My server is a phone now
seg6.space
seg6.space
In linguistics, the terms “theme” and “rheme” refer to the status of information in a sentence [1]. The theme of a sentence is something that is known or can be assumed from the context (that is, old information), while the rheme is information that is newly introduced in the sentence, often as a comment on the theme. The tendency in many (most? all?) languages is for the theme to precede the rheme in connected discourse.
Without context, a reader is therefore likely to assume from “my X is a Y now” that X is something the writer had previously and that Y is something new. Because most people have a phone but few have their own servers, I was initially confused by the title “My server is a phone now.” I had to read the article to realize that the word order is in fact appropriate: for the writer, the server is old information (something he has had for a while) while the phone (as a server, at least) is new.
In English, theme and rheme might be inferred not only from the position in a sentence but also from how a noun is marked. Possessives like “my” are generally used before nouns referring to things that are previously known or assumed; the definite article “the” is similar. In contrast, the indefinite article “a” (or “an”) usually marks information being newly introduced, so it is more common with rhemes.
I worked as a Japanese-to-English translator for many years. Because the grammatical structure of Japanese sentences is very different from that of English, sentence-by-sentence translation often ends up with paragraphs where the information flow is awkward and unnatural. I gradually learned that I could produce smoother translations by reorganizing my English sentences into an old-to-new information sequence.
PS: fascinating insight.
> Sometimes, particularly in speaking, when we want to focus on something important, we bring it to the front of the clause. This is called ‘fronting’.
https://dictionary.cambridge.org/grammar/british-grammar/fro...
The author uses "server" as a role. So "my server is a phone now" translates to "the server role in my setup is now played by a phone".
Maybe this is just a rewording of what you said. You clearly know a lot more than me about linguistics.
This theme-followed-by-rheme is only a tendency, not a hard and fast rule. In casual conversation, people do tend to structure their sentences that way without even thinking about it. In writing or translation, though, a sentence might need to contain a lot of different elements and, in the process of arranging them, a writer might put the rheme before the theme. Readers might think the resulting sentence seems a bit awkward, but they probably won’t be able to pin down the reason. This “My server is a phone now” title was an exception, because (as the comments below show) some people were confused enough to remark on it.
The current title is just a bad English construction. Not a big deal and absolutely forgivable since I don’t expect everyone to know perfect English.
Servilo, the tool that serves something.
Servulo, someone who serves something.
Servanto, some agent (in grammatical sense, so very broadly applicable) which serves something.
Servisto, some professional serving something.
--
That's a big tangential to TFA though, add other noted it suffers from wording ambiguity that don't need any other language to be dispelled. Compare for example "Dæmons dwells on phone now", just as short, far less ambiguity.
It's borderline clickbait, unless the author is an English graduate and assumes all their reader will be too. A phone running a server is not very interesting and has been shared many many times on HN before, while I've not heard of a phone running on a server before (which I why I clicked at all).
As it is, I expected the server to now have a modem, to be operating as a phone.
To me it was a clear and a very fitting title.
(But my first language is german, maybe that has some influence)
The author should have swapped the title with the subtitle: "rooting a CMF Phone 1 to run my personal infrastructure at home".
It would be very interesting to see a poll results on how many of us thought that the article is about a server doing phone stuff and not a phone doing server stuff.
I mean, turning a server rack into a phone just isn't something that makes any sense, so it didn't seem like a valid option to consider. I've thrown it at a few LLMs now to see what happens and they also make the same assumption with no extra info, that it's a server now being hosted on a phone. That's the statistics talking. It's very unlikely to be the opposite.
If I said "my water is ice now," one would assume freezing rather than melting.
So in this case one would assume a server became a phone, when actually a phone became a server.
(English is my first language.)
I think most of the time you’d play with the order it would be for comic effect, like someone saying “my dinner is pretzels now” after having too big a snack.
(I then realized that I have a cluster of 3 phone chips I use as my "server". Phone chips are pretty OK.)
Do you think my personal assistant becomes an AI agent or do you think an AI agent is acting as my personal assistant?
However I still think it is phrased poorly as it does initially seem to imply your personal assistant has become an AI agent, but then I have to use additional knowledge I posses (that people can not become AI agents) to correct the interpretation.
