
Yep, it's not a problem if you know how it works and don't have CG-NAT.
My point is that newbies don't know, and whilst the IP rotation problem might force some of them to learn about DNS, it will also put a bunch of people off entirely.
My ISP is supposed to be moving users whose contract rollover to CGNAT.
This is why IPv6 needs to be available everywhere, including on mobile.
My ISP, in New Hampshire, Fidium, does not list IPv6 anywhere on their website. I asked the tech installing it. He told me no one uses it.
So, he cleared that up I guess.
”One day, I’m going to graduate and move out for university”
Seems to me like University would be wasted time. You’re well on your way. If you can configure reverse proxies on nginx, host node apps, route ssh, and keep uptime - there’s a job for you today…
Do university on the side if you really want the paper but you already possess the skills to succeed. The insatiable desire to learn.
Best of luck my friend.
Do university on the side if you really want the paper.
There are other good reasons for uni, like social and networking opportunities (doing a startup with like-minded students, maybe exploring the uni party life, learning how research is done...).
None of the reasons are worth going into debt for though. You do indeed already have the skills needed to get a good job in the field. Uni is an option for you but definitely not a requirement.
Feel free to go for a study field which is tangential to your interests (e.g. UX design, business...). Maybe you won't be doing any of that for a living but it's one of the best opportunities you get to learn how people you will end up working with talk and think. You can certainly learn the stuff you major in on your own.
Going to Uni for networking and social reasons (as long as you aren’t going into debt) is a great reason to go.
I was simply suggesting that one already has the skills to go work and make a pretty handsome salary doing what they are doing while learning what they want to learn.
It’s a whole lot easier to fund a woodworking shop when you have money. It’s a whole lot easier to experiment with 3D printing when you have money. You see what I’m getting at. It’s not the ends, it’s the means.
You can learn the easy stuff by tinkering. It's fun! Many people are smart enough to teach themselves anything. But actually sitting down and doing the work is the hard part. How many people are going to struggle through "Algorithms" by Cormen without the external pressure of coursework and exams? Almost nobody. And that's the problem. It's not that people can't learn by self-study, it's that nobody wants to.
There is no alternative for learning to the structured environment provided by a good university.
I like that kind of stuff and will gladly sit through it again and again until it’s ingrained in my brain. Some people love to learn. Teach me something or lose my attention.
You are right though that there’s a huge difference between doing something as a hobby vs as a job. If you don’t have a passion for it, it will feel like every other job. No matter what you are doing. However, hard work and discipline are good traits to start early so I’m still in the camp of “Go and work and make money instead of go into debt to learn”.
” There is no alternative for learning to the structured environment provided by a good university.”
We can agree to disagree. I found the R&D lab of some of the companies I worked for far surpassed what my peers were doing in their labs.
It's too soon to lock into a career path like this. Insatiable desire to learn is something that can lead to success in a great many fields other than the one that helped you set up your Minecraft server. OP can use college to find the field where society needs them most and there are the most opportunities. As "Server technician 12", they will make fast money but they won't be discovered or mentored appropriate to their talent.
Good for you! I get a feeling that there are fewer and fewer individuals learning the basics (or sometimes not-so-basic) of hardware and operating systems. It is invaluable knowledge, even when working with PaaS. For instance, having experience with IIS since NT4 and Apache since the 1.x days is invaluable in diagnosing Azure Web Sites.
Even people actively in the industry don’t know. About 10 years ago I was working in a datacenter and one of the severs needed some hardware replaced (new ram or something). The guy who ran the team that used the server (who sat maybe 50 yards from the server room for many years), asked if he could watch me do the replacement. He had never seen the inside of a server before and was curious what it was like. This blew my mind. I didn’t get how he could be so close to it for so long and never see it. He made it sound like he never even saw the inside of a home PC as well.
At the time it wasn’t like there was a lot of protocols in place. He could have walked up to one of a dozen people and asked for a tour on any given day and gotten it, no questions asked. Not to mention random servers sitting around the cube farm for whatever reason. He just never thought about it before that day.
I provision servers for a living and usually the only times they get physically touched are when we rack them and the 2nd and last time is when we unrack them. Occasionally, one will have a hardware replacement. But for the most part a server runs continuously for 3-5 years. Even a memory upgrade involves unbundling so much cable (power, network, storage, KVM) that it's easier to rack a new one with more memory.
I'm not surprised your team leader, if of a certain age, hasn't seen the inside of a modern server. We push our datacenters further and further out into the countryside. My annual trip to our DC was a 4 hour drive and even then it was just to do annual inventory.
In a modern datacenter I totally get it. I’ve visited one of our current data centers and had to get a lot of clearance, and go through several layers of security. I couldn’t even drive in the parking lot without my name being on the list and proof of my identity. Most people will never get that chance; I have a lot of co-workers who have asked and were all denied.
However, at the time, in that office/datacenter, the barrier to entry for him would have just been asking and 2 minutes of his time (or however long he wanted to look around). The computer rooms were on the way to the bathroom. That place was all pets, no cattle, so we were constantly doing whatever it took to keep the hardware running. There was no shortage of opportunities.
Most dont even have a PC that they assemble themselves. They have very little idea of the internals. Nor do they follow and read enough about hardware news and articles. These PC could be used as Server. While not professional rack based ones, but you will still need config them.
A lot of these basic understanding and fundamental learning experience is gone. We are not far from Front End developers that dont understanding anything about HTML and CSS but only React.
We hear this every day on this website and it's just not true. Making computers accessible to people makes it more likely that they'll get into hardware enthusiasm and assembling a desktop from parts.
Fun fact.
When I was a 16-year-old, it would have been significantly easier for me to get a $200 home server than a $5 Digital Ocean VPS.
The online shopping service over here (Allegro, basically Amazon and eBay all in one) is perfectly happy to do pay-on-delivery for a small fee. You probably need to be 18 to sign up, but no Polish person ever actually cared about such restrictions and there's no age check.
On the other hand, cloud services require you to have a debit or credit card, which you can't get without an intervention from your parents if at all.
You can say lots of bad things about Bitcoin and crypto in general, but the youth rights angle can't be overemphasized enough.
You probably need to be 18 to sign up, but no Polish person ever actually cared about such restrictions and there's no age check.
You need to be 13.
On the other hand, cloud services require you to have a debit or credit card, which you can't get without an intervention from your parents if at all.
With parental permission, a 13-year-old can open a free bank account with a debit card that works online. It's not a new thing, I had one 15 years ago.
With parental permission
Is the constraint the parent comment is referring to.
With parental permission
Back in the 90s I was a minor and doing similar things, I don't think I could have explained what I was doing to my parents. Even if I had the funds, they wouldn't believe it wasn't some scam or an illegitimate thing. I think teenagers who are tinkering don't always have buy-in or understating from their parents, so getting that kind of permission can be tough.
I think I didn't have a debit card until 18. Then I got old Unix boxes on eBay.
You can say lots of bad things about Bitcoin and crypto in general, but the youth rights angle can't be overemphasized enough.
You're not in favor of parents controlling what their kids own?
