The ’tar’ package no longer ships the helper scripts /usr/bin/backup and /usr/bin/restore.

If you rely on these tools, install them separately with: pacman -S tar-scripts. The core tar archiving utility itself is unaffected.

It’s tar-1.35-5

But I have just seen the word vomit (in slightly different context though), and I realised it’s so cool!

I’m still going to call my own local copy-pastas slop, not going to rename. But this vomit thing feels funny.

I mean, just why?! I don’t have the U.S. being in my dictionary or something. Not that I tore the country name that often too.

Sounds potentially very interesting. I was thinking of migrating from open ssh to some simpler implementation. I don’t really see any value in having password authentication, and I was postponing regenerating my RSA key with the Ed25519.

It implements hybrid post-quantum key exchange using the mlkem768x25519-sha256 algorithm, and curve25519-sha256 is supported for key exchange as a fallback algorithm for older clients. Ed25519 and ecdsa-sha2-nistp256 are the only supported public key algorithms, AES-128-GCM is the only supported encryption algorithm and SHA-256 is the only supported hash algorithm. It should be fairly easy to add other algorithms that are supported by the underlying backends, but I’d like to keep the set as small as possible to reduce attack surface.

This and Windows being out of scope, paint me interested!

Updating Arch Linux at the farm’s office PC.

Turned out, it’s

 rocm-llvm-2:7.2.4-2-x86_64                                               998.3 MiB 
 rocm-llvm-2:7.2.4-2-x86_64                                              1374.9 MiB

( 2/70) upgrading rocm-llvm
New optional dependencies for rocm-llvm
    rocm-openmp: OpenMP host runtime and GPU offloading

sudo pacman -Syu rocm-openmp

Packages (1) rocm-openmp-7.2.4-1

Total Download Size:    37.12 MiB
Total Installed Size:  131.84 MiB

Installed rocm-openmp, just in case. Have no idea whether I need it.

I consider doing that because Mac mini is awesome, and so are MacBooks!

I don’t know. It terrified me, a little. Not much, but still. Time surely flies!

If your forceis above 1 then it’s OK to have AI power up your force.

Say, 2.3 to the power of 3 is 12.169.

But if you’re a beginner and your force is bellow 1, power upping it makes things worse.

Say, 0.2 to the power of 3 is 0.008.

Top Laptops for use with FreeBSD

• HP Laptop 15-ef2xxx
• Lenovo Yoga 11e
• Dell System XPS L702X
• Dell Inspiron 7348
• Framework Laptop 16 (AMD Ryzen 7040 Series)
• ASUS TUF Gaming F15 (2023)
• HP ENVY 17 Notebook PC
• Dell XPS 9315
• Lenovo ThinkPad T430
• Lenovo ThinkPad E14 Gen 2

There are many more models,

• HP Laptop 15-ef2xxx
• Lenovo Yoga 11e
• Lenovo G40 70
• Dell System XPS L702X
• Lenovo IdeaPad 5 15ALC05
• Dell Inspiron 7348
• Framework Laptop 16 (AMD Ryzen 7040 Series)
• TUXEDO InfinityBook Pro AMD Gen9
• ASUS TUF Gaming F15 (2023)
• HP ENVY 17 Notebook PC
• Dell XPS 9315
• Lenovo ThinkPad T430
• Dell Latitude E4300
• HP 250 15.6 inch G9 Notebook PC
• Lenovo ThinkPad E14 Gen 2 Intel
• MacBook Pro 13 inch (2016)
• Lenovo ThinkPad E595
• Lenovo ThinkPad X13 Gen 4 (AMD)
• Lenovo ThinkPad T14s Gen 4 (AMD)
• Dell XPS 17 9700
• Lenovo ThinkBook G6
• Lenovo ThinkPad E15 Gen 2
• MSI GE60 2OE
• Framework Laptop 13 (AMD Ryzen 7040 Series)
• Lenovo ThinkPad X270
• Framework Laptop 13 (13th Gen Intel Core)
• Mechrevo JIAOLONG Dragon 16 Pro 2025
• Beelink SER8
• HP Compaq Presario CQ56 Notebook PC
• HP ProBook x360 435 G7
• Lenovo ThinkPad T14 Gen 2 (Intel)
• Dell Latitude E6410
• Dell XPS 17 9730
• MSI GE60 2OC
• Acer Aspire A315-24PT
• Dell Latitude E7440
• Lenovo Yoga Pro 7
• Dell Latitude 7490
• Dell Latitude 7400
• Microsoft Surface Book with Performance Base
• Lenovo ThinkPad T14 Gen 2 (AMD)
• Lenovo IdeaPad 1
• Lenovo ThinkPad T490
• Dell Inspiron 3543
• Dell Inspiron 5570

I’m not sure, but I think I’d love to have a FreeBSD laptop to play around with. Or maybe a desktop at home… We’ll see.

I use Arch Linux everywhere. And for using USE flags, I need Gentoo! I have never used it, and it’s too complex… or was too complex. Now, I feel I’m ready, but I have no device to run Gentoo, yet. I have my primary computer, running Arch. But…

I want to have FreeBSD, and I want to have Gentoo. For FreeBSD, I decided I’d buy a used Mac mini of Intel-era. Eyeing a 2010 server model for $30, which is a nice price, considering there are two 500 GB disks in there, and it’s in a working condition. That would work with FreeBSD no problem (and I don’t care about Wi-Fi and Bluetooth, I can ditch them even, or just disable them in software).

I want to have OpenBSD somewhere too. But I have no idea in mind right now, so I’d put it for later.

And for Gentoo… I have no idea where to start. I thought of getting a used (and probably faulty) MacBook Air M1 for that, but I’m holding off on that idea. Better next year. Or, ideally, when that laptop become obsolete.

Check out FUTO keyboard with the larger 250 MB model

Someone recommends.

Only wimps use tape backup. Real men just upload their important stuff on ftp and let the rest of the world mirror it.

— Linus Torvalds

Linux-kernel Mailing List, www.webcitation.org. July 20, 1996. 

• As a cherry on top, empty blog. I wonder who are the people who have nothing to say, then they post like one silly piece of word salad, then they don’t post for years again. I’d prefer reading someone who actually posts regularly, and is able toto communicate their mind clearly.

Here … I mean, why did you start so many projects, and why did you abandoned them? I too have an ADHD, like many others these days, I believe. Yet, for me, a GPT allows make progress when I’m stuck and don’t know where to start to continue working on my abandoned project. The massive improvement to me is the ability to communicate. To have my mind written, the problem clearly formulated. I have never paid for any subscription and I would never pay them. It would be less convenient when this shit won’t be given for free to folks like me. By that point, I expect I’d just run some less powerful local model. The value I get is from the ability to communicate and have feedback, sometimes (oftentimes) on topic. Yes, it’s slop. The longer the chat, the sloppier it is. But I usually have up to 10 messages conversations, and I use it only in my browser, for like an hour or two a day, tops. Most times, it’s much less than that. About 4 hours a week, summarised.

I usually hit the free limit with Claude and that’s where I stop. I almost never hit that limit with ChatGPT for some reason. Weirdly, Gemini was absolutely useless to me. I tried it thrice and never opened it again. I have no use for it.

I am not that good to juggle gazillion of projects and languages too. So I don’t. I work on one project, sometimes two or three, but they are highly related, they are almost the same, or solve the same thing from different angels.

I’d love to discuss it, but this particular piece has zero value. To me, there’s not much to discuss.

And, it’s irrelevant to my ADHD. Perhaps because I actually do something with it, and I have no Instagram, Facebook et al. I don’t doom scroll, so the dopamine is here only when I’m able to write some tiny helpful script with one prompt. When things go into some complex work, a GPT always produces incredible useless shit. So, I’ve learned there’s no point in using it that way. So, I use it with very small tasks only.

There screens simultaneously, I can only laugh at these delusional folks who believe they’re being efficient.