I think a better phrasing would be “I now use an AI agent as a personal assistant”. No confusion there. Your phrase can be understood, just like the article title, but it’s not well-written English IMO.
Sure but they will have other types of ambiguity
"My car is a Jeep now" is most likely saying "I switched to using a Jeep as my car" because, short of a large merger, it's not likely that the same car changed its brand.
The same server can become a phone and also the identification of my server can be switched to a phone.
> "| switched to using a Jeep as my car"
"My server is a phone now" is exactly the same: "I switched to using a phone as my server"
> because, short of a large merger, it's not likely that the same car changed its brand.
I find a server rack becoming a phone that I can carry around and call people equally unlikely. And because of the improbability, it's ruled out of my interpretation.
1. Server becomes phone 2. Phone becomes server
I guess more people find 1 more probable than 2? That's surprising to me and especially on hacker news.
HN comments sometimes try to argue this is impractical or impossible
would've been better
--
> built in battery backup
Isn't it always recommended to remove battery if you plan to use a phone or a laptop as a server? ( Fire hazard )
or is it fine to just limit the charging to 80%?
But in other hand, it is probably a pain in the ass to remove the battery and do some workarounds to make sure the phone correctly behaves without it.
'Very quickly' seems to be measured in decades from what I see
Most people I know have their laptop docked, either near-permanently at home or taking it between home and a desk with dock (so it's nearly always plugged in when in use)
The 'servers' I've had were all old laptops - yes, with battery (don't have to buy a separate UPS if one is already included). I've now also had two non-server laptops that were plugged into power pretty much all the time. I treated my first laptop perfectly: taking out the battery anytime I'd be in a classroom for several hours and it was already full enough (you know, back when sliding out your battery was possible in the first place), but I'm not noticing any faster degradation on the systems I treat poorly. I'll activate the 80% cap function where it exists (that is, in the latest phones and my work laptop's UEFI config panel), but the evidence of it being helpful has yet to materialize
I have read that studies show that charging to 100% causes degradation and I trust those more than random people's anecdotes (and I'm a random person), but I'm starting to wonder if I've maybe just misunderstood the study results. Maybe it's like how smoking increases your chance for some cancers but you might also be fine in the end. I see people with plenty of much younger and better-treated batteries that can't sustain literally 30 seconds of use whereas mine are reasonably close to the new state (some haven't had many charge cycles, just sat at a stressy high voltage all the time, which was indeed supposed to destroy them), and I don't know how to explain that otherwise
Anyway, my point is, maybe his phone has some mechanism similar to laptops to prevent this kind of thing, but in the end it's a phone. It's not the intended use for it.
The phone's battery can be reported to Home Assistant, where an automation can be triggered to turn off power to the phone when the battery gets above a certain percentage and turned back on when it goes below.
Not the best solution, but certainly an alternative from messing around with trying to get the phone to work without a battery.
Agree! But do get one or two of those "small automatic fire extinguisher" or whatever they are, that you can attach on stuff, for electrical panels and what not, and put right next to it together with a traditional fire alarm, and you're good to go :)
Putting the phone into a metal box, and the box, onto something not flammable, would help.
Maybe eventually everything around it can be so wet/burned out that the battery fire slowly just ends? The room would probably be fucked anyway...
Edit: Or I should say, it was 20 years ago. Maybe modern stuff is worse, but my M2 MacBook Air and Framework 13 Intel Core both power on without a battery.
By saying "my server is now ..." implies a change of server platform to a phone but not where it came from.
Your offering: "my phone is a server now" implies a change of phone usage but does not indicate where the server came from.
The article itself tells all.
Except the original title is "my server is a phone now", not "my server is now...".
When I read the title, I literally thought "wow, I've seen a lot of people make servers out of old unused phones, but this is the first time I've heard someone make a phone out of an old unused server!"
LiPo batteries will fail over time if left fully charged. If you don't need the battery it's best to keep it somewhere around 40%. If 80% is the only option, that's usually good enough.
The primary failure mode starts when the cells are charged near their maximum voltage and kept there.
See also, this atomic clock wristwatch: http://leapsecond.com/pages/atomic-bill/
In normal conditions, you use the product while it’s active. But with a server you are not daily checking.
You might have a bug in your changing limit. I think the constant higher temperature could also become an issue.
Not really. No sign of a phone. Just a computer — same as what all other servers run on.