I'm in favor of kids being able to do on the internet what they were already able to do in meatspace.
In the 80's, a kid of basically any age could go to see or rent a movie, buy a book, make purchases at a store, play an arcade game with their friends or throw some cash into a payphone and call anybody they liked, all without their parent's knowledge or permission. There were very specific things that have been decided as "age inappropriate" by a democratic process of law (although one that didn't include the kids itself), and those very specific things (like alcohol consumption or X-rated movies) were forbidden, everything else was fair game.
In the modern age, most of those things are done online, and most platforms require you to sign a contract to participate, which requires you to either be 13, 16 or 18, depending on platform and jurisdiction. Even if you get past that by lying about your age, you still need a payment method, which you can't easily get without explicit parental permission. There are exceptions, some platforms support gift cards that you can get over-the-counter, without any form of ID, but the point still stands.
Those Dell R720s are a pretty sweet spot. This past summer I got a couple to replace some old dev/stg hardware and picked one up for my home lab: $305 landed with 256GB RAM and dual CPUs, 24 cores, and dual power supplies. It normally runs quietly, though can ramp up if under heavy use.
I added 8 1TB SSDs, and another 8 2TB spinning 2.5" drives. I had to buy a new RAID controller because the stock one wouldn't boot with all the drives in it (they are unsupported drives), and a 10gig NIC. All in for around $450.
how much Watt does it pull?
A dual socket supermicro with haswell/broadwell with only one sata ssd idles around 80 waitt. Full load up to 500 watt.
One that we have in production over the last day has been averaging 197 watts over the last week. So that's probably the upper limit of what I'm going to see on mine for my home lab use. I'll have lower use, and I have 8 7.2K discs where that has 14 10K. I don't have my home machine DRAC set up right now.
I had an HP proliant with 4 xeons, 40 total cores (80t), 192GB of ram, and 8 spindles. The 4 (four) power supplies ran about 600W idle and the BTU graph went to 1500W equivalent. I ran it on 240VAC on 2 supplies and it was generally around 1kw/h for my usage.
It was electrically equivalent to a bitcoin miner.
Nope, my son has them (5 in total) stashed under the stairs where they still manage to generate enough heat & noise to make me question the placement. Still, pretty chuffed that he's managed to do it all himself.
Nope what? What do you disagree with?
Actually looks like a 'nope' of agreement.
- Most 16 year olds don't have servers in their room
- No indeed, some have them in the cupboard under the stairs
Of course, some 16 year-olds, like Harry Potter, have their room in the cupboard under the stairs, so the logic doesn't follow implicitly.
It follows, that's just a third category which agrees and can be prefixed with 'nope'
Probably referring to the title.
"Nope most 16yr olds do not in fact have servers in their room"
Or at least that's how I read it.
Congrats! At 16, I only had a 486DX2-66 with a 28.8K modem and a UPS.
For other teenagers and teenagers-in-spirit to accomplish this feat:
- Search for used enterprise servers on eBay. HP, Lenovo, Dell, and Supermicro are the typical brands to choose from.
- Before purchase, find the specification docs, drivers, and firmware on the specific model to be sure it's something that's complete and usable, or can be completed for reasonable $$.
- If you plan on running XenServer or VMware ESXi to then run multiple virtual machines on it, make sure the hardware is compatible by checking the OS hardware compatibility list (HCL) before buying it.
- Avoid 1U servers because they're louder than 2U+ designs due to having to use shorter, smaller fans that spin very fast.
- Make sure the CPU is at least as fast as a computer you own, or it might be a very expensive doorstop or an oversized "Raspberry PI": https://www.cpubenchmark.net
- There is a gotcha with Dell, Lenovo, and HP servers using AMD CPUs where they are vendor locked. The plus side is sometimes sellers offer locked CPUs cheap enough that it makes sense.
- ECC RAM. Friends don't let friends drive non-ECC RAM. https://cr.yp.to/hardware/ecc.html
- SSD. Because spinning rust is for storage, not the OS. https://web.archive.org/web/20110831080738/http://buyafuckin...
- Try to get something with lights-out remote management (iLO or iDRAC) if it's already licensed.
- When received, it's probably going to be dirty and have some scratches. Open it up and clean it with compressed air outside.
- Bonus points: Convince the parental unit(s) to put a 19" patch panel and half rack in the garage. :]
All of this advice is merely nice-to-have. You can easily just use an old desktop or laptop to do the same.
I've homelabbed for a long time and have rackmounted systems full of old desktop parts, rather than buying or seeking out used server hardware. It just isn't required.
ECC RAM? Nah. It's useful if you can find it cheap and your CPU supports it but it's overstated as a problem IME.
I do agree you should buy 2U or even 3/4U equipment, if only so you can mount much larger (and thus quieter) fans.
My file server is a 2011 mac mini, I attached two HDDs (5TB and 8TB) then installed services like Jellyfin and Gitea.
Thank you! Those are all really good tips. I heard stuff similar to some of them getting started myself, and I wish I heard the rest too. :)
This is cool, but honestly... Go learn how to paint or play an instrument.
join a band and create art.
You will have a ton of time to learn to code and build servers.
I was one of these early bloomers for being a programmer, and you know where it got me?
The same fucking place as all of the other people who started in their mid 20s. Its a waste of time.
Well... I make hundreds of thousands of dollars and have a stable career and I mostly like my job... BUT! My point stands.
There are absolutely diminishing returns for being a "good" programmer, and then it opens up to social stuff. Being able to impress someone with your paintings. Being able to both implement and design a page. Being able to jam with your co-workers after work every once in a while.
If anything focus on school and drill math as hard as you can, it is a HUGE hill to climb later in life.
Yikes, this is a pretty judgmental post. I spent my time as a kid joining bands and being in plays, but if someone loves building homelabs, what’s wrong with that?
To your point about diminishing returns in technical skills vs social skills, as a senior engineer, I completely agree, but I have the opposite takeaway! I love to build at home and expand my knowledge of pure computing, because it’s a love of mine that I don’t actually get to do much of at work.
Fair, edited to make it less judgemental. Honestly I don't care what people do with their time, anything is better than reality TV.
I just deeply regret my mis-spent youth expecting a pot of gold at the end of a rainbow that was not there
What an absolutely horrible take.
Here's a kid who's found something he's passionate about and loves what he's doing and all you can do is say "get a life".
We rarely get to pick what we're passionate about. We go through life and, if we're lucky, we stumble upon something that piques our interest. We don't pick when that happens either. You could find your passion when you're 7 or when you're 70, or anywhere in between. It could be technology, cooking, baking, art, music, gardening, exercise, a sport, or any other thing on an endless list.
People who are passionate about something, anything, are the most interesting people to talk to.
Great work! What’s your social media platform? Curious about your tech stack — and what content/interest/community you’re serving.
Thank you! The social media platform is a Twitter-like website for my friends (roughly 3000 of them registered, but smaller set of daily active users) that I started building in 2020. I made it to learn new technologies, so I've rewritten parts of it from scratch a few times. It's got a a SvelteKit/NuxtJS 2 frontend (in the middle of a long rewrite) with a seperate ExpressJS API (which others have used to make their own clients, like an Android app!) that's spun off of an older iteration of the site where the frontend was coupled as templated HTML.
