Readit NewsMy applications have remained the same for many years my octoprint and retropie don't require more FLOPs as time goes on but I'd really enjoy a modern board that has fewer headaches. Works on any normal USB port instead of requiring specialized power supplies, doesn't brown out and reset as much, doesn't heat up as much, etc. I suspect "a pi 3, but now with fewer headaches" would sell better than "a pi 3 but even more headaches and bigger numbers that you don't want".
I live in an old stone farmhouse with my office in a stone garage across a nice poured concrete driveway. There's wires from A to B under all that, but nobody except an unknown electrician from the 80s could tell you even where they come out at either end.
Powerline kinda worked, with crap download speed and just abysmal upload (0.1mbps max), and I limped along with it for years.
When we upgraded to Fibre, that left the old phone line spare, and as luck would have it went straight from the office to the router cabinet in the house. But 80s electrician guy didn't use Cat5, so my genius attempt to use it as ethernet cable ended up slower than the powerline.
My eventual solution was a crazy powerful point-to-point wifi beam blasting straight through the 3 foot thick stone wall to a receiver in the garage below the office. It sets birds on fire from time to time if they fly through it while Helldivers is downloading an update, but it gets the job done.
Still, I might look in to getting one of these things as an upgrade.
Thanks for the writeup!
I've worked with DB people and running lines under driveways for telco and cableco is BIG business and they will not find your request to bury fiber or cat5 to be even remotely unusual.
The bad news about directional boring is they usually want "like a kilobuck" just to show up. Its a lot of heavy equipment and a lot of dudes to operate it all.
The good news is if they're already down the road they'll come by and bore for like $20/foot because its a small job (usually they only charge $10/foot for long runs)
Permitting depends a lot on where you live, some places treat it as a cash cow and they will brutally milk you, others don't require a permit at all. The equipment takes up a fair amount of space on each side, probably more than you'd expect. Scheduling is like dealing with an arborist. "OMG I need this partially collapsed tree removed immediately its an emergency I have homeowners insurance please arrive in the next hour" well thats multiple kilobucks "Meh please remove this tree sometime and I don't care when" well thats like $250, probably less if cash.
I've seen people spend thousands of dollars on DB or crazy laser/wireless comm gear to avoid spending hundreds of dollars on a stone mason. Try not to pay someone to DB under a stone wall, its usually cheaper to hire a stone mason twice and he will leave the wall in better condition than before you started. All masonry is temporary unless its maintained. Similar logic might apply to driveways, most concrete cracks so if you're hiring a guy to fix the crack you may want to bury a conduit before he fixes it. Replacing an entire driveway is expensive, replacing a sidewalk sized path is surprisingly cheap. If you want sidewalk poured (like for a walkway in your garden or around a swimming pool) its about $50/foot and a driveway would have to be thicker and better prepped, but the section could be narrower than a sidewalk. The point being don't accept a DB bid over $50/ft because its cheaper to replace the concrete at $50/ft.
Telephones only want a twisted pair. Ethernet, popular with businesses for decades, also wants a twisted pair. Now, that pair must meet much stricter criteria to be suitable, such as Category 5 (for 100Mbit) or Category 5e (1000Mbit ie Gigabit) - but it really is just twisted pair cable, merely a tighter specification than your phone.
Suppose you are a sparky (electrician) and you have some jobs where you are to install telephone connections, some where you put in "Ethernet" (presumably 100baseT would be fine) and some they specifically want you to wire for Gigabit.
You could go to your wholesaler and buy a reel of Cat3 phone cable, a reel of Cat5 100baseT Ethernet, and a third reel of Cat 5e Gigabit cable, and take the right one for each job. So long as you do this flawlessly you can probably save a few pounds every year by using a slightly cheaper cable for some jobs.
Or, you can buy one reel of Cat5e and use that for all these jobs and since it's the same reel you can't have the wrong one and don't need to check paperwork to know you've put the correct cable in a duct etc. Thought that was a phone line but now the client insists it's data? No problem, they're the exact same cable, just smile and agree.
When I bought the place where I live now I wanted GigE to this desk, even though the DSL comes into a different room. I didn't love the idea of cutting holes in walls but I was resigned to maybe needing that, except there's a phone extension in this room (like the author says, we do love phone extensions) and so that room the DSL comes into has a twisted pair to here. I opened up the box, and I'm like huh, that's Cat5e, and sure enough this entire building was wired with Cat5e because like I said, why not, it's basically the same cable, why carry a separate reel?