The entire OS runs from a Corsair MP600 MINI in a DockCase USB-C NVMe enclosure. No internal drive modification needed. When I want to game, I unplug the drive and reboot into my main Windows install, and game normally on the 4080. When I want to do LLM stuff, I plug the drive back in, reboot into NixOS, and both GPUs are available.

From I Put a Datacenter GPU in My Gaming PC for £200 article.

Please Do Not Vibe Fuck Up This Software #929

I didn’t expect to realise that rsync is maintained by one person, who decides to vibe code their own very popular project.

Some comments from my research on that issue.

Rsync has to be one of the worst spaghetti projects I’ve worked with. It’s an incredibly decent tool, but it’s code is an exact opposite of what you’d expect. And it’s written in C. I’m not surprised letting Claude loose on it for roughly 2 months already caused visible breakage. The question is, with it being very obviously a bad idea, can the maintainer still be trusted if he let something like this happen?

26k code changes in 2 months….. rsync was 67k LOC as of 236417c (latest not obviously vibecoded commit it seems?)

edit the file  /etc/zsh/zshenv , add the following lines to the file:

export XDG_DATA_HOME=$HOME/.local/share
export XDG_CONFIG_HOME=$HOME/.config
export XDG_STATE_HOME=$HOME/.local/state
export XDG_CACHE_HOME=$HOME/.cache
export ZDOTDIR=$HOME/.config/zsh 

You can then put .zshrc and other zsh config into the .config/zsh/ directory.

I won’t recommend on buying one though. It’s simply too expensive for just a board, but perhaps it’s cheap enough, if you consider getting the same modern hardware. Or not much more. It’s stylish and all.

Me, I’d go with any used PC you can get, it’ll deal with the whatever task gracefully enough, and does not have to look awful, be noisy and all that.

Not that fast though. It comes with conditions.

• She’d have her Lenovo ThinkPad,
• and she’d be required to write a report where exactly MacBook was better.

After which the management would decide whether the employees of the design team of Nestle are worthy of MacBooks.

LOL. Imagine what a company that is. That alone is humiliating enough.

Then, we have no idea which MacBook that would be. A Pro? Our bets are:

• MacBook Pro 13, the cheapest. The most sane option. It’s good enough, it has ports, and it’s pro enough.
• MacBook Air, also 13. Obviously. Those people won’t waste a cent on an employee’s comfort. Same thing in principal, but USB-C everything makes it a bit harder for an average Joe to deal with the laptop. They may not care, since the office packed with USB-C displays, so… This option is more likely, but hey, isn’t an Air is just 8 GB RAM laptop in its basic configuration?
• MacBook Neo. I would laugh my ass off if they’d actually buy a Neo! It would be as if the company decided to sabotage this whole migration thing for designers. Giving an average employee a month to accomodate is idiotic enough.

And one option I’m particularly proud of:

• They’d bring some stupid-ass old MacBook circa 2009 and say ‘hah, told ya!’

Idiots.

We’ll see what would that be. A sane company would bring a 16 MacBook Pro for a designer, at least. And wouldn’t think twice. For myself, for any serious work, I’d rather get a 16 MacBook Pro than anything else. Personally, I could live with an Air too. But that’s me. I wouldn’t just do most of serious work on a laptop, when I can have a desktop.

Air 15

Air 15 it is! With a 24 GB RAM, huh! Well, my apologies the folks who bought it, it’s a pretty decent computer. Hats off, guys. I was really scared it would be fucking Neo. But it wasn’t.

If you’re running Windows, always assume that if the US Authorities or Microsoft itself want to spy on you as an individual or on do a little industrial espionage on your company (which US agencies also do), they’ll just use a backdoor already present or at worse push an update to your machines(s) to create said backdoor.

Treat any and all software made by US companies as a foreign agent.

All the shit that the US Government and companies say about China, is pure Projection - the result of a mental process of “what would we do if we were the ones making those devices”. (And, yeah, China probably does that shit too)

If it ain’t Open Source, you got it as a binary or it can self-update, that software is somebody else’s agent and you’re trusting their ethics and goodwill when you have it running in your system outside a sandbox.

Link.

Basically that’s it.

• Is your hardware ten years old or more?
• Do you want a system made up of software that is on average 3 years old?
• Do you want absolutely ridiculous stability for the uptime memes?
• Are you a fan of the idea that every design decision should be done by a committee of theoretically democratically chosen developers but is actually just whoever wants the job because there is never any real transparency or motion about when the meetings are, much less when elections are?
• Does the idea of your operating system being compatible not because its good but because it’s just the largest base thanks to corporate investment make you moist?

Then pick Debian.

If you answered no to literally any of those options then go ahead and pick an Arch flavour, or Arch itself.

I’m not tolerating an Arch flavour but otherwise good comparison.

I wonder how many years into the future would it take me to have all this. Not just money wise, but also requirement wise. I don’t need much these days, but I’m slowly growing my hardware needs. I’m pretty sure I’d go with tiny efficient machines, like Raspberry Pi, Orange Pi, or alike. Mac mini too. I’m more like a Mac stadium guy, but I still think I’d have a rack or two at some point. I’m really curious when would that happen. Hundreds of thousands in servers, huh.

On the very opposite end, most of my servers are about tens of dollars each, no more. I believe I can cover most of my projects with servers like this. Even some professional projects! What a time to be alive! Many folks of the past were housing complex things on machine much slower than that!

“Microsoft’s ability to acquire successful companies and then destroy them needs to be studied. Today, we’re talking about GitHub.”

— Windowscentral.com topic sentence.

👌

i also only download flac files, and i keep them in my ~/music/lossless directory. i use picard to organize that, and wrote a bash script to keep a synchronized opus format copy in ~/music/lossy. on my phone i use termux/ssh to rsync the lossy files to my phone and avoid streaming altogether. for reference, my lossless directory is 221gb, and lossy is 19gb.

And

Same almost. I have an ~800gb main library of mostly lossless files that I squash to around 150gb by transcoding to 196k or something opus that i put locally on my phone. I also strip embedded cover art which can save a stupid amount of space sometimes; relying on folder hierarchy with cover.jpg/png files. (Bitrate is pretty overkill for me so I may drop it to 128-160…)

I haven’t had the time to manage the tags properly on my reference library*, but my folder hierarchy encodes artist/album/title with optional years and track numbers. I wrote a linter script to check the structure, that every folder has a cover art image, and to warn about lossy formats not in directories suffixed with [lossy] (purely for documentation purposes; not used in script logic).

My transcode script generates tags from the folder and filenames, only copying genre tags if they exist and stripping everything else. Lossless files are transcoded while structure, art, and lossy files are copied. Then that result is synced to my mobile devices. So whenever I add music my workflow is to just name file folders properly and download or extract art then I just lint, transcode, and then resync.

*(Tags of my reference library don’t matter so much to me, but the squashed lib needs consistent tags for mobile apps for behave as I intend)

The protocol Apple uses under the hood is AWDL (Apple Wireless Direct Link), which is a proprietary peer-to-peer layer that runs alongside your existing WiFi connection without dropping it. It uses a time-sliced channel-hopping mechanism so the radio can serve both infrastructure WiFi and the direct peer link simultaneously.

That’s the part that’s hard to replicate. LocalSend and most alternatives need an existing shared network because they’re just TCP/IP, they have no way to negotiate a direct radio link without OS-level support. Even Android’s QuickShare, which does peer-to-peer via WiFi Direct, drops your existing WiFi connection on older devices because the radio can only be associated with one BSS at a time.

The EU interoperability mandate lxgr mentions would theoretically require Apple to expose this, but AWDL interop would mean licensing or reverse-engineering some fairly deep radio scheduling logic, so I’d expect compliance via a different (probably slower) path.

— Some HN dude

It’s quite old, from 2011. The seller ignores me.

I don’t understand these kind of folks. They sell rubbish, yet they are not aware of it, are they? I message them, they read my message, yet they chose to ignore me. Why? Not that I write them to tell them they sell shit and they want too much for it. (Yet, it’s actually true.)