I used MongoDB from the start, but I'm beginning to regret that - It's getting hard to maintain relational consistency and ensure all the data going in the DB matches the same shape as the site scales. Maybe it's time to migrate to Postgres.
I mix up a few things from the rest of the world for my site - comments are nested like Reddit/HN, and posts/comments are rich text limited HTML like a blog. There's a really minimal algorithm that's relatively hard to mess with: chronological feed from people you follow, and a seperate Explore page with top users and trending posts (based on the recent likes)
Profiles have seperate 'walls' from posts so they basically have little comment sections for people to use.
The new site also has profile customization so people can change the accent colour that their posts and profiles appear as in other people's feeds, and there's little widgets to customize profile 'Sidebars'.
It's nice to have a little quiet place seperate from the rest of the world, with its own ecosystem of clients, tools, bots, etc. (actually this year somebody made a Spotify Wrapped style recap website using the API!).
I have to say, this would be an impressive set of achievements for someone much older - great work and glad you're obviously having so much fun with it all!
Thank you!
There are so many ways to overcome ISP limitations to drive your own server at home.
Sure, but ... he's 16.
I'm hoping he figures it out, one of those eurika moments are absolutely a blast.
This comment is nonsensical
Dock’n’Roll looks interesting, but, alas, he didn’t share the source code (or did he?)
Thanks. I haven't had time to clean up the code, but I would be happy to share it if there's enough interest. :)
It's mostly tailored to my use case but it should run fine in most environments.
I think there is a lot of interest in tools like this. Self-hosted and hobby development movement is on the rise, and industry development practices don't really fit hobby developers, we need something more lightweight.
Of course not; most 16-year-olds are enjoying their teenage years instead. (as they should) And build social skills.
Haha that's totally fair - but I'm definitely still enjoying my teenage years and building social skills. I work on my hobby stuff in the free time I have after enjoying my teenage years, hanging out with my friends (and studying for school) :)
Thanks for the concern though.
Thanks for the reply. No concern though, i was just too nerdy myself w/ computers and missed out on important teenager / young adult stuff. This time never comes back! Most dorks here wouldn't understand though.
Very cool to see! I find it interesting that being young you didn't automatically gravitate towards Linux. It is IMHO much more interesting to use than windows, is free in a variety of ways, and gives you the power to customize as much as you desire. You did join the light side after all however so all's well that ends well :)
gkrellm might have single handedly brought me to Linux. So much bling
In my case was WindowMaker converting it into something totally alien to a NeXT; it was unique with it's GTK themings, translucid RXVT's and yet running much faster and stable than Windows.
How did you acquire so much knowledge in such a short time? Can you share your process with me?
Never underestimate the value of motivated experimentation.
1) Set a goal (e.g. host a Minecraft server for your friends)
2) Spend all your free time trying to achieve it
3) Go to 1)
It says (on his website) that he started coding at the age of 7, so I guess he's had at least 9 years of mixed IT experience.
But learning the basics of setting up a server isn't that hard. One of my first jobs involved a lot of similar work, and I had never touched a server in my life - purely a coder prior to that. Literally my first task was to go down to the server room, install a new server, and install Ubuntu server on it, and set up the server. Took me a week to get that done.
Most senior software engineers don’t have servers in their rooms neither.
Of course not, they can afford to put them in a different room or a dedicated server closet, to keep the noise away from where they work/sleep...
How loud is it with the mod you made to the fan?
Much quieter, but still loud enough that I was told to move it from the living room to an unused room :)
It's always interesting seeing people like this online. I didn't start programming until I was 15, and didn't know how to deploy anything for real users until many years later. At 13, I don't think I knew how to do anything useful.
+1. At 22, working as a full time software engineer at a tech company, I still didn't think I knew how to do anything useful
This is awesome! It's like a retro version of my youth; as someone 25 years older, I realize that not nearly as many people are interested in homelab/self-hosting stuff now. And it's not retro for retro's sake; Varun is also working on the latest tech as well.
(also, it took me too long to realize how close their name is to /var/run which made me smile when I thought of it)
Thanks! Never noticed /var/run. Might be an interesting idea for a personal logo or something :)
Fantastic! I actually used your Quickz in a medical conference a month ago. Thanks for building that!
Thank you, that's really cool. I hope it worked well!
I completely get this - got my first "server" at 11 and a ex-dc "real" server at 13 from ebay after I got my first job.
Really glad people are still doing this! (I learned a lot from getting £5 PCs from the local dump, which now don't sell electronics and now "retro" PCs are becoming more and more expensive - I wasn't sure if this ability for kids to buy a cheap PC and tinker (which may or may not help lead to feeling comfortable moving on to running their own servers etc.)
My first real "server hardware" was a Sparcstation 5 I bought under a bridge at the Dallas first Saturday swap meet. I was maybe... 15? Before that, it was Linux on whatever commodity PC my dad had passed down to me. Running Linux on that Microchannel-based PS/2 was a bitch and a half for a 5th grader.
I had _absolutely_ no idea what I was doing, but I feel like my general comfort with diving into things where I still have no idea what I'm doing started then and there.
I'm sorry for his endangered hearing. Servers belong in the DC where their noise isn't such a problem...
The noise levels are bearable, the constant sound is comparable to the noise generated by other home appliances (although the original location was not ideal). We moved it to my an unused room so it's not audible anymore.
I really enjoyed reading this because it was so similar to how I got into the field. I felt myself reading about you running windows and going 'oh wait until you discover ubuntu, oh you did!', reading about all the services and going 'oh wait until you discover docker, oh you did!', 'nginx, oh i should tell him about openresty! wait hes using it!'
Great job in getting so stuck in and teaching yourself so much, it took me until 26 to get to where you're at now!
Thank you! I'm glad you enjoyed it :)
Ahah I wonder how much of the EFZ you already completed just by doing all those projects
Frankly, having accompanied a few EFZ trainees, I’d say about 170%…
When I was 16 I started an ISP in my bedroom. I had an 8 port octo serial cable into a computer for people to dial into, and an ISDN line going out. It's how I learned linux. Setup SMTP mailservers etc.
That's a pretty dope setup. I was running a web hosting company at that age (and likely near the same time period? late 1990s into the early 2000s). Couldn't have done the ISP thing because not even ISDN was available in my area.
I started this way too, but many decades before. I wonder if OP found the R720 idrac “hack” on my blog! Kudos fella, you’re just scratching the surface of the fun to be had. Bring a towel!
Thanks! I found the command here: https://blog.filegarden.net/2020/10/06/reduce-the-fan-noise-...
Is that your blog?
Raspberry Pi
I'm surprised no one has mentioned Raspberry Pi.
You get a decently powerful compute node, no noise, tiny footprint, for ~$100.
Beelinks on Amazon have better cost-benefit unless you use the RPis without accessories.