So I changed the face plates from telephone to Ethernet, and I'm done.
Real cat5 and ethernet connectors just work and phone cable and phone plugs just work, but if you mix them you'll get all manner of expensive labor costs trying to figure out jury rigged solutions.
At one client they used two pair for business phone system, we're on a cable pulling team and one guy punches down the blue and green pairs the other side punches down blue and orange pairs (essentially a 568A vs 568B violation) and we spend SOME EXPENSIVE TIME trying to figure out why the cable toner "proves" we are on the same cable so it can't be a wiring fault.
Or the stereotype of the halfway colorblind guy at the far end working in the ceiling, on a ladder, in the dark, swaps the orange and brown pairs as happens sometimes.
Oh even funnier is there's always "that guy" who is too lazy to pull an additional cable to a new phone, so he steals some pairs from a nearby phone, somehow knocking out both phones in the process. Such a headache.
Labor for troubleshooting miswired cables/jacks is SO expensive its just cheaper at work to install phone lines using phone line parts and ethernet using ethernet parts.
The arrival of VOIP phones around Y2K, somewhat after my time, must make life so much easier. And now nobody uses wired phones everyone has a smartphone.
At home if you're doing one line and its a hobby so your time is free, then your strategy does work.
Its really slow. Too slow to use 2FA or in some cases, verify email addresses or recover passwords.
Most people can't handle a notification on their watch every minute, or several spam every five minutes, so "large numbers of people" are shutting off notifications on their phones. And human nature being what it is, they're not going to be turned back on again. So the era of getting a notification when you get an email is coming to a close. "Important Immediate Attention Stuff" moved to text messages a long time ago anyway, at least for me. The list of technologies you can no longer reach me on, always increases over time...
The people who submit fake/AI issues and PRs for resume purposes seem to exclusively use github and this is a substantial expense for the real users (see the recent Curl discussion). Gitlab doesn't have those people (or at least its a wildly smaller problem).
One social downside is noobs will insist a project does not exist if its not on github, even if you send them the gitlab URL. Almost have to physically cut and paste the gitlab URL into their browser for them before they will believe. You can either do nothing which filters them, not a bad idea, or create a clone or placeholder on GH that basically links back to the real repo on GL. I don't know if that is allowed in the ToS for GH but people certainly do it a lot.
Consider how a SCM like git or bitkeeper is more complicated than a wrapper for LXC. For some odd reason Docker has almost 100x as many employees as bitkeeper. They're just too big. It would be like trying to create a startup of "/bin/ls as a service" with at least 50 employees and 49 of them would not be able to generate enough revenue to break even much less turn into a billion dollar "LSaaS" tech unicorn. There's not enough meat for the pack. FreeBSD has jails and all of FreeBSD (not just jails, the whole thing) is about a third the size of Docker... hmm.
An alternative to having an appropriate sized company would be giving up on profit. There probably is no way to make "real" money doing what Docker is doing, not "real" in the context of 1500+ employees. It would be very real if they could get their current revenue with 20 employees, but ... That is not bad, that just means they're better off as an IRS 501(c)(3) approved charity rather than trying to become a startup unicorn. Large organizations like the Red Cross are a valuable and important addition to the community, despite not being a successful tech unicorn. They got a lot of money from In-Q-Tel so they're already kind of taxpayer funded (via CIA) so going outright charity wouldn't be a stretch.
A good business analogy for Docker would be the small day care my kids attended. They were based in a small church building which permanently limited the size of their state license. It doesn't matter if they hire 3 caregivers or 1500, they only have space for an 8 kid license and revenue will never exceed 8 kids. They can hire 1500 caregivers using VC funds but they'll never get more than 8 kids of revenue. They are not working in a field where they can scale to a billion dollars of revenue. There's nothing "wrong" about a daycare that rents a room of a church, employs a couple "early childhood education major" college grads right around minimum wage, and the kids have fun. Thats Docker. There is perhaps a bigger third problem that they probably sold themselves to investors as an unstoppable money printing machine. Whoops. Nobody makes that mistake with the local church daycare. To some extent lack of due diligence is the fault of the investors. We'd never have had docker without their ... selfless financial donation.
This is an example of the HN "Jump to Conclusions Mat" where there is an instant jump to extremely high level politics and philosophy and skipping over the more practical mundane problems.
A more practical issue is the author has zero interest in being sued if my LoRA connected emergency stop button for my CNC milling machine crashes and the machine then hurts someone (possibly myself).