I have a collection of these folks. They sell something, and it’s pretty obvious they’d sell it another year or two. Yet, they refuse to even fucking reply in a fucking chat! Not that they did not read my message. They clearly did saw it.

The fuck are you posting your rubbish on these platforms then, huh?!

I litreally bought a mouse for $2 then some other idiot sells a broken mouse. Yet, the idiot failed to even reply to me. Some people, I pinged them in a week or two, asking similar question again, if what they sell is still on sell. Crickets. Yet, the ad is still active.

The fuck? What’s their logic?

Buying a ThinkPad means using a device made by Lenovo, a company which has repeatedly shipped devices infected with malware and backdoors (sometimes for profit) see: https://en.wikipedia.org/wiki/Lenovo#Security_and_privacy_in…

I hope it pays well, those Windows users are subsidizing some great hardware 😂

I would like to have one ☝️

One day.

Don’t forget to do vim keybindings in your shell, kids! Tell your parents to set -o vi

Oh, so much to learn!

Also,

firenvim may be of interest!

Also,

I toss in goyo.vim for when I’m writing prose.

Also,

There’s a similar plugin called zen-mode.nvimfor neovim, which has a few more bells and whistles like tmux handling and Wezterm support.

A great thread by the way.

It is a reply to this conversation.

The linked article is The Downfall and Enshittification of Microsoft in 2026

Person 1:

2025 wasn’t bad either. Retiring Windows 10, while requiring new hardware to upgrade during an economic downturn, and aggressively pushing unpopular features nobody asked for.

Whatever led to this sequence of floundering and cartoonishly stepping on rakes feels like it should be a case study.

Person 2:

Just pure greed. Never been in short supply at MS HQ, but this time short-term-ism has finally defeated the “long con” faction.

So, the Person 3 writes a truly great comment:

I don’t agree (having worked with the Windows team for a few years). There were several factors:

TPM 2 for Windows 11 was an important baseline for a load of security features that they were being hammered for not having. Apple and most Android phones had security coprocessors that did more than TPM 2 and so it was viewed as table stakes for security by a lot of government and corporate customers. The Microsoft sales org is focussed entirely on these kind of customers because of the company!s scale. Getting a $1M contract is barely worth a sales person’s time. The ones that move the dial in their org are at least tens, if not hundreds of millions of dollars. This means that they simply do not talk to SMEs, let alone individuals. They have a partner ecosystem to sell to these people but the feedback channels are not good.

They have an ingrained culture of lying to management. Partly this is incentives being misaligned, but there’s also fear of things leaking to journalists. No one wants a slide deck about how far behind a competitor they are to leak, so mid-level managers delete those slides when they present to senior leadership. This means that senior leadership has no idea how bad things are. They get a rose-tinted picture and reduce headcount in the org (and the mid-level manager gets promoted for making things good).

Windows is full of technical debt. The virtual memory subsystem alone is more code than a minimal Linux configuration. It has a load of copied and pasted code that made SQL Server marginally faster on one specific thing. Only three people in the company understand it (and one is semi-retired). Win32k is a massive attack surface that has vast numbers of things in the kernel that have no business being there. All of this happened for good reasons in the ‘90s but modern computers look nothing like the computers for which all of these optimisations were done and they hurt both maintainability and performance. Worse, most ‘anti-malware’ systems hook into these things and they can’t change them significantly without other vendors crying antitrust and claiming they did it to make Defender the only usable option. Apple isn’t a convicted monopolist, so was able to kick all third-party code out of the kernel. Windows still lets third parties run code in interrupt handlers (ever wondered why interactive performance is so bad on Windows? Take a look at how many things run in ISRs: every other OS moved to a policy of ‘ISRs wake threads, they don’t do anything directly’ twenty years ago).

A lot of the senior people on both the product and engineering sides don’t use competitors’ products at all, so they have no idea of what the competition looks like. They’ve been using Windows and Office for decades and believe it’s better. And they’re so used to Windows and Office that, if they do briefly try something else, the fact that it’s different makes it seem worse.

There are a bunch of cultural issues that reward shipping new features without any reference to user demand or how they integrate with other things. It’s not quite at the level of Google’s ‘get promoted for shipping a new chat system’ phase, but it’s quite bad.

But, first and foremost, senior leadership has no vision. I asked a bunch of senior folks what they thought computing should look like in 10-20 years. None of them had an answer. You can’t define a strategy unless you have an idea of what the strategy is aiming towards. And, because this comes from the top, no one with any kind of vision stays long enough to get promoted, with very few exceptions. It was noticeable that about half of the people I respected and enjoyed working with quit in the year before I did, and none of the ones I thought were a waste of space did (some got promoted).

There are more issues, and now that the stock price is tanking the board should be asking Nadella some hard questions.

While I truly appreciate some mindful pieces of wisdom, all I want to say is ‘Microsoft is full of idiots, lol!’ when I see any discussion about them. They’re like cartoonish villans, but also very stupid, like it’s some stupid movie.

Interesting thing is, the Person 3 says ‘hey, they aren’t all idiots!’ and then they tell how idiotic M$ is :)

Also, an interesting series of articles: How Microsoft Vaporized a Trillion Dollars. Inside the complacency and decisions that eroded trust in Azure—from a former Azure Core engineer.

Linux is just superior on Desktops now. (I mention desktop only because servers… only incompetent idiots run their servers on Windows. Plus a couple of very niche usecases where its necessary, I believe.) And now we have games. This user nicely summarised the situation:

And we probably shouldn’t discount the actual areas of improvement either. It used to be that for a large segment of the population, Windows could get almost infinitely bad before they would consider switching, because they play games and you just couldn’t really get games to work on Linux. The fact that the trade-off is no longer “I will not be able to play any games”, but instead “some games may be a bit more buggy or a few have anti-cheat which makes them not work but most just work out of the box”, is an immense achievement.

Ah, and the blog post itself, it’s good.

WHAT. THE. FUCK?!

Why? Fuck you!

No, really. WHY? I have this strong impression that Microsoft is just full of absolute idiots. I used to have MacBooks throught my entire life, and even this Dell laptop I’ve got at work recently, it has its trackpad centred. This trackpad right here… it’s off by millimeters.

Oh, what a piece of crap hardware!

Grep was produced in a night based on the source code of Ed.

Grep is in fact short for the ed command g//p shortened to g/re/p and then dropping the slashes.

— Some random stranger in a comment

I use Unity where I can and Xfce where I can’t

Oh. I like this phrase.

I use Sway WM where I can, and Gnome where I cannot.

• output power on|off|toggle

  Turns on or off the specified output.

  As opposed to the enable and disable commands, the output keeps its current workspaces and windows.

• output dpms on|off|toggle

  Deprecated. Alias for power.

I have no idea how long that was, just noticed the update. Well, thankfully nothing is broken, but I updated my configs.

So,

:%s/dpms/power/g                                                                    

it would be.

Firefox is ending support for Windows 7, 8 and 8.1; Mozilla recommend switching to Linux, for people who can’t run Windows 10 or 11

I would recommend Linux also for people who can run Windows 10 or 11.

That’s exactly the issue, and the bigger one. As compared to a ‘repairable’ phone. Like FairPhone et al.

Theoretically, a very first Google Pixel 1 could be upgraded software-wise with the newest Android, while the hardware can work longer. Heck, even my Google Nexus 6P and 5X could be upgraded! If not my beloved Google Nexus 4.

So, I’m not really interested in repairability (except the battery replacement, obviously) as much as in software longevity.

My iPhone 4S is still going strong, even despite numerous falls. The software though, it made it useless. I use one as a digital voice recorder sometimes, but that’s the only idea I had. It can shoot nice pictures too (in a bright daylight), and be a nice smartphone for a kid. (Because it’s very limited, and also small, and also cheap.) But the software made it useless.

Who loves hardware? Well, I do and this is my blog so I win.