This is super cool; I wish I had this level of initiative when I was 16. I'm sure you have a bright career ahead of you!
at university we had a dorm wide private file sharing network, it stretched across all the dorms and over to a sister school as well
I visited the top file sharer and he had a rack in his dorm room too, looked just like that
kind of funny for 18 year olds
I think you will like university
When I was a teenager I had a bbs running on my computer. When I wanted to play, I just opened the line with a phone so it would be busy if someone wanted to call. Oh the memories.
I had rack mount equipment running in my house for a while; not sure how he sleeps with it in his bedroom. The best part of the "cloud" was getting all those boat anchors out of my life.
I don't even lift. This 16 year old guy during the covid is going the same way as me, but I'm in my 50s. The story is my story is just like me that Im very happy with him and Im sending him the best of wishes and ---if you can read this --- Im proud of you. And I am jealous. Man, you are 16 and you must know that you are achieving so much, especially because you are free and brave. We need more people like you. Edit: Im not a dev oc. I am a lawyer.
I remember going through the entire qmailrocks.org guide on a PIII with 128mb of RAM (Running Fedora 1 or 2) and then saying "ok, send me some emails!" and my buddy immediately crashed it within like 30 seconds. We were laughing so hard.
Smartphone in hand, but no Raspberry Pi experience, no server anywhere
Most 16-year-olds don't have servers in their rooms
..but some have mainframes [1].
I didn't ever have servers in my room... i kept them in the basement! I had a PBX running asterisk for the house landline, a file server, a monitoring server, and a few sandbox boxes. All old PII/PIII machines. My PBX called me (on my cisco desk phone) every morning to wake me up, and made me answer some basic math questions or else it called back. I sometimes forget how much i was up to with technology and linux.
we're in a golden era of home servers really. You can pick up affordable refurbed servers like this, or just run a mini PC. The amount of performance per dollar and per watt is incredible.
I encourage everyone to run their own services. Some stuff I've run * Social Media automation with Puppetter * Smokeping network monitoring * Flan for network security scanning * PiHOle / AdGuard DNS Home blockers
Get a Pi, a NUC or Beelink Mini PC for ~ $50-$150 and you can run a dozen VMs or docker images.
16 year old me would be super jealous. Awesome work, and I hope you keep flying high
As others have said, I was this kid too. Go go go! Yay for homelabs.
Let me know if you’re ever interested in a paid internship, Varun! Very impressive work putting this all together.
I love stuff like this. When I was 13, my "server" was the old family desktop running a Minecraft server I could access remotely via Hamachi. I really only programmed as a necessity to use Arduino. My tinkering was more of the "take apart batteries and set things on fire" variety.
Recently bought a Raspberry Pi 4, spun it up and installed various things on Ubuntu running from an SD Card. I now have a server running in my room again after probably a decade!
I'm not 16. :p
I sometimes wonder what got me into servers as a kid. I guess part of it was that the internet and open source were new, and it seemed like to be a "seed" in the network, rather than a "leech," you needed a server. Plus I had already gotten into building PCs, which is to say, just assembling and disassembling cases, motherboards, and parts. So maybe for my teenage brain it seemed like having a server made you part of some secret internet club, and used a skill I already had.
Around the year 1999, I was fortunate to literally have access to my Mom's basement with adequate power and a dedicated DSL modem with a static IP. Which was a total stroke of luck, that a small Seattle ISP called Speakeasy was willing to set that up, even though I was all the way in a suburb of New York. I guess I was part of their "nationwide expansion plan."
I was 15. Once the house got this static IP and a fast enough uplink (I remember it being 1.5 Mbps down and 768 Kbps up -- that was enough!) ... I felt compelled to make good use of it. So I would cycle through old desktop PCs and use them as servers.
In 2004, I was away at college but still had access to that wired basement. Using some money from summer jobs and odd jobs during the school year, I bought a 1U rackmount server off eBay to store there as a Linux server, and this let me set up a small array of hard drives in it. The static IP let me do remote development via SSH as if that basement were a colo. My biggest issue is that I had no monitoring for power and WAN uplink.
I still have a record of the specs: Intel Entry Server Platform, P4 2.8Ghz, 1GB of RAM, two SATA hard disk slots, and built-in dual-NIC (one 100Mbps and one 1000Mbps), which I managed to get for around $600. I then managed to find two 250GB drives to put in there. I ended up using that server as my web development platform from 2004-2009.
I also remember one summer where I did remote development over the LAN on that server, working from my desktop development machine (also Linux) via the 1000 Mbps local LAN port, which felt like living in the future. I was using sshfs and NFS and the like to test code directly on the web server and turn over web app releases to the client for review. This was 2005.
After which I eventually found cheap colo and cloud options. Funny enough that static IP DSL line was in my Mom's house until she sold it in 2012. Thanks for all the fish, Speakeasy!
p.s. there's this fascinating 2002 article still available describing Speakeasy and giving a little vignette of late 1990s Dotcom 1.0 history, where people who ran a popular internet cafe in Seattle could fancy themselves telecom executives! https://archive.seattletimes.com/archive/?date=20020703&slug...
You know, it's certainly less common, but it's not rare like it used to be.
When I was in high school in the 90s, I did consulting gigs until I was able to buy a 486 laptop. I was then the only kid in my entire 1200 person high school with a laptop, period.
So, let's do some terrible math and assume that about one in a thousand teens in the 90s had servers in their rooms. 30ish years later, I'm willing to be it's more like one in fourty to sixty.
Even 1-2% of a population represents a lot of individual 16-year-olds with servers in their rooms.
What's really wild is that today they can find each other easily. You're not limited by folks you can reach via FIDOnet.
When I was 13, I had a laptop 'server' with TeamViewer running, which allowed me to play games and run Windows from the school computers on the TeamViewer flash client.
It definitely impressed a few people, but was horribly slow. Later on I learned about SSH, and used that for all my remote management.
How much do you pay for the energy, and how thick earpads do you have with these optiplexes in your room?
Happy to see young people hacking away like this. Sometimes I look at teens, and even those tech-oriented seem to be so stuck in the smartphone/tablet world, that makes me a bit sad inside.
They should. A Prodesk or RPi is cheap :)
I had an AMD K6 with about 128MB of ram when I was 16 - running FreeBSD 2.2 on I think a 10GB drive with a TGUI9680 AGP graphics card and a Hayes v.92 modem. Before that I had a PC AT with I think a 20MB drive and VGA graphics; it was a significant upgrade. Now you know about how old I am.
This is great, good on this kid. Networking was a mystery to me until I was like 18 so I missed out on some good years of self hosting. If I could go back, I'd buy a book on basic networking at 14 with some birthday money. That would be huge. I just didn't even know what to search for at the time. With all the free time I had in high school, this would have been a ton of fun to play with.
This kid’s got a future.
As for taking your servers with you. You don't have to! Use tailscale (or wireguard). Just downgrade your hardware a bit beforehand perhaps in order to decrease burden.
When I was 16 I was definitely hosting services for people on a monthly VPS, but it wasn't until a few years later when my cloud provider imploded with 2 hours notice while I was in class and I lost all my files that I started to self-host with my own hardware over a VPN.