"My server runs on a computer with a form factor that is atypical for server workloads" gets a little closer to the heart of the article, but even then it seems the real meat of it is about Android, not the hardware (although that is mentioned too).
How about "My server runs on Android"?
Turkish is my native language, and its sentence structure is pretty different from English, so that occasionally sneaks into how I phrase things. I can see why “my server runs on a phone now” is less ambiguous.
Anyway, I’ve enjoyed the accidental linguistics thread :)
For a long time, even though I consider myself fluent in English and primarily speak English every day, both at home and at work, I’ve still felt like I’m not quite where I’d like to be in terms of proficiency. That feeling actually kept me from writing until fairly recently, because I was always worried that whatever I put down wouldn’t come across as coherent (or maybe polished?) enough.
I used to write a lot in my native language back in the day, so I think part of that hesitation came from knowing how naturally writing used to come to me and feeling like I couldn’t quite express myself the same way in English. More recently, though, LLMs have helped me with that. I feel much more comfortable putting my thoughts into writing again.
I actually feel like this is such an interesting topic to explore. I’d be curious to somehow experiment on myself and see whether I’ve started sounding like an LLM recently, and if so, whether we might eventually see people converging toward speaking, or even thinking, in the same way. Scary thought honestly.
> At that point I had neither a server nor a phone
I don't know why I like it, I just do. Perhaps it is the slightly humorous tone. Anyways, nice article!
without root, you also can't bind to ports on any android phone, and rooting often requires an unlocked bootloader and not all android phones support bootloader unlocking. You also need an unlocked bootloader to install postmarket os.
so without binding to ports you can't run a web server unless you run it through a proxy server which the author did, or containers at native speed on postmarket os because you will not be able to flash the phone with postmarket os if you have missing kernel syscalls on android nor root the phone for the speed increase on android depending on your android kernel version i assume.
so with android containers may or may not work without emulating x86, and without root chances are it will be slow specially if you are forced to emulate it as x86 because it turns out you have missing syscalls, nor root to make them faster if they exist on android. again no chances to install postmarket os nor rooting to make containers faster, without bootloader unlocking.
not all android phones can have their bootloaders unlocked unless you find an exploit. https://github.com/zenfyrdev/bootloader-unlock-wall-of-shame
Rooting is required to install postmarketos, and making termux faster and bind to ports on Android. You don't need postmarketos if your android version can run containers on termux with all the required kernel syscalls
To root or install postmarketos you need an unlocked bootloader.
Termux is only slower without root if you use proot. PRoot provides a limited implementation of chroot that runs unprivileged in userspace, but this also causes it to be pretty slow. If the server software you want to run is available in the Termux repos, you can use that and have it run at full-speed.
Termux works by patching all programs to treat /data/data/com.termux/files/usr as the root directory and /data/data/com.termux/files/home as the home directory. This introduces some compatibility issues, but it doesn't make it any slower.
The Termux project has a great wiki at https://wiki.termux.com/
Even in this environment of inflated hardware prices, you can get one for a reasonable price.
Although I do see the appeal in getting unconventional hardware to do what you want, just for the love of the game.
For a home sever that's mostly idle, it is almost like commuting in a pickup truck.
I had a Nuc 9 extreme with a 10gb SFP card in it.
It averaged 69w. The Mac Mini I replaced it with uses just under 5w.
A Mini is possible the worst bang for buck. Also, the horror that that is Desktop Docker can’t be over emphasised. I know there are better ways of using Docker on a Mac, but having got something working I can’t face rebuilding it.
I also have some Fujitsu office machine I picked up for $50 as my "I need to run some random task on a physical linux box" machine with an i5-7500 and it idles at 15W
OrbStack is my go-to when it comes to running docker on Mac. It also uses macOS native containerization APIs so containers will be a bit more performant than they would (at least use to) be on Docker Desktop.
Another good approach if you’re fine without a GUI or want to bring your own is just ‚brew install docker‘.
I’ve got it working, but compared to doing it in compose, it’s awful.
I've got an Optiplex with an i5 and two hard drives and that idles around 12W when idling.
Some mod Xeon systems to use regular PC PSU like Be Quiet! ones and Noctua fans, etc. but my HP Z440 workstation is so quiet already I didn't need to.