Stolen from Nick Craver’s Stack Overflow: The Hardware - 2016 Edition

When I tell you how we do <something>, a few things happen:

• You might learn something cool you didn’t know about.
• We might learn we’re doing it wrong.
• We’ll both find a better way, together…and we share that too.
• It helps eliminate the perception that “the big boys” always do it right. No, we screw up too.

Do not expose anything without authentication.

And absolutely do not expose a folder with something like python -m http.server -b 0.0.0.0 8080 if you have .git in it, someone will help themselves to it immediately.

If you are aware of this, funnel works fine and is not insecure.

Tailscale IMHO failing in educating people about this danger. They do mention in on the docs, but I think it should be a big red warning when you start it, because people clearly does not realise this.

I took a quick look a while ago and watching just part of the CT firehose, I found 35 .git folders in 30 minutes.

No idea if there was anything sensitive I just did a HEAD check against .git/index if I recall.

https://infosec.exchange/@gnyman/115571998182819369

Ironically, Russia probing defences in Europe is functioning like Chaos Monkey — revealing vulnerabilities and triggering hardening.

Oh, Chaos Monkey, I do like this name!

Also, have never heard of Chaos engineering.

• The Pi 3B has 10/100 Ethernet, 802.11n (single-band) WiFi, Bluetooth 4.1. Power idled at 1.4W and peaked at 3.7W.

• The Pi 3B+ removed the 10/100 Ethernet in favor of USB Ethernet (~300Mbps w/USB2.0). CPU cores were overclocked from 1.2GHz to 1.4GHz (so a heatsink is more necessary), with ~15% increase in benchmark performance. It added 802.11ac (dual band) WiFi and Bluetooth 4.2 w/BLE. Power idled at 1.9W and peaked at 5.1W. This is also the only 3-model supporting PoE (w/ extra HAT).

• The Pi 3A+ removed Ethernet and reduced USB to a single port. The RAM was reduced from 1GB to 512MB. Power idled at 1.13W and peaked at 4.1W. The A+ form factor is more compact. Overall the 3A+ is smaller, cheaper, and less power draw than the 3B+ (but not as low as the 3B).

The lowest power draw with acceptable performance is the 3B. For slightly more power draw and more CPU performance, go with 3A+. For “everything” (including PoE) the 3B+ is it.

If you want the 3A+ but don’t need the video, want a smaller form factor, and half the power draw, the Pi Zero 2 W is it. Though the Pi Zero 2 W is supposed to be cheapest, due to demand it’s often sold out or more expensive. The 3A+ is still cheap (~$25) and available, with the downside of the higher power draw and larger form factor.

(disabling HDMI, LEDs, Wifi, Bluetooth, etc reduces power draw more. in testing, the 3A+ drew less power than the Zero 2 W with everything disabled. all of them draw ~0.1W when powered off)

Source

I recently worked for a company where using Windows was the norm (Linux is allowed, but be prepared to be on your own with company infrastructure). It felt ridiculous at first, then I decided to try and be more open to it, maybe I don’t know something. I even tried working on Windows for a week, but couldn’t stand it, everything just took so much more effort to get done and there were annoying ads on my lockscreen.

Three months in, I’m realizing my first feeling was right. Using Windows was just one of symptoms of poor culture, low standards for developer’s experience, neglect for code maintenance and devops. A few examples of what they considered normal:

• svn with a custom-made GUI Java tool to make PRs; it’s buggy, bugs are neither fixed, nor documented, each new dev just learns to work around them,
• culture of passing documentation orally; many projects with no READMEs, you’re just expected to ask around if you have questions,
• workstations are connected to internet over wifi, which acts funky couple of times a week, and you write in the group chat to ask someone to restart the router,
• legacy projects that are built with a spaghetti of Makefiles with quirky behaviors not documented anywhere,
• company file sharing is on Samba, use instructions can’t tell you what protocol version and settings to use, they just go though clicking steps in Windows GUI on how to access it, Linux users are on their own,
• instructions for setting up dev evironment going though GUI clicking steps, again no technical spec of what formats, versions and protocols we’re using — figure this out yourself (many Windows users end up doing thing under WSL anyway, but some system setting still require clicking).

To me, apart from just being technically inferior, Windows is a signal that company has a problem with culture. Now if I’m interviewing and they say they use Windows, I instantly walk out, not trying that again. I can accept if someone uses Windows due to external pressing circumstances, and we’re on the same page about it being a necessary evil, but it’s a huge red flag if the see nothing wrong with it.

And for those stupid enough to ask why, or those stupid enough to have arguments for the sake of arguments, the author continued:

You mean Linux kernel? Judging by the standards of C projects, its dev docs https://docs.kernel.org/ are excellent, there’s also books on it. Many opensource Linux tools I used have alright user docs and mans (nobody will use it without docs), they’re certainly better than no docs at all which is the case this company. I’m saying that a competent dev will leave docs to what he wrote, because he has a habit of doing it, and it costs him little. Even for something as small as a hobby project on Github. And I’m also saying that a competent dev won’t tolerate a lot of things that Windows user experience does to him (unless he’s paid extra or Windows is somehow inherent for the problem at hand).

The common link is neglect to and lack of awareness of developer’s convenience, good tools, and good maintenance practices. I’m claiming that this is causing both the choice of Windows and the things I listed. The Windows and docs are not linked directly, but there’s two links to analyze: poor culture → bad docs, poor culture → Windows. Which of the two are you skeptical of?

I admit that my point is not bulletproof solid, one needs to share my sentiment on Windows to a degree to buy the example, and it’s alright to not be convinced if you come from a different background. But I do think we may have a fun conversation if we try to find where the mismatch is.

Really, that’s a great write up, and I’d like it to be not far to link it, when I need it. Thanks, ph14nix.

Commercial OSes (both Windows and macOS) are also both American, and lots of people are trying to de-Americanize.

I’d rather de-Americanise it, you know.

• An unmodified iPhone SE (2022 model) with OS support until 2032. This runs all my authentication, banking, health, etc. It is in airplane mode 99% of the time unless I need it.
• The second is a Pixel 9a with Graphene OS for daily use, routing and internet access.

This is expensive, but I found it to be the only viable solution to this problem.

Well, I use older versions, but I think two phones is the best approach.

What’s the logic behind this? I cannot wrap my head around this.

• What about the need to physically drive to there to do something with the server? Cannot they hire someone to do the job? It’s just ridiculous.
• Why destroy the hardware? What a waste! Weren’t your disks encrypted? If they were, you rewrite everything a couple of times, and it’s impossible to recover. Destroying everything else just cries incompetence to me. The fuck were you having on your servers? No, really. What users data? The database of questions and comments? Aren’t they public already? Hashed passwords? What?

This feels like they’re cursed there, and their administration brain rot has filled everyone. They migrated to someone else’s data centre. Because what? Today’s price is cheaper, isn’t it? All that is just somewhat weird to me. Drop the five nines uptime requirement. Optimise the software stack to be more efficient and run on less servers, so you could manage spikes more efficiently. Sell half of your servers, if you predict you won’t ever need it.

I simply don’t understand that as of now. I may lack quite a bit of knowledge, that’s for sure. But I see that as some cultural thing, plus the red eyed crowd to somewhat forgetting we’re the minority. People in general won’t care if, say, stack overflow would go offline for a week. Will it though? Of course it won’t. So even going offline for a while for whatever reason is not the end of the world.

It feels like these too big to fail companies are simply unable to scale down when needed.

I won’t ever believe being in the ‘cloud’ is sane strategy for a company that can manage their servers by themselves.

Pomerium with Pocket ID.

I had the same difficulties with the next gen VPNs:

• high battery drain
• poor integration with the rest of the OS networking (DoH, firewalls, oher VPNs, etc.)
• challenges with extended family
• security concerns: a compromised program on a device has all the access that any program on the device needs. This is the usual BeyondCorp critique.