Really cool! As a 16 year old I also had servers in my room. I still have them. They're towers, but they have ECC RAM, Xeon processors, and lots of drives. Arch Linux is the base operating system, managed by Ansible. Everything is connected together with a mesh network. I have Nomad and Consul running on each server, and self-hosting has been great fun.
I wish we can scrap everything and start over We would make the internet a safer place to be We could let the energy flow easier And protocols are more secure
When 18-19 yrs old I used to have a rack in my room with 5 HP blades + 1 router + 1 switch. I had 4 startups. All failed...
Excellent! You are off to a good start.
You will likely find very few non-16 year olds doing that either. Locally I mean. I’m sure that at least five in this site do :-)
You should know that a lot of IKEA coffee tables and shelves are the exact width of a equipment rack. You can buy ~$8 rails and turn that table into a rack for that server. though the weight may be an issue.
Edit: found one example: https://boingboing.net/2020/08/14/lack-rack-ikeas-cheapest-t...
Oh the days when I was 14, compiling KDE on FreeBSD overnight. Fun times.
I love your enthusiasm, varun, and how you're chasin the things that exhilarate you. uni will probably do excellent things for your curiosity, best wishes
When I was 16 it took me a while to realise a "server" is just a loud PC with a weird shape that isn't optimised for use in the home. PCs cases are easier to work in and can fit larger (quieter) fans. Rack mount is good for networking equipment once you get above an 8 port switch, though.
But they should. Imagine something like a Helm in every household
Having a server in your room was a hell of a lot more common 30 years ago than it is today. I knew guys who ran BBSs off a spare phone line, one dude that was running a small ISP out of a studio apartment, and several guys running everything from personal websites to file sharing nodes off of dedicated equipment stashed in a back room.
Never owned a "server" per say, but I used to build and sell computers in highschool, right out of my locker or the back of my rusted-out Chevy. Also bought and resold air guns and pellets out of my locker too, I was cyberpunk before I knew what that was.
It took me a long time to get internet access because my dad spent a lot of time on the phone, and I could only dialup at really short intervals which were better for BBSing. I was in university before I really had a chance to internet.
This is great. Back when I was a teenager I found some ancient servers on craigslist and picked them up, they ran in my bedroom closet and sounded like a jet engine. Think my parents were concerned
At 16, I had a 24 hour BBS running in my bedroom.
My first "server" was my dad's old PC, which was moved to the basement and given an 802.11b PCI card so that it could connect to the internet. I pirated a cracked copy of Windows Server 2003 from eDonkey to install on it and built an ASP.Net website in VB. I wanted a domain name, so I used a dynamic DNS client to set my .TK domain to point at my home's IP (with port forwarding on the router—running on 8080 since Verizon blocked 80), updating every ten minutes.
I learned so much from the experience. If I was a kid today, I'm not sure I'd run a server in my home, though. For as much as I learned, there's far more that you can do and learn running code on other people's computers. I spent so much time trying to get the thing to just _run_ in the first place that I missed out on learning and doing a lot of the things that having the damn thing in the first place would have enabled.
For two years, I taught a class of high school students how to make websites. We used Glitch, and I'm still consistently amazed at how much you can do with very little to get started. I can only wonder what I'd have created if I had what kids today have.
You’re a natural but don’t forget to slow down, get done with school, find meaning outside of tech, and then change the world. You’re already on your way.
When I was 16 my “server” was a hand-me-down laptop with a broken hinge that I kept in a drawer in my dresser because I thought it was cool
I bought an old rackmount server from a colleague who had in turn bought a bunch in an auction.
I lived in New York City at the time and had to physically carry the server for about 20 blocks home from the train station. The entire way home I though "Man, this thing is HEAVY"
I then got home, plugged it in and realized it was LOUD. I though "I know, I'll disconnect all the fans". Turns out even with all of the fans disconnected it was still the loudest thing in the apartment.
I ended up ditching that server and just buying a refurbished small form factor HP desktop machine. That worked great, was quiet and ran find until a Hurricane Sandy related power surge fried it.
1200 baud modem on an atari st at age 6, building pcs by age 9, seller's permit to buy computer parts at cost at age 10, turbo/borland c at age 11 (also some vb and mark williams coherent, but we don't talk about that) slip and slackware at age 12, dug trench for phone line and full-time bbs up at age 13, first software job at 15.
Imagine the hearing damage having a server in your room would bring (speaking as someone who got tinitus from a defective Toshiba laptop harddrive)
I had a server in my room at 13 :)
It's cause I wanted a Linux box to host Minecraft servers on, that didn't take up resources on my main computer. It ended up serving the purpose rather well.
Keep at it -- your skills will be growing in demand as the cloud pendulum swings back.
Windows server. This guy is a pervert.
I decided to stick with the pre-installed Windows 10 Pro (yes, as a server OS!) simply because I was familiar with it.
Hah! Me at 13 was precisely the same, though with Windows XP Pro that I suspected wasn't entirely legitimate but the seller didn't say and I didn't ask.
That old thing ran so well for its low specs, it was crazy, I was really surprised how little you needed to run nongraphical software such as a web server, ftp server, database, email, dns, torrents, vnc, samba, openarena game server... I'm probably forgetting things. The only thing it didn't run was Minecraft, it didn't launch even with everything else shut down. Oh yeah, and the game server didn't have the best ping, but that might have been the 10 megabit expansion card as network adapter.
Can encourage anyone to buy a second-hand system if you find it fun to self host! Especially a laptop (with the battery removed for fire risk) uses very little energy compared to old desktops, or a raspberry pi (alternative) is also an option nowadays.
My nephew, a sharp kid who started his freshman year at college in September, saw me working on my computer a year ago.
He asked what I was doing and I said, “making a website”. He asked, reasonably, how do you do that?
Not wanting to get into details I said “it’s just writing some files that chrome can then read”. He asked, “what’s a file”?
I told him, but I was floored. He’s not every kid, but it’s had me thinking about how younger folks (I’m 38) will see computers.
Has he never copied a file to a usb stick? Or emailed an attachment? I don’t understand how this is possible.
Phone and Chromebook. Kids through highschool might never see anything else. Assignments are in google drive, etc.
Gdrive still has a concept of files and directories. It is kind of primitive in that way.
It does, but using it is really optional. It works equally well if you leave everything in the default Drive directory and use search to find whatever you need.
If you're not used to thinking in terms of files, it just looks like a list of your most recent documents, with a search bar to find anything else.
I just read through the circa 2010 retrospective on how the Drive team tried to come up with their brand. 99% of the concepts were very based on the traditional physical idioms of files, mechanical hard disks, etc. iCloud was released midway through the project and so then they started making “cloud file” and “cloud disk” concepts. It was nearly an accident they ended up with an abstract shape like Chrome. But ultimately Google Drive icon has stood the test of time while files and disks have faded in to history.
Kids (almost) never use Drive. They access everything though apps like Doc, Sheets, etc.
Source: was until recently a classroom teacher for 7-10 year olds.
I can't speak for iPhones, but Android has a file system and the ability to create directories. I think it would be weird if Apple products didn't have that as well, even if it's a bit hidden. I don't have any form of evidence to back this up, but I think that this has more to do with apps, not the device that kids use. Having everything be an app makes it seem like you have multiple independent things on your device, and the device itself is nothing more than a way to connect to those apps.