In my case it's an headless server and is sits between the vacuum cleaner and the cat's food: there are literally only two cables connected to it, the power cable and the ethernet cable.
I can live with the footprint of one tower.
I run many VMs and containers inside those VMs: Git server, pastebin, NAS, containerized/VM-isolated AI agents, etc. but I don't need it to be up 24/7. I basically only need those when I'm working. So electricity usage ain't a big concern to me as it's more like running 8/7.
I don't feel like I'm commuting in a pickup truck: I feel like I have the perfect machine for the task.
I'm using a PCIe expansion card to get double NVMe drives for the cache disk, and then three (up to five) 3.5" spinning 8TB drives for long term.
I tried really hard to get it on 32GB but sadly the motherboard only has two slots and only supports 8GB DDR3 modules. I literally ordered some of the rare late-generation 16GB ones and it refused to boot, with no firmware updates available from the vendor (Advantech, board is an AIMB-274).
Buy a refurbished Xeon workstation with DDR4 RAM. If you're in the US or China they're basically free and if you're in the EU, they aren't too pricey (like you can definitely find ones for less than 200 EUR). Slap your double NVMe expansion card, your two NVMEs and your 3 3.5" spinning rust in it. If the workstation doesn't come with a Xeon with lots of cores, just buy one on eBay for $25.
ASRock H67M-ITX, i3 2.5GHz, 16GB DDR3 1333Mhz, 2TB+4TB SSDs, DVD-RW, in this Mini-ITX Antec case[1].
My first home server is a 2nd hand gaming PC that I bought for 500$ with 32gb and 7700rx video card that I use for GPU based tasks but it's quite audible when it's working so wife and I agreed it has no place in the living room anymore and is being deported to the corner of my office.
You can still find really good bargains for home servers from refurbished good computers, and given how well solved tech reliability is now, it's really not worth buying new.
Because there are so many of them (at least in Denmark where I live) they are pretty cheap and it's easy to get one with an I-processor.
My current machine's idle draw is also very low at 5-6w.*
*Based on some quick research it does seem that I was a bit lucky with mine, so I guess it's fair to mention that mileage may vary.
Old Optiplex are also good deals. I bought a Dell Optiplex 5070 MT for $150 for myself and 2 Dell Optiplex 3070 SFF computers for $75 each to replace my mom's aging Windows 7 PC. All 3 computers take DDR4 and M.2 NVMe SSDs and have an Intel i5-9500 CPU that will also run Windows 11 pretty well, if you're into masochism that is.
The future used to be so cool. Having access to Wikipedia and live reporting always on hand was going to make everyone so informed, easier communications were going to make democracy and grassroots organizing flourish... We were all going to have our own server on our pockets...
We were naive, but I sorely miss the tech-optimism that to some extent carried me through uni.
"Prior to deployment, smartphones must be processed to remove all but the motherboard, which contains the core compute functionality. Note that the motherboard is responsible for the largest fraction of embodied carbon (approximately 50% based on internal carbon footprinting assessments), so this effort targets the most impactful components."
And they accomplished the author's first goal to run real GNU/Linux on the phones.
https://research.google/blog/a-low-carbon-computing-platform...
The new AGI goalpost will be "able to deploy Linux on everyone's old phone with drivers for all the hardware ".
That said... there are Chinese rackable phone KVM thing, available on Chinese e-commerce sites, intended for social media fraud farms(faking likes, aiding account thefts, sending scam DMs, calls, SMS, etc). Taking them apart should be already common in those scenes. Reportedly those are operating at some scales globally in various third world scam malls - shopping malls by architecture, but operated by gangs and inhabited by scam shops. So it's readily possible, if necessary.
https://chinese.alibaba.com/product-detail/Mobile-Phone-Group-Control-Equipment-Motherboard-1601460564510.html
https://www.alibaba.com/product-detail/For-Samsung-Custom-Packaging-OTG-Memory_1601872798649.html
https://aliexpress.com/item/1005010070778997.html
https://aliexpress.com/item/1005012481751631.html
https://www.tiktok.com/@xbus260430425/video/7671530564876258573
https://www.tiktok.com/@xbus260430425/video/7663701554905582862
https://www.instagram.com/reel/DP_UrdhEm8R/
Took me few minutes, but looks like one query that works for these on Alibaba/AliExpress/taobao is "click farming phone". "手机 主板 服務器"(phone mainboard server) might also work in some places.I see bunch of multi GPU crypto mining style chassis as well as just USB mass charger styled ones. This is something I just came across randomly so I have no experiences or connections for these, but hope this helps. It's kind of cool looking if it weren't for scams.
edit: note, it's not clear if they come with the phone PCBs or phones are customer supplied. The "boards" mentioned in some of listing seem to imply that it refers to adapter boards for the phones, through subtle wording
I’ve wondered for a while if they could be used as an extension to your laptop like the old 80s PCs where you would plug in hardware.