So I went looking for something BeyondCorp-esque:

• All entrypoints are authenticated through a small set of programs: I don’t trust random programs to implement auth correctly.
• Authentication to live as close to the backend as possible: the further it is, the more exposed that backend is to the rest of my infra.
• WebAuthn for everything interactive, with SSO.
• mTLS for non-interactive (OpenSSH, NFS, HTTP monitoring)
• Identity to be provided via OIDC, since Pocket ID is very simple, and there are decent hosted OIDC, which I’m open to (esp for social login).

I’ve mostly now migrated to:

• Pomerium on each of my servers, exposing all its HTTP/TCP services via HTTPS. It had a convenient free-tier hosted control plane, but I’ve since migrated to self-hosted.
• Pocket ID for SSO based on WebAuthn.

I like this because:

• Pomerium is designed for my usecase, making it easy to setup, and hard to get wrong.
• It’s easy for family.
• I don’t have to deal with the problems and limitations that next gen VPNs have.

It’s not perfect though:

• Some Android apps don’t support WebAuthn or mTLS, forcing me to use the website instead.
• Setting it up is more complex than the next gen VPNs.
• It doesn’t address lack of IP reachability. I’m fortunate for that not to be an issue.

Answer:

Its the kernel drivers, not firmware. There is no bios or acpi, so the kernel itself has to support a specifc board. In practice it means there is a dtb file that configures it and the actual drivers in the kernel.

Manufacturers hack it together, flash to device and publish the sources, but dont bother with upstreaming and move on.

Same story as android devices not having updates two years after release.

Source

It bundles things like a bar, launcher, and control centre that one normally has to assemble piece by piece in WMs like Niri and Sway.

Dank Material Shell on GitHub

Looks cool. I won’t ever use it (for myself), but I may recommend it to someone.

I never travel without my GL-AXT1800. Saved me so many times.

What is the benefit of this over, for example, an iPhone hotspot?

Run one wireguard server in your home and one client instance on this router and now all of your devices can share the same residential VPN connection. No fraud blocks or extra extra verifications from your banking apps, no million suspicious login detected from all your social accounts, use your home netflix account, etc. All without your individual devices running a VPN app.

The obvious competitor here is Tailscale. But let’s say, reasons, and Tailscale isn’t an option. Then you go down the path… TwinGate, Teleport, Netbird, Pomerium, Netmaker, ZeroTier, etc…

The logo is cool though!

Not that karma matters; except to say that it probably informs the reader if there’s a grey comment saying something; it’s a signal that this is not the correct opinion.

From lobsters.

As I’ve been saying recently:

I think we are well past the point where belief in a rules-based world order turns from optimism to delusion. Entrench yourself behind entire libraries of law books if you wish, it only makes for spectacular kindling.

To paraphrase Thucydides: the nuclear capable countries do what they want, and the non-nuclear countries suffer what they must.

Just a reminder they can car accident anyone anywhere from anywhere and having an old car will not save anyone when a Tesla with AI enabled cloud cameras is approaching head on.

Oh, that reminds me of the film by Obamas.

A while ago I had a discussion with my friends that it is possible that in the future if 5G is sufficiently cheap, smart tvs come with a 5G SIM so they can force ads and updates even if you refuse to connect it to WiFi. I wonder if this will ever be a real thing. Either 5G, 6G or whatever comes next.

Sure thing, I don’t want this in my room. But the idea is quite good.

And this idea is even better.

I fear this won’t even required SIM cards. I’m worried that Apple’s Find My and Amazon’s Sidewalk networks are the precursors of this: They’re effectively company controlled p2p networks that lets the company use their customers’ internet access points like a commodity. If one customer refuses to give a device access to the internet, they could use that network to route it through the access point of another customer.

Also, personal experience: My own ISP (in Germany) experimented with some similar stuff a few years ago: They mandated use of their own home routers where only they had root access. At some point, they pushed an OTA update that made the router announce a second Wifi network in addition to the customer’s. This was meant as a public hotspot that people walking down the street could connect to after installing an app from the ISP and buying a ticket.

The customer that “owned” the router wasn’t charged for that traffic and the hotspot was isolated from the LAN (or at least the ISP promised that), but it still felt intrusive to just repurpose a device sitting in my living room as “public” infrastructure.

(The ISP initially wanted to do this on an “opt-out” basis, which caused a public uproar thankfully. I think eventually they switched to opt-in and then scrapped the idea entirely.)

remember the “S” in “IoT” stands for “security”

Haha, you’ve got me!

It’s taking a cloud-based product, de-clouding it, and then connecting it to your own private ‘cloud’. Pretty cool all things told.

[0] https://kazlauskas.me/entries/tailscale-doesnt-quite-suck

[1] https://tailscale.com/blog/tailscale-sucks

I don’t get the appeal of Tailscale for simple homelab use. I have OpenVPN and it’s trivial. Hit the toggle and I’m connected, no fuss.

I’d prefer to have WireGuard, which is better in a lot of ways compared to OpenVPN. It’s far more flexible, secure, configurable and efficient.

So I’m assuming the person means just regular simple WireGuard vs Tailscale.

Tailscale (and similar services) is an abstraction on top of Wireguard. This gives you a few benefits:

1. You get a mesh network out of the box without having to keep track of Wireguard peers. It saves a bunch of work once you’re beyond the ~5 node range.

2. You can quickly share access to your network with others — think family & friends.

3. You have the ability to easily define fine grained connectivity policies. For example, machines in the “untrusted” group cannot reach machines in the “trusted” group.

4. It “just works”. No need to worry about NAT or port forwarding, especially when dealing with devices in your home network.

The common folklore is just FUD. The main issue is deliverability to the likes of Google, Microsoft, Yahoo, etc. You need a clean fixed IP in non-residential block and a sufficiently aged domain or your mail will be flagged as spam or rejected. Alternatively, you can use a relay service for outbound email. Besides the deliverability issue, hosting email is fairly trivial from a technical standpoint; on Linux, the standard utilities are Postfix, Dovecot and OpenDKIM. The server is for my own use, so I don’t even bother with spam and AV filters.

Even if you can’t send email at all (unlikely if you use an outbound relay), there are very significant privacy benefits to having your own server. I send very few emails relative to the number I receive. You couldn’t pay me enough to go back to one of big commercial providers.

The CI situation on codeberg is definitely their weakest link; IIRC it doesn’t even let you SSH into your builds to debug. I was very happy to discover https://yojo.emersion.fr/ which could allow you to easily build your codeberg projects on sourcehut, which has top-tier CI.

Keep in mind that when flash memory is read, you don’t get back 0 or 1. You get back (roughly) a floating point value — so you might get back 0.1, or 0.8. There’s extensive code in SSD controllers to reassemble/error correct/compensate for that, and LDPC-ish encoding schemes.

Modern controllers have a good idea how healthy the flash is. They will move data around to compensate for weakness. They’re doing far more to detect and correct errors than a file system ever will, at least at the single-device level.

It’s hard to get away from the basic question, though — when is the data going to go “poof!” and disappear? That is when your restore system will be tested.

I have my root on a RAID-0 set of 3 SSDs. I plan to move it to 6 SSDs, I just need reshuffle my disks for that.

  sudo pv -X /dev/sda

or even just:

  sudo cat /dev/sda >/dev/null

and it’s pretty inefficient if the device doesn’t actually have much data, because it also reads (and discards) empty space.

for copy-on-write filesystems that store checksums along with the data, you can request proper integrity checks and also get the nicely formatted report about how well that went.

for btrfs:

  sudo btrfs scrub start -B /

or zfs:

  sudo zpool scrub -a -w

for classic (non-copy-on-write) filesystems that mostly consist of empty space I sometimes do this:

  sudo tar -cf - / | cat >/dev/null

the cat and redirection to /dev/null is necessary because GNU tar contains an optimization that doesn’t actually read anything when it detects /dev/null as the target.

On macOS / Linux you can use dd to “copy” everything from /dev/yourssd to /dev/null. Just be careful not to do it the other way! https://www.man7.org/linux/man-pages/man1/dd.1.html I have no idea if forcing a read is good / the right way. I’m just answering how to do it.