I can answer for my kids. Everything they do is in the cloud and accessible on any device. I'm not sure they'd understand why anyone would want a thumb drive.
As for attachments on emails, they don't (even for a 100% online highschool). To turn in a paper they can just hit "share" and select their teacher.
Files and filesystems are almost dead to them. There is a little hope... My son googled to learn why a game was going slow and answers indicated he should delete the temp files. That was his first experience needing the file explorer.
As someone who grew up with DOS and turned in CS homework on a floppy disk, it is almost hard to watch.
We'd better rectify that. The future we're headed to where everyone is a slave to several massive IT corps is not one we should wish on anyone.
(Cue Doctorow links to enshittification and war on general purpose computering.)
The death of the floppy disk has been nice to watch though.
I haven't used a USB stick in years as a tech person and virtually all my day to day document processing is in the cloud. It wouldn't surprise me if the person in question does know what a file is, but doesn't work with them enough as files to really get it.
It’s probably my paranoia but I assume something will go wrong with the network at the conference where I’m presenting. I usually have multiple types of backups.
Kids don't really use email these days, or usb sticks. They share more than we ever did, but directly via apps or mobile which doesn't work with "files". It seems reasonable in this case the thing missing is the abstraction of a "file". Think of how your phone works, you share pictures, video, text, posts, etc. but never files. Sharing to email is my phone's 5th or 6th priority option. It comes after a boatload of social media options, messaging, online storage, even some sort of AWS application I've never even launched.
If your whole computer life is using Safari or clicking apps, then yes, you might not ever see a "file".
Course they have. But to them they see it as sharing the state of the app rather than saving and loading data.
I gather most of his use was on his phone and a school-issued chromebook
A workmate (I do labouring and was doing traffic control that day) brought her teenage son a PS5. I don't know why parents would purchase consoles for their children as you don't learn any transferable skills (I guess you could be a streamer).
I'm not saying being a sadist and give the kid a second hand laptop with *BSD because then the kid will want nothing to do with computers, especially if the laptop is cheap and doesn't have a working driver support (This is where Linux is better from my personal experience).
This is also why I like HTML, I got a free internet CD with Netscape Navigator in 1996 when I was 12 and the three working examples (two with Netscape Navigator and one was a html email with Netscape Mail and News client) I had, I was able to play around with and create and learn basic HTML until I got a modem in 1998.
It's plausible that parents may just want their children to enjoy some leisure? To not feel like every experience in childhood should be preparation for labour?
As a child, I was disappointed when I got a BBC B instead of a Megadrive. But I definitely learned more, and some of my leisure time was programming in BBC Basic.
Weren’t there like 10 years between the BBC and the Megadrive?
Parent comment has big “alpha grindset” energy. Like they hadn’t considered the concept of “toys”.
In my own anecdotal experience, if you are a computer gamer, you miss out on all the "console exclusive" stuff. Stuff like Gran Tourismo or the latest Kingdom Hearts or Final Fantasy or whatever. I can't really remember. Then when it is a game everyone plays, it tends to be something where the console gamers can't play with the PC gamers so it's like
"Hey I just got the new counterstrike!"
"Oh that's awesome! I have that too! We should play sometime!"
"Yea what's your PS5 handle?"
"My what?"
"You don't have a PS5?"
"Nah I have a PC"
and yea.....
That is my anecdotal experience though, and it was always bolstered by the benefits of PC gaming - cheaper games, ample mods, lots of sales, and the biggest game library of any platform. Plus better performance. And I could text chat on my keyboard while playing Quake 3 back then haha. Sadly text chat in games has died off as an art, in favor of voice chat and all the problems voice chat comes with. lol.
Edit: And of course, the PC is a tool, the game console is just a dumb appliance.
Well it plays all ways. PC misses out on the console exclusives, consoles miss out on the PC exclusives, consoles themselves have different exclusives and don't interop. For people getting into gaming when they're young, a big influence is probably what their friend group is playing on, which in turn is probably inspired by whatever their parents decide to let them on. That or whatever games is latest hotness.
When I was young Minecraft forcefully evangelized my friend group, which had previously been more nintendo/xbox, into PC gamers. The origins of a lot of my current interests and career go back to the initial questions of: Why is minecraft running at low FPS? How do I set up a minecraft server? It's a little different nowadays, when some of the biggest multiplayer titles are cross-platform.
Moral of the story: make friends with PC nerds.
I never had much time for consoles and I don’t feel like I missed out on much. Popular games come and go. The joy of doing whatever I can imagine with my PC is more than enough consolation.
There are many people in tech industry who began to learn about coding precisely because of video games. Age of Empires II was essentially the first gateway to my career.
It was my gateway to a lifelong love of history and cultures.
Actually, playing video games is an activity that let you practice real-time thinking. Other activities with real-time thinking include music instruments, sports and social interactions. By real-time thinking, I mean an activity that requires you to think and act within short timelapses. It appears also that activities involving real-time thinking can be quite fun, and few course in schools involve real-time thinking.
Now, learning new languages like html or how to fine-tune LLM can be quite fun. Learning how to draw nice pictures, or how to write nice books can be quite fun too. It's nice to let children explore different activities (or free internet CD if we speak about 1996) so that they can find the ones that resonate with their inner motivations.
I also learned HTML in 1996 (when I was also 12), in part from a cover story in I wanna say PC Magazine. I had a modem and had access to an unlimited AOL account thanks to my school’s computer lab teacher not doing a great job of hiding the password for the account (she also trusted me and my best friend quite a bit…trust we only abused insofar as we stole AOL from the county, back when it was still hourly and not unlimited for everyone), but it’s a similar story.
But you know what got me interested in computers to begin with? My Super Nintendo (and I guess my NES before that). So I rebuke the idea that video game consoles don’t teach transferable skills (what the fuck that is supposed to mean) and that they don’t unlock wonder and excitement about technology in lots and lots of people, especially kids.
What is he studying? I feel like a lot of older people in non technical fields, even if they work with files, could not reason through exactly how to define a file.
What is your definition of a file?
Does it include bundles of files?
What metadata does it include?
Can it exist in more than one place? When a file is opened, is the file what is in processor and cache/RAM or what gets serialized to some persistent media?
Metaphorically, is a file a sheet of paper or folded tan cardstock that holds many sheets of paper?
Is a file unitary or can it contain other files? Is a directory a file? Is a link? Is a device a file?
I mean, a file is an abstraction of several of the choices you are asking about in these questions, and different people will answer different things. ps. I've worked on filesystem drivers...
Ok, can you "reason through exactly how to define a file?"
Is defining something as an "abstraction" the same as saying you can't provide a definition?
Since you are a driver worker, do you consider something a file before it is written to a filesystem?
I read you as needlessly combative here, getting far from the points made. Commenter said they told a kid that websites are just a collection of files, kid replied "what's a file?", and I chimed in that that's not so strange, because even people who work with them might have a hard time nailing down a definition. None of this is incompatible with your soap box, but you're writing as if we disagree.