So ‘make it a server’ is the obvious way I guess, and then you could do things like render webpages on there and then stream them as images to your actual computer (like the now shutdown Mighty app), or run ffmpeg on there to convert a video, etc
Any day now!
Doesn't the CPU throttle if the phone gets too hot?
Or maybe some parts like the eMMC aren't spec'd for things like constant logging?
things might have gotten better though, I last had convos with people who did this maybe 6 years ago at this point.
I have an old oneplus 6 and all I did was plug it into my computer and told Claude what I want and it just worked.
I even told it to make sure the battery doesn’t charge above 80% to not fry the battery and it couldn’t find the setting so wrote something to make it work.
I feel like this whole thing would have taken a weekend pre AI - instead I did it on a whim and spent about 5 minutes on it and it took less than 2 hours of wall clock time.
People immediately see the cost of a tablet, the issues with charging, building the app and updating the data, customers not understanding the interface etc.
Yet, they don't seem to grasp all of these are basically the same for a QR code: you're now maintaining an online service just for your restaurant. People still need to understand your site. You'll be dealing with their browser, extensions, languages etc. And you still need the QR code somewhere on the table, visible, explain that it needs to be scanned (by one person ? everyone ? do they need to keep the tab open ? etc.). These will all be costs, you pay either upfront by hiring experts to make the experience good, or on the go with crappy customer experience, your staff spending more time than needed working around the system and explaining it to your customers, and those not coming back.
There are super cheap and simplistic ways to deal with all of that, and I assume GP has seen many of these.
PS: The weirdest experience I ever had was someone at our table scanning the code, to copy/paste it on the browser on their personal profile (default was work browser). They missed a digit in the url params and ended up opening the order menu of a table at the other side of the restaurant, and just started ordering, then waiting for a pretty long time while the other table was dealing with the mess.
1) It's optimized for the restaurant first, and the patrons second or not at all, and it shows.
2) Someone thinks "hey, we could sell ads on this", and now it's even more a piece of garbage. (This is also what happened with all the little kiosk-tablets on the tables at some restaurants. We just set them face-down on their screen and talk to a human.)
3) Most of them never anticipate unexpected problems. They might have an "order X" button, but they often don't even have a "call server" button. And when they do, it often doesn't work, because the other end isn't hooked up to something that gets people's attention.
The much lower-tech and more enjoyable solution I've seen used successfully many many times is a call button on the table. Press the button if you want service. Bonus if it has a "press this other button if you want to be left alone".
Having backup tablets on hand is a good hybrid.
Besides that, there are people who have no smartphone or an empty battery, can't use a smartphone (one of my grandparents can read a menu with a pocket magnifier, but a touchscreen phone is too cumbersome and so they don't know how to use that; another grandparent has memory issues and never learned how to use a smartphone), don't have signal (you need to provide WiFi), have an allergy and need to ask questions or ensure that some ingredient is really taken out, etc. I'm not against the option to order impersonally but a good kitchen communication system would go a lot deeper than slapping wetware with slopware
Using QR codes to not having to talk to a human and to "optimize" humans out of the loop in a restaurant sounds like a great scenario for an episode of black mirror.
QR code menus are one the shittiest experience people going to a restaurant to enjoy a meal can have.
> Do you not like this option?
The entire world fucking hates it. It's totally dumbfucktarded.
Sorry to be blunt.
pmOS is a proper port with native Linux drivers, but there’s also Libhybris / Halium which runs Android drivers in a more typical “Linux” userland. Ubuntu Touch uses that, so it could be a good starting point. Sadly, I don’t see CMF Phone 1 in the devices list: https://devices.ubuntu-touch.io/
Yes there are ways to bind to server ports with official APIs, example can be requested to Fable, Claude, or whoever you feel like it.