/dev/null is “empty” if you read it, its /dev/zero you dont want to dd onto your disk.

About 10-15 years ago, the scourge I was fighting was social media monitoring services, companies paid by big brands to watch sentiment across forums and other online communities. I was running a very popular and completely free (and ad-free) discussion forum in my spare time, and their scraping was irritating for two reasons. First, they were monetising my community when I wasn’t. Second, their crawlers would hit the servers as hard as they could, creating real load issues. I kept having to beg our hosting sponsor for more capacity.

Once I figured out what was happening, I blocked their user agent. Within a week they were scraping with a generic one. I blocked their IP range; a week later they were back on a different range. So I built a filter that would pseudo-randomly[0] inject company names[1] into forum posts. Then any time I re-identified[2] their bot, I enabled that filter for their requests.

The scraping stopped within two days and never came back.

[0] Random but deterministic based on post ID, so the injected text stayed consistent.

[1] I collated a list of around 100 major consumer brands, plus every company name the monitoring services proudly listed as clients on their own websites.

[2] This was back around 2009 or so, so things weren’t nearly as sophisticated as they are today, both in terms of bots and anti-bot strategies. One of the most effective tools I remember deploying back then was analysis of all HTTP headers. Bots would spoof a browser UA, but almost none would get the full header set right, things like Accept-Encoding or Accept-Language were either absent, or static strings that didn’t exactly match what the real browser would ever send.

— Source (original)

Thank you very much for the observation about headers. I just looked closer at the bot traffic I’m currently receiving on my small fediverse server and noticed that it’s user agents of old Chrome versions but also that the Accept-Language header is never set, which is indeed something that no real Chromium browser would do. So I added a rule to my nginx config to return a 403 to these requests. The amount of these per second seems to have started declining.

There was thig feature (more a bug to me) when I was scrolling through the long email I have written, and I did not notice I paste some shit I constantly select with my mouse. That’s the habit of mine, and it helps me concentrate on the reading. The pasting happens when you press the mouse wheel. I wasn’t actually meaning to press it, rather scrolling. And my mouse is perfectly fine, and in a perfectly healthy condition! If it was broken, the shit would be pasted much more often. I know it by experience, because before this mouse, I had this broken mouse.

‘Don’t use it’ they say.

FUCKING WHAT?! You can select the shit not meaning it, and then you can middle-click not meaning. And not even noticing! That happened to me. Often, when I was editing CSS for my website, and obviously that was broken. But that wasn’t a big issue. I was accidentally pasting shit into the layout partials for my website, and while there was no issue purely because I just noticed that at some point.

But also, I was writing a long email, and I was pasting some irrelevant shit in there, and did not notice it, because the whole block of text looked very similar. A block of text I meant to send, and another block of text I did not mean to send.

That’s a fucking security issue. Here’s the solution for sway, thankfully that was implemented.

# idiotic feature when middle-click (the scroll wheel) mouse button (usually button-2) pastes things

primary_selection disabled

Here’s a thing though, you need to reboot. Or re-enter sway, if you will. Reloading the config won’t help. But that’s a minor annoyance for a desktop. Otherwise, I have no idea how to battle with this shit. Last time I’ve heard, it’s disabled in Wayland-only sessions like Gnome. And I assume that’s treated as an obsolete junk. Really, who ever needs that? What for? There’s cmd+c (if you’re lucky to have Apple keyboard or a keyboard with cmd, and you know how to make X86Copy, ’Paste, and ’Cut work.

set $mod Mod4

bindsym --to-code $mod+z exec wtype -P XF86Undo
bindsym --to-code $mod+Shift+z exec wtype -P XF86Redo
bindsym --to-code $mod+x exec wtype -P XF86Cut
bindsym --to-code $mod+c exec wtype -P XF86Copy
bindsym --to-code $mod+v exec wtype -P XF86Paste

Where Mod4 is the left CMD key.

verbosity=1
bootlogo=false
console=both
disp_mode=1920x1080p60
overlay_prefix=sun8i-h3
overlays=usbhost2 usbhost3 tve
rootdev=UUID=b4aabcac-142e-30f4-b70f-c216082688e0
rootfstype=ext4
usbstoragequirks=0x2537:0x1066:u,0x2537:0x1068:u

While after the removal it looks like this:

verbosity=1
bootlogo=false
console=both
overlay_prefix=sun8i-h3
overlays=usbhost2 usbhost3
rootdev=UUID=0b51a721-639c-41e1-ac3f-c33c4b9d1533
rootfstype=ext4
extraargs=video=HDMI-A-1:d video=Composite-1:d
usbstoragequirks=0x2537:0x1066:u,0x2537:0x1068:u

.\setup.exe /product server /auto upgrade /EULA accept /migratedrivers all /ShowOOBE none /Compat IgnoreWarning /Telemetry Disable

I’ve used this to upgrade 10 to 11 on non approved hardware, going back to at least 2nd Gen. Intel CPUs. I’ve used it to upgrade existing Pro, EDU and IOT that didn’t want to upgrade.

The install window will say server but it isn’t.

I use Xiqi printer for barcode printing on a self-adhesive paper. It’s about $11 only, battery powered, light, bluetooth.

— Someone on the internet.

CMD will open

Type (no quotes) “net user Prefferedusername /add” (replacing Prefferedusername with the user name you wish to use) and press enter.

Next type “net localgroup administrators Prefferedusername/add” and press enter.

Next type “ net user Prefferedusername/active:yes” and press enter.

Next type “net user Prefferedusername/expires:never” and press enter.

Next type “net user administrator /active:no” and press enter.

Next type “net user defaultUser0 /delete” (this is case sensitive make sure the “U” is capitalized) and press enter.

Next type “regedit” and press enter.

This opens registry editor navigate to “HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\OOBE”

Delete DefaultAccountAction, DefaultAccountSAMName, and DefaultAccountSID

Right click on LaunchUserOOBE and rename it to SkipMachineOOBE and make sure the value is set to 1.

Close registry editor and type shutdown /r /t 0

— Source

I remember there was just one release that worked. But I don’t remember how it went through. All other times, I had to refer to Fedora’s documentation.

Yesterday’s evening, I pressed Download Upgrade, yet it hung, the interface. I needed to leave it for like 15 minutes or so, and it showed that it downloaded the upgrade. Then, I pressed ‘Restart to upgrade’ (or whatever the name of the button is), and it did nothing. Precisely, it asked for my password, but did nothing afterwards. I pressed it like 5 times, entered my password 5 times, and then 5 times of nothing. Then I noticed the entire interface is fucking hung. I said ‘fuck it,’ reached to my laptop, sshed into the computer and systemctl poweroffed it.

This morning, I repeated the same thing, and nothing really happened. Nothing. This afternoon, I decdided I’m going to update it via ssh, as always. There were new updates (regular updates) and the upgrade too. I pressed both buttons. It was super long to download regular packages updates, but it downloaded the upgrade to Fedora 43. I pressed Reboot, expecting nothing, as always. Yet this time it fucking worked, and it prompted me with the question whether I’d like to reboot or not. Well, ‘yes’ I said. So here we are, this thing is being updated now.

It’s been about 10 or 15 minutes now, and it’s at 46% now. That’s with the regular SATA-3 SSD and

A word to the wise: Do not set any cron jobs for 2:00 am or 3:00 am on Sunday morning! Or to be safe, on other mornings besides Sunday as well, since jobs originally set to run on some particular day may eventually be changed to run on another day, or every day.

Most of the time such cron jobs will run fine, but if they run every Sunday morning, then twice per year they will run at the exact time daylight savings time (aka summer time) kicks in or ends, sometimes with very strange results.

On Linux with vixie-cron we saw two cron jobs run something like once per second between 3:00 and 3:01 when the most recent daylight savings time began. Thus they ran about 60 times, stepping all over each other and making a noisy mess in email. No serious harm was done, but that’s only because they were not tasks capable of causing serious harm.