Or my "emergency alert" transponder fails when I'm in the wilderness and someone (maybe me) dies instead of being rescued.
The wildest part of the story which isn't being covered is this is an example of one guy doing all the work to produce something more capable than the entire meshtastic project in about a year. A real life example of 10x or 100x engineers. How can meshtastic accomplish so little if one guy accomplished so much? Historically it was not THAT bad where having more than one person work on a network protocol never killed progress for decnet or banyan vines or SNA or any other old time protocol, but maybe its a mesh network thing that having more than one cook in the kitchen eliminates all progress.
Unfortunately, being a pretty much 1 person project he doesn't have the legal skills to realize the license as written is awful and needs rewriting to achieve his goals, assuming his goals are even a good idea...
I've set this up and used it on my LAN at home. Its a LOT more than just LoRA or just meshtastic and its pretty cool and works well. The app on my phone works well. Being abandonware I'm shutting it down "when I get around to it". The ratio of Meshcore to Meshtastic users/traffic is around 20:1 in my area so I'll be setting up Meshcore to fit in. Mesh LoRA is very local just like cell phone service; I'm well aware there are parts of the world operating at an opposite ratio of popularity where you "have to" use meshtastic to fit in. That is not where I live so I must use meshcore.
Meshtastic isn't used here, so I can't mesh so cross that off. Reticulum works perfectly and is abandonware so cross that off. Meshcore has its ... interesting pay money to unlock features scheme, I can't decide if I like or dislike that, I'd like to cross that off but its the only remaining protocol. I could write my own and GPLv2 it but if a superior system (reticulum) can't get buy in, my better licensed system would also be unused. I think I am stuck having to use Meshcore, I about 95% like that and 5% dislike that.
I do find it amusing that I used ham radio AX.25 packet radio in the late 80s, early 90s, some times this century, I know all about digipeating problems and hidden transmitter problems and all the stuff the "kids" refuse to do a literature search for and seem surprised when it bites them. Really this mesh stuff is just ham packet radio from 1981 except the total cost of a station is like $15 vs back in the day it was oh at least $1000. I had a node running linux AX25 back in the 90s and I'm sure I had a couple thousand bucks in equipment by the time I was done, mostly repurposed later on. I still have several hardware TNCs in some closet or shelf somewhere...
That's the implication of your comment.
Healthy economies "should" have a reward for specialization where both supplier and purchaser win. There is no reward anymore for economic specialization in the context of ice cream; its cheaper to make your own, now. This is a troubling long term implication for any *-as-a-service
There's a second even worse economic implication in that ice cream has long been affordable to 100% of US households... Now due to permanent long term economic decline its seen as acceptable losses for some not to afford it anymore. Again, troubling long term implications.
https://ygo-assets-websites-editorial-emea.yougov.net/docume...
Yes it includes audiobooks in "books".
physical books were around three times more popular than ebooks or audiobooks.
75% did not read anything to children (kind of surprising 25% of the population has access to pre-literate children)
15% don't read books they own, which is surprisingly high. A third borrowed their books from the library.
54% of the population inaccurately think they "own" an ebook as opposed to reality. 40% "a book you accessed for free online" Sure thats all project gutenberg LOL.
Mysteries and Crime are top of the charts. I have no idea if "computer books" count as 11% other non-fiction or academic or hobbies.
Only 51% have a library card. I know they are cracking down hard at my library, show up physically with proof of residence or it gets cancelled. Its harder to get a library card in my community than to vote, get a job, or register for school, your community may vary.
Most people go to the library less than once a month. This sounds about right.
Shockingly 20% of people never go to the library just to hang out. As a parent of older kids I do that a LOT, drop them off then go silently read or compute or whatever at the library. The attempt at turning libraries from book warehouses into makerspaces seems to not be working very well according to this survey.
People own a surprisingly small number of books. A "large full height bookcase" puts you in the elite. I'm kind of surprised at that.
Virtually no one hoards digital or audio books, I am apparently a far extreme outlier in that regard LOL. I'm easily five figures each. From, uh, totally legit sources.
Most people actually own about two dozen books and think most other people own about twice as many around fifty.
Since I was a little kid I always read a little before bedtime and it seems this is very popular.
Most people don't organize their books but think they have an easy time finding them (not unlike how people organize computer files...)
Surprisingly there is zero to very minimal demographic difference in every category among people who do not read, which I find very surprising and unlikely.