Also, you can't install things on termux if you have android 6 or below. You must install postmarketos and thus an unlocked bootloader.
<uses-permission android:name="android.permission.INTERNET" />
Then you need to write the server handling code on a background application thread, with perfectly working server sockets.
This only works on wlan, or local network when using bluethooth, but who wants data modem for a server anyway.
With Gemini's compliment,
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.withContext
import java.io.BufferedReader
import java.io.InputStreamReader
import java.net.ServerSocket
import java.net.Socket
import java.net.SocketException
class AndroidTcpServer {
private var serverSocket: ServerSocket? = null
private var isRunning = false
// Call this function from a CoroutineScope (e.g., lifecycleScope.launch)
async suspend fun startServer(port: Int = 8080) {
withContext(Dispatchers.IO) {
try {
serverSocket = ServerSocket(port)
isRunning = true
// Loop to continuously accept incoming client connections
while (isRunning) {
val clientSocket = serverSocket?.accept() // Blocks until a client connects
if (clientSocket != null) {
// Handle each client connection in a separate background routine
handleClient(clientSocket)
}
}
} catch (e: SocketException) {
// Triggered when stopServer() closes the socket
println("Server socket closed: ${e.message}")
} catch (e: Exception) {
e.printStackTrace()
} finally {
stopServer()
}
}
}
private suspend fun handleClient(socket: Socket) {
withContext(Dispatchers.IO) {
try {
val reader = BufferedReader(InputStreamReader(socket.getInputStream()))
val outputStream = socket.getOutputStream()
// Read incoming text from client
val incomingMessage = reader.readLine()
println("Received from client: $incomingMessage")
// Send a simple response back to the client
val response = "Hello from Android Server!\n"
outputStream.write(response.toByteArray())
outputStream.flush()
} catch (e: Exception) {
e.printStackTrace()
} finally {
// Always close individual client connections
socket.close()
}
}
}
fun stopServer() {
isRunning = false
try {
serverSocket?.close()
} catch (e: Exception) {
e.printStackTrace()
}
}
}
As you might imagine by my comment history, I am a big fan of letting Android be Android, with Android native apps, not stuffing GNU/Linux into an APK.Containers are nice, but they're not strictly necessary. On Google Pixel 6 and newer phones, you can take it a step further and root your phone and run a KVM-accellerated arm64 Linux virtual machine on your phone. There's a port of Limbo PC emulator that includes support for the KVM accelleration, and it works quite well! It even allows you to pass through USB devices to the guest.
> Java and Kotlin as regular Android app already provide all the juice required to do the job.
To do what job? What job can Java and Kotlin do on a phone to replace a Hetzner VPS?
Keeping it plugged in like that is going to harm the battery.
That and the fact that my Xiaomi 12 runs USB 2.0 ....
I'm stuck running stuff on MacBook at night when I don't use it because of that.
Electronic price need to become reasonable again, a 8gb raspberry pi 5 is about 200€ right now. It's absolutely insane.
My pi 4 did cost me 40€ a few years ago.
i'm thinking more about long term.
Termux. Proot ubuntu. Serving OWUI and syncthing. Works fine. Had to remove pw protection so runit starts services at reboot. Therefore encrypted at rest any sensitive data
I Ssh in into it with ai harness to make changes. Made me not want to sell it after all.
I run kube in a VM, and it was ressource constrained.
I hooked an old samsung to my pc and have let claude handle everything.
I only had to unlock the bootloader, flash the image claude uploaded.
It rebuilt postmarketos multiples times to enable various flag, and now there is a kube node running on it.
In the end, I abandoned the idea and opted on purchasing an Android TV Box that has been installed with Raspbian for my home server. Sure it may not be as powerful as my phone, but it came with full blown USB that I can use to add any storage I want and it was cheap (today, it was approximately USD 11-12).
LXC does provide a security boundary, more than Docker but less than a full blown VM.
A real html page maxes the cores easily but it works. Keeping the pages open instead of closing / reopening tabs helps.
For reference I tried running it in a container inside my low end Synology DS223j but any container running, even doing nothing, ruins disk performances, weird bug.
> An open source anti-detect browser built for AI agents.