Feel free to wish for or agitate for or fund or write a better open source job scheduler that everyone will use, one that will ensure no overlapping runs, allow specifying time limits, etc. Better tools exist, but until one of them achieves cron’s level of ubiquity, we have to live with cron at least some places and sometimes.

Alternatively, where possible set the server timezone to UTC so that no daylight savings changes will happen at all.

Or most preferable: Governments of the world, stop the twice-yearly dance of daylight saving time altogether.

But in the meantime this particular problem can be entirely avoided by just not scheduling any cron jobs to run on Sunday morning at 2:00 or 3:00 server time.

— Source

I start my jobs at 03:50, I don’t know why I chose this particular time.

My current setup is Fedora Silverblue, and an Arch container for development. Silverblue as a base system is rock solid, but it being immutable means you need a container, Nix or something else.

Fedora as a development environment is also fine, but I ran into a few issues with packages either not being available or being really out of date. The bigger underlying issue here is Fedora’s dumb packaging requirements (= every dependency must be a package, even if that doesn’t make sense), resulting in (at least I suspect) many people just not bothering packaging their project for Fedora.

An Arch container means you can work around this by, well, using Arch. Since it’s not your base system the usual issues with Arch (e.g. it not always being stable) don’t matter much, because in the worst case you rebuild your container (something I’ve automated down to a simple make arch). And Arch being Arch means pretty much every package is available, either as an official package or an AUR package.

My main gripe with Silverblue is that running development environments inside a container is a bit wonky, and the tooling surrounding it is a hit and miss. For example, Toolbox is generally more stable/of better quality compared to Distrobox, but lacks features that seem pretty important (e.g. generating .desktop files).

And the whole thread is just golden!

This is so great way of formulating this:

Beyond dissing Hansson, which is admittedly why I clicked, this turned out to be a neat laundry list of what not to do when writing orchestration scripts. Worth reading from that perspective alone.

Aaaaaannndd

there’s a huge grey area separating a noob installing Fedora and a noob installing Omarchy.

… called brain 😂

I would like to point out that it is entirely Apple’s choice that their operating system does not run on my new Lenovo ThinkCentre. I’m not saying it was a bad business decision — they are a hardware company, after all — but it is particularly galling to realize that if I bought a new Mac, I would be subsidizing the development of an operating system that contains code whose sole purpose is to lock me into a specific hardware platform. I realize that most people don’t look at it that way, but there it is.

What interesting about this citation is that it is from the year 2006, almost twenty years ago. When Steve Jobs was still alive, and would be for 5 more years. The very first iPhone would be invented (demonstrated to the public) in a year.

David Heinemeier Hansson has a problem with Windows as a programming platform.

— Jeff Atwood at 26 Feb 2008

Pro tip: If you’ve arrived here via a link aggregator, feel free to skip ahead to the Summary for a conveniently digestible tl;dr that spares you all the tedious details, yet still provides enough ammunition to trash-talk this post in the comments of whatever platform you stumbled upon it.

I like it, and thought of stealing it. The phrase. However, it feels like I cannot be bothered to truly give a — what’s the polite word? — I mean, I don’t mind others thrashing on my blog posts. I’m too long on the internet.

The original post — A Word on Omarchy — is quite good. Flagged on this idiotic hacker news (search the page for the word ‘flagged’), obviously. Not even surprised.

Omwhat?

The name Omarchy appears to be a portmanteau of Arch, the Linux distribution that Hansson’s configuration is based upon, and お任せ, which translates to omakase and means to leave something up to someone else (任せる, makaseru, to entrust). When ordering omakase in a restaurant, you’re leaving it up to the chef to serve you whatever they think is best. Oma(kase) + (A)rch + y is supposedly where the name comes from.

This was a fantastic overview of what Omarchy is, and I’m so happy I was able to fully understand its details! I did not expect it, truly! But hey, I’m using Arch for over six years now, almost daily.

Almost, because due to being a father of two kids, sometimes there are days I don’t touch my computer, at all.

This is just funny!

To all evidence, the creator of Omarchy has no such concern. They were all-in on Windows last year, then a couple months later it was Omakub. About a year later, it was this thing. It seems there’s a good chance they’ll be on to the next new shiny thing before they have to worry much about maintaining this one.

I tried making it my default browser because of:

Set as default browser BOOST CHATGPT LIMITS Unlock 7 days of extended limits on messaging, file uploads, data analysis, and image generation on ChatGPT Atlas.

You miss the most questionable bit which is asking for keychain access. I said no to that one.

A browser using your keychain seems like the least questionable bit, if anything.

Right, but most browsers aren’t owned by money-losing startups desperate for any bit of training data they can get their hands on as scaling taps out.

I really doubt OpenAI consciously wants my passwords, but I could absolutely see a poorly-coded (or vibe-coded, lol) OpenAI process somehow getting my keychain into their training set anyway, and then somebody being able to ask Chat-GPT 6, “hey, what’s Analemma_’s gmail password?” and it happily supplying it. The dismal state of LLM scraper behavior and its support (or lack thereof) of adherence to best practices lends credibility to this.

— Source.

I think Chrome is going to be vibe-coded as well, pretty soon. LOL.

For those who hasn’t seen it, just dropping down the control center consumes 14 watts of CPU/GPU power as demonstrated by various YouTube videos.

Yeah, I can imagine! My iPhone 12 mini wastes the battery so much quicker!

In big engineering programs engineers are paid bounties for every kg they remove from the design. We need software developer bounties for removing CPU cycles and memory.

In the car industry, this is the case because there is a direct incentive for the company to make a vehicle lighter.

A single kg is enormous likely saving tens of dollars per car. Which sounds like nothing, but if a million of these cars are made the cost saving are in the tens of millions.

What does a software company save by their software running 10% faster on user hardware? Exactly nothing. In the case of Apple they even have some incentives to make their software worse to get people to buy new devices.

The curse of software engineering has always been that there is very rarely a reward for your software being better. Software companies mostly make it on features and “good enough” stability.

The tragedy I see these days is that building scalable services is the height of tech fashion, and every engineer wants a fashionable CV that will help her get the next job. Nobody has ever gotten a job with a CV that said “I don’t know AWS and Kubernetes, but I know how to fulfill all your SLAs for a $100/mo infra bill,” because every hiring manager has stopped reading before the comma, and that engineer has starved to death.

— Source: Why is everything so scalable?

Oh, yes, and I’m that guy! I’m over six years (at least) into this, and I have no idea about AWS and Kubernetes so far. Well, I know k8s, but mostly theoretically, and I have no real opportunity to master that knowledge. Same with AWS. And I’m on that side that does not think it’s an issue. I’m pretty good with computers though. My servers

There were zero Android phones in 2007. The G1 launched in 2008 as a brand new platform with no third party apps. In 2007, Windows Mobile was the popular platform for people who liked loading their own software on their phone.

If you were around for the early Android scene you might have heard of XDA Developers — which was named for the O2 XDA Windows Mobile phone.

— Source

Oh wow, I didn’t know about this XDA Windows Mobile phone. I bet these days XDA the community is much stronger brand than that Windows phone.

…

And so I seriously dislike the experience of using a Mac with Liquid Glass. The UI has become the star, but the drunken star, blurry, illegible, and physically unstable. It makes making things way more of a struggle than it used to be.

We had pretty good Mac UI, but Apple took the bad parts of it — the translucency and blurriness already there — and dialed it way up and called it content-centric. But it seems to me the opposite. Liquid Glass is Liquid-Glass-centric.

Oh, and I’m so happy I’m a Linux user now!

For those not on latest macOS and not on Apple Silicon but on a somewhat recent Intel Mac, there’s still ‘sudo fdesetup authrestart’ (check with ‘fdesetup supportsauthrestart’ before whether your Mac supports that). Then you can reboot the machine without being physically present by providing unlocking credentials prior to reboot.

Thanks!