Everything is a file!
You're a file.
I’m a file. You’re a file. Everything is a file!
Went to a trade high school, the usual classes but a focus on carpentry.
I keep hearing about how the #1 profession every kid wants to be is YouTuber or influencer or something like that. But don't these professionals rely on high-end graphics/audio/video editing software to produce polished videos? Doesn't all of that require a high degree of computer skills, including using files?
Sure, but if they reach that level or have some business acumen, they may just pay other people to deal with the video production and files.
Plenty of tools are popping up in this space. This seems like a growing market.
The concept of a file itself is pretty interesting in itself though, as this buffer that lives on disk and is referenced to indirectly. In Linux, everything is a file. In Emacs, everything is a buffer, and then maybe a file.
Sure, the concept of file itself is an abstraction.
It's a whole different thing to use an abstraction in order to make something understandable and to just handwave everything technical about it. It's not just a different thing.
K12 CS Framework only mentions "file" 4 times: https://k12cs.org/wp-content/uploads/2016/09/K%E2%80%9312-CS...
Computer file: https://en.wikipedia.org/wiki/Computer_file
File system: https://en.wikipedia.org/wiki/File_system
Plenty of bright people don't understand the first thing about computers, even in the generations that came before Gen Z. It just means their calling is potentially in something else. It is possible that the majority of techie youngsters are so used to the abstractions around dealing with filesystems on modern technology that files and filesystems elude them. It's also possibly your nephew is just sharp in another area that only requires understanding how to open and use a web browser, and that plenty of other youths exist with similar cultures around computers as we had growing up.
Then again, a Gen Z software developer at the company I work at was confused on how to download a torrent (not that it matters but the contents of the torrent were legal.) That was at least a bit of a shock for me, as it's difficult for me to imagine someone getting into tech without having gone through the usual piracy phase (that phase lasting shorter or longer depending on the person.)
perhaps writing the -apps, and -data, for the site, might have more signifigant meaning [despite not being entirely correct] relating to a mobile context.
I think your nephew is the exception, not the rule.
Has he never filed paper in a filing cabinet?
Our abstractions are out of date.
I managed my own file system by writing down tape counter numbers on the cardboard cover of the cassette tape.
This reminds me of when I took an Intro to Programming with C++ course at a community college almost 10 years ago. There weren’t any basic computer usage prerequisites for this course. So I guess I shouldn’t have been so surprised that at least one classmate did not know how to locate the source code file they created for their Hello World task in the first lecture.
I totally relate to this, but from the exact opposite perspective. I'm old-school, everything is a "file" to me. It's still weird to me that modern tools have abstracted away from "files" to just data. I'm not saying it's bad, it's just so different than how I came up.
It’s ok, lots of “normal” folks today don’t really know what a file is either. They just know it’s an icon you can drag around.
The kid at least asks questions!
I’m not sure kids, especially recently born ones will ‘see’ computers. In the same way that most people don’t ‘see’ the abundant ressources around us. Air, water taps, power, cars, internet etc. These things will always have been there. Most children’s have a rectangle with multiple circles on it pointing at them from their first hour on earth.
Non-IT parents think their kids "are good with computers" because they can tap the next video on YT.
IT parents show their kids what an OS is, what a File Explorer is, etc. Perhaps it's useless, but it sheds a light on that's behind those curtains, and kids start understanding the mechanics of 'stuff'.
In the same we should be worried that people don't know how to grow their own food, procure their own heat, or build and thrive in a community. Not being snarky, I am genuinely worried about that
Great comparison, that's even more worrying. If we suddenly loose fossil fuels and fertilizers, the lack of agricultural skills will probably kill billions.
Unless you’re in full apocalypse mode, every man for themselves, I’d expect for the most case people will learn from their neighbors quickly enough — as long as a survivable subset of the population knows what they’re doing
What’s going to kill billions is the lack of industrialized farming productivity, but you don’t expect every individual to know how to resolve that
I doubt even the lack of industrialized farming productivity will "kill billions".
It will certainly switch diets and most probably cause hunger and malnutrition for some time. But we can support billions of people, given the right diets. We cannot support billions of people's mcnuggets, beef burgers or cornflakes with milk. About a third of my food comes from stuff I make, grow and harvest myself. I can up that, if I need, but preparing, growing and harvesting are a hobby, not a full-time endeavor. I'm certain I can't up it to 100%. Maybe not even 50%. But some neighbor, friends and family can help out (and I them) and I'm certain with some ~100 people we could be self-sufficient. Sober, sure. But not dead.
(I would die from lack of insulin within months, though. TII. So I am aware of how dependent on industry and high tech I am)
No, without industrial scale fertilizer production, there is no way to grow enough food for our current population. Natural processes simply can't fix nitrogen fast enough. The main reason populations started exploding in the 19-20th century was because we found a way to produce nitrogen fertilizer in absurd quantities very cheaply.
If that link in our global industrial chain breaks, a lot of people will die. We can, of course, still grow food, but not anywhere near as much and it will take a lot more care and attention to not deplete all available soil in a generation or two.
In a hypothetical apocalyptic scenario, we might lose a lot of institutional knowledge about agriculture. It's not hard to imagine some average joes trying to farm the wrong way and accidentally starting another Dust Bowl. After all, that's how it happened the first time.
Feeding 8 billion people is so, so much more complicated than just putting some plants in the ground. That we can do it at all is an absolute miracle of technology. We simply could not do it with only naturally available resources. Not by a long shot.
According to most sources I can get to with my amateur knowledge, a giant portion of proteins is "wasted" by "producing meat". Another large portion is wasted because of poor logistics and wasteful behavior.
If everyone stops eating meat, and starts eating the proteins themselves, we win some 1000% to 25000% (sources range from 10 to 25x the proteins in/output when producing meat). Meat is just a terribly inefficient way to produce proteins. (Edit: to be clear, "meat" differs a lot: chicken apparently being rather efficient, more even than cheese, beef being amongst the worst).
And if then the resulting food sources are grown efficient, and all, or a lot of waste can be eliminated by shortening supply chains, sourcing JIT (no avocado's in winter in Europe, no tomatoes in snow), I'm pretty sure we'll get a long way.
But, sure. If some rich countries keep demanding giant portions of global protein in e.g. soybeans and corn to feed cattle to make burgers for 7 meals a week, then certainly: people in countries who cannot afford to compete with the soy-prices will starve.
We’ve settled on a system of markets to deal with scarcity. I believe the 1,000% to 25,000% win might have some losers you’re not telling me about.
Market is about supply and demand. People demand meat. Regardless of how inefficiënt it is produced.
And, no surprise, this isn't a free market at all. At least in Europe, meat, dairy (and on lesser extent crops) are heavily subsidised.
https://ourworldindata.org/grapher/feed-required-to-produce-... has a quick overview. My numbers aren't wrong, but, as always, they do need context (i.e. not all input is currently human consumable. Oos, many input is grass, which is rather ineffective land-use, i.e. the same km² could produce a multitude of human edible foods)
1/3rd of value or flavour is relatively easy. 1/3rd of protein or calories is a lot harder.