Sounds legit, circumventing the locks people are trying to add to their properties in an attempt to keep costs down and the service available for people that are actually trying to use the service rather than bots roaming around. Really happy for you that you managed to run ten of those. Not a nuisance at all.
> I wanted to turn [the old tablet] into a smooth, distraction free reading device, so I [got root and] transferred a few books with iFile and had exactly what I wanted [...] until the reading workflow left the book. \n\n Reading is not really offline anymore. Finding articles, checking references, downloading PDFs, opening links, and asking an LLM about something confusing all happen on the web. [...] [The ancient Safari engine can't provide] JavaScript, CSS, media, and browser APIs expected by current websites [and Apple didn't allow third-party engines.] [...] Since I was on vacation, this also felt like the perfect pet project to hand to an LLM and see how far it could get. Out of that experiment, Surf was born.
The point seems to have been getting iOS 6 to run a modern browser, which was experimentally built with an LLM, not so much have LLMs roam and masquerade themselves on the web via it
This article may come in handy when we look into it.
Extreme distributed regional servers - a dream come true.
The promise of the fabled lowest single tower latency, blast radius containment and resilient to the teeth has always been where my heart truly lies.
But it still seems too tedious/unreliable and most android devices cannot charge at the same time as using OTG even with special cables, needs kernel support
You can also just insert an SD card to do the write-heavy operations and swap it out when it wears out (you make backups, right? :D)
But the flash storage in a phone is going to be the eMMC variety which typically means a proper controller and TLC NAND. Not as robust probably as a "proper" SSD/NVME, but a lot more durable than most micro SD cards. And it'll have health indicators.
Also if your phone has SD card slot you can try "industrial" SD cards. They are small in capacity but ideal for write load.
For me I have a 24x7 running intel gen 11 nuc, i am lucky to get it installed with 32G memory and two NVME disks years ago and cannot image add those memory and NVME disk with much higher price
Just run some docker and VM on that, use cloudflare tunnel or tailscale funnel if you want to expose some service to the wild internet
I recently even find a pixel 5 and passthrough it to the openclaw VM on that nuc, let it browse the X to find some posts and relax me from doom scrolling, though seems timeline still bad and no much valued post find.
How hard would it be to combine the power of a couple phones for a server? I'm sure someone has done something like that already, right?
Would have been a fun load test had you hosted your blog on the phone.
Any efficient blog software will handle HN fine on anything that you can call a smartphone, so like maybe not a Nokia with GPRS and 6MB on-board storage, but something like the Samsung S2 from 2011 that basically shipped as a Linux box -- just need to adb push ARM binaries for PHP and a webserver
which one would people suggest? I got two already, both locked. Google Pixel 3a XL 64GB. Cannot root it, Sinlock. Not possible to install postmarktOS. Xiaomi Redmi Note 8T 64GB. Same thing, some lock, need a SIM card, some proprietary BS.. not possible to root easily.. I really wanted to build a cluster of mobiles, but its hard to find cheap and good mobiles for such experiment... and if i have to pay 500 buckets, i can do better with some Mini PC...
My goal is to reproduce the following experiment https://research.google/blog/a-low-carbon-computing-platform...
Instead, I got the same old post of someone serving web pages from a mobile phone :(
I asked Claude to set up Twilio (a phone API), and after setting everything up for it, now I can give my Claude agent instructions in any format where I have access to it (such as our Jira, by email, or when I'm writing a prompt by hand in Claude Code) to call me if it gets stuck or has any questions or needs anything from me, or to call me when it's finished.
so now I sometimes get to relax at home and get a call where Claude tells me that I have a Jira ticket that needs me to answer a question. so I open up Jira, answer the question, and go about my day.
(I also have it call me to remind me about appointments and requires me to press a key to acknowledge it, so that I definitely don't miss the call and confirm that I got it - I did this because I was constantly missing reminders and calendar appointments.)
So, while my server isn't literally a phone, it does call me when it has a question and to remind me about my appointments.
Running electricity through wood, arguably insane.
Or set the phone to never charge above 80%
Thta said, keeping it on charger for months will eventually swell its battery, true.
On top of that, pmOS can work with non-mainline Linux, so with the Android kernel, which is somewhat guaranteed to have support for most / all the devices. You shouldn't use it if you care about security though, but that's a different matter