Need those mechanized 2000ac farms growing corn, canola or wheat to supplement the homegrown tomatoes/potatoes/peppers/herbs/apples. But it’s impressive what can be done with a somewhat small backyard in terms of value/flavour and not a lot of effort.
Depends on the local deer population.
Only slightly kidding. They reproduce quickly and are currently plentiful. Overabundant, even, in the central US.
oh, for sure, but they said "make, grow and harvest" themselves, so I excluded that from my thinking.
Get something else to do the work can accomplish a lot. As could fishing depending on where you are.
My diet is vegetarian, so a lot of flavor is that staple.
My tiny lot brought me all the potatoes, corn, beans and pumpkin I need and some more. Our five walnut trees, shared with 12+ households, produce some more protein.
I obviously still buy rice and a lot of wheat and rye meal to bake my bread and pizza and to make pasta.
My point was that I don't need the large industrialized food industry to feed myself. My point was not that I can do without a society with specialized and efficient food production.
Yeah water and sunlight is too much for anyone to comprehend. Don't pretend that someone couldn't learn it in a weekend.
Plant seeds
Water
Sunlight
Pick weeds
Wait
Did you forget a /s ?
Growing food is much harder than that.
Growing a few tomatoes or something is easy. But growing enough food to sustain yourself and your family is much more involved and requires a lot of training and experience.
Most people could likely do it. But it would take them 3-5 years of failures before they could reliably grow enough food for themselves. Probably more on the scale of a whole family.
Growing food year round for your entire family is quite literally impossible unless you have acres of land.
Why do you think it takes that much land to feed an entire (average[1]) family?
Besides plants, you could have animals, and get milk and meat from them. Chickens for eggs and meat too.
It doesn't take too much land for it in my opinion. And if you live in a community that does that sort of thing, you could trade meat for tomatoes, and vice versa, or whatever, without relying on megacorps to give you your calories
Localism is the way
[1]https://www.statista.com/statistics/183648/average-size-of-h...
My gut feeling (before searching) was that it would take around 2 acres to feed a person. Searching for data, it seems like estimates ranged from 1-3 acres per person, meaning that feeding an entire family would quite literally take acres of land.
Can you link your source/s?
Here's the google search I used:
https://www.google.com/search?q=how+much+land+is+needed+to+f...
None of them I found credibly suggested less than an acre per person.
Alright, thanks
You have to feed the animals, too.
Check out the famous book, “Five Acres And Independence”. That was the basic size of a family farm in the USA a hundred years ago.
There’s a widely quoted myth that the UK’s ‘dig for victory’ campaign during WWII deemed a standard allotment plot (~250m2) sufficient to feed a family of 4.
More people at least knowing and having some practice with these things makes communities more antifragile.
At our current rate of consolidation and specialization we're going to get to the point where so few people know how to do particular high tech things we may be at risk of someone taking those people out and leaving a massive capability hole across multiple industries.
Or just burnout, changing jobs or them leaving tech alltogether. I'd argue that many companies are already incapable of keeping their tech stack running as intended due to knowledge deficits.
we've done it to ourselves, by embracing MASSIVE complexity at every turn to get the fancy new features.
i keep thinking about how we don't actually need 95% of the code running at my job to actually accomplish my company's goals. it's insane how we just kept going with the flow while features and maintenance effort exploded.
that said, anyone know how well the BSD's work as a desktop operating system? mostly worried about hardware i guess, since it's already an issue on linux from time to time. but man the linux ecosystem, while wonderful, suffers from massive complexity too, or at least it looks more and more like it the longer i use it.
If only tech was at the point where it would be as reliable as your local supermarket or your house.
When i'll leave for family home later, i'll take my laptop with me. My family is big, so the chances are high that i'll need to run some ADB or photorec on a SD card. I hate it[1]. There is some knowledge/reliability deficit with computers that houses/heatings/food aquisition dont have.
[1] https://www.youtube.com/watch?v=7HvtEhxINvQ
from my experience, heating has plenty of reliability deficit, probably getting way worse in recent years. fancy boilers with computers in them break way more often than they have any right to.
Agreed. Likely in the search of efficiency, which oftentimes brings complexity, which then involves more precision or more pieces to a puzzle, or both.
There is something to be said about the dead simple ways of things for sure, and all depending on object/item but a "sweet spot" where something is pretty darn efficient yet not overly complex.
Our endless forwards towards "the most efficient" everything always carries consequences. People / society included.
Or how a car works. I am always surprised that a lot of young people have no idea how many cylinders their car has or how to do an oil change, never mind any real repair.
Although, maybe that's ok. My car has 0 cylinders and doesn't require oil changes.
That concern must be about as old as agricultural societies at this point. Specialization and division of labor really is nothing new.
On the other hand, single-board computers are now both more capable and more affordable than ever, and arguably more open than most of the WIntel PCs I cut my teeth on (at least until I got my own and installed Linux on it).
The UX of entry level laptops is terrible. Windows 11 + vendor bloatware are poisoning the well, I completely get why people are going iPad-only
except the ease and frictionless experience of iPads and most phones, combined with curated ecosysmtems of apps, means you never have to see how the sausage is made, or learn how to make your own
That’s the whole point. They’re an appliance like a dish washer, fridge or TV.
They’re not marketed as a computer.
Tomorrow I'll strike a blow for this. I got my (12 year old) grandson a Pi 5 for Christmas. I plan to spend as much time as he'll allow on-boarding and hopefully he'll learn the fundamentals. At the very least he'll need to learn how to open a terminal and launch the Docker container that runs a Minecraft server. :) There are other things that are included with the full RpiOS install that he can explore as well, like Scratch, Mathematica and Wolfram. And a few games that I tried and are absolutely horrible.
His father, my son, earns a pretty good living writing code (and now managing a group that does this) and should be highly motivated to assist with motivation.
Good man.
I gave my granddaughter (strictly speaking, my wife's grand daughter, but she calls me granddad and always has. She also now has a sister) a laptop running Ubuntu at the start of the first covid lock down in the UK.
She still uses it. Mostly to watch crap on Youtube, I gather, but who am I to judge? It is tethered to my home router via OpenVPN and I patch it from time to time. I have an ISO27000 registration to keep on top of 8)
That laptop was a customer cast off, destined for the skip.
https://www.theverge.com/22684730/students-file-folder-direc...
“Take their phones away and get ‘em on Windows 98.”
That seems unnecessarily cruel though.
No, those abstractions are invisible.
Most teenagers simply don't understand that a phone and a tablet and a desktop and a laptop and a cloud server are all computers. They don't understand that an app and a web page and that thing you install on a desktop are all programs.
As soon as you educate them such that those abstractions become visible, the questions start to flow naturally.
Yeah just teaching my 17 year old to torrent stuff. That was a learning curve and a half for her. So you don't know what File Explorer is?
I don't think we were ever normal.
The entry barrier got so much lower with learning overall.
Humans just don't care in avg how shit works.
What if files just aren’t relevant as an abstraction any more, 40+ years later? As most data lives somewhere in a database the cloud anyway, the concept of a file as a somewhat tangible, autonomous chunk of information may just be outdated.