I have been using Gentoo for close to 13 years now. What really helps with the long compile times (I don’t have the fastest computers; my current one is a Ryzen-APU-based one from 2019) is to run emerges with low priority in the background. Out of sight, out of mind.

You can do that by setting

PORTAGE_SCHEDULING_POLICY="idle"

in /etc/portage/make.conf. This tells portage to run the emerge processes with an extremely low priority, and updates more or less trickle down in the background without disrupting system performance noticeably. You can also set other portage scheduling policies.

— Source

Cool, nice to know!

You see, it’d be much easier not to be such a nerd. When I set up a computer for someone else, it’s Windows or LMDE or Fedora. It’s such a no-brainer. The relationship with a computer can be simple: like/dislike, works/doesn’t work, can fix/can’t fix. One can actually accept that something can’t be fixed or changed! Really admirable. But that’s not me. I actually care what is the DHCP daemon, who’s resolving the DNS queries and how aggressive the OOM killer is going to be.

Haha! Yes, indeed! It’d be much easier not to be such a nerd!

I think I cannot escape that path myself. I do care what’s my DHCP daemon either.

Source: Half an year on Alpine: just musl aside

• Can be restricted to specific devices
• Must be available under GPLv3, including anti-tivoization provisions (forced bootloader unlock)
• May not attempt to use TPMs, DRM, or other systems to support assertions about client devices Not part of the hardware:
• May only interact with hardware through public, documented, APIs in the “part of hardware” category
• Using alternatives from competitors must be fully supported
• When made by a company that also makes hardware, must also work on competitors’ hardware (at least one, more if technically feasible)
• May be under a proprietary license
• Must not attempt to assert anything regarding the hardware, so things like Google Safteynet are now illegal. Security boundary must be shifted to consider client devices insecure This is, I think, a good compromise to allow software developers to get paid without taking away ownership of hardware devices. Developers can be paid for “part of the hardware” software with money from selling the hardware, and “not part of the hardware” software can be trivially commercialized under a proprietary license. But, there is no way for a user to end up unable to control their hardware, or incentivized to configure it in a specific way.

This is the reason the 3-2-1 rule for backing up exists.

They might be singing this song now. (To the tune of ‘Yesterday’ from the Beatles).

Yesterday,
All those backups seemed a waste of pay.
Now my database has gone away.
Oh I believe in yesterday.

Suddenly,
There’s not half the files there used to be,
And there’s a deadline
hanging over me.
The system crashed so suddenly.

I pushed something wrong
What it was I could not say.
Now my data’s gone
and I long for yesterday-ay-ay-ay.

Yesterday,
The need for back-ups seemed so far away.
Thought all my data was here to stay,
Now I believe in yesterday.

• https://github.com/vinceliuice/MacTahoe-gtk-theme
• https://github.com/vinceliuice/WhiteSur-gtk-theme

They look pretty good!

Bugs aside bcachefs is the most advanced filesystem I am aware of. It really delivered on that promise of “just add all of your disks, then dump your data on and it does the right thing.”

For example: Do you have disks of different sizes? That’s fine just works. Do you have disks of different speeds? That’s fine it will measure them and read from the faster ones. Or you can configure to write to the fast disks first then migrate them to the slower disks later. You can also configure to migrate to the faster disks during read.

On top of that all of the settings were configurable at the file or directory level. So if you have one directory that is huge but replaceable just mark it as unreplicated. You have one directory that needs consistent great performance? Mark it as always stored on the fast disks (unless they run out of space of course). Do you have a directory of videos that don’t need faster storage? Mark them as HDD only. Or maybe write to SDD first for fast writes but then migrate to HDD as you don’t need to read them very fast.

And again, all of these settings are configurable at runtime. No need for different partitions, datasets or whatever. No need for different mountpoints for each different service directory. Just tag the directory with the settings you want and you are good. Change your mind? No problem, just update the directory settings. You get a lot of advantages from having a single filesystem for example you can hardlink or reflink between any directories on the system, no worrying if they happen to be different datasets. You also get whole-system atomic snapshots.

Most people compare it to ZFS but I don’t think it really compares. ZFS is really just RAID + a CoW filesystem. Sure, there is good integration for things like read errors but fundamentally there isn’t really anything that it does that RAID can’t do (the ZFS AnyRaid project may change this). You basicaslly build up your vdevs using regular RAID patterns then put a filesystem on top of it. bcachefs is a step further in flexibility.

That being said a lot of these differences matter most for more “home” use. For larger scale you typically provision a server with a full set of matching drives and you more or less know what you are using it for. In that case ZFS really isn’t missing much functionality and of course it is rock solid.

Short, and in simple words! Kudos, Kevin!

What a fucking idiotic name for a software, and what a fucking correct it is, it’s literally the worst interface I’ve ever seen on any platform. Change this idiotic name, improve the interface, and you’d get a very strong graphical editor not much worse (if at all) than Photoshop, you dorks.

And playing with a graphical editor in Linux, I realised I don’t care whether it is cross-platform or not. Previously, when I used proprietary operating systems, I treated any software as whether it’s worth my time to invest my resources into it, by evaluating whether it is available on other operating systems. That cross-platformness, it allowed jumping ships. I heavily used Adobe Creative Suite on Windows, but then I switched to macOS. And upon switching, nothing bad happened, I could use the software I mastered.

These days, being a Linux user, I take pleasure in the system being open-source, and the entire software stack being open-source too. Or, free and open-source, as it’s commonly named. Also, there is this name, libre software.

I assume the same would apply for FreeBSD and other BSDs too, but I just have very little experience with them. However, enough to know I can switch my system for FreeBSD pretty easily. There’s some difference, and I’d need to learn some new concepts (jails?), but overall, I can transfer all the experience I have.

Today, I realised that on Linux, I just don’t care about Windows or macOS any more! Will Gimp work on Windows or macOS? I know it will, but I don’t care. Same with Krita, Blender, my browser of choice (I use Firefox), my window manager (I love sway), my graphical file manager (I adore Nautilus), my video player (mpv), my terminal emulator (foot). I don’t care whether they are available on Windows or macOS. I’d like them to work on FreeBSD, because I consider this system to become my main desktop system one day, but made not for 100% of my machines. But also I know that most of Linux software works on FreeBSD.

Technically, I see no perspectives with Linux becoming hostile. I assume we’d be able to recompile our own kernel without bloat when needed. If needed. That’s why I mostly don’t care whether the software I use is available on other platforms. While I can imagine giving up and returning to macOS on a laptop (only because of their MacBooks), I cannot imagine a scenario I’d use Windows ever again in my life.

And so, if some software is heavily dependent on, say, systemd, like Gnome is, I have no issues with that. Sad there’s no support for FreeBSD now (and alpine, and some other systemd-less systems), but it’s still no Windows.

I’m willing to study Blender, and I think It’s available on Windows and macOS. But what’s the point? If you need maximum resources for 3D, why would you ever consider using Windows for that? Feels incredibly stupid. I can understand macOS, because of their hardware. But again, in heavy workloads, I expect a loaded Threadripper paired with a dedicated powerful GPU to utilise the software much better.

I wonder, will dropping these other platforms simplify software development for Linux?

If there was one thing clear from the Snowden leaks, it was that the NSA really lamented the rise of encrypted web traffic back in the late 2000s and early 2010s. Wouldn’t it have been very convenient for the NSA to have a TLS termination point covering the majority of the web?

I think Cloudflare is just that, or at least a part of it (PRISM). Cloudflare sits in the US, so they can be legally surveiled/overtaken by the NSA, and it would be a punishable crime for anyone working at Cloudflare and knowing about this to speak up about it (FISA laws).

Worried about DDoS attacks? Most providers include that in their VPS offerings. For instance, Hetzner does. It’s not at Cloudflare’s scale, but you most likely won’t need that level of protection anyway (you need to be a viable target in the first place).

Cloudflare is just obnoxious. I can’t believe how many well-read, technical people use it.

Now, as I think of it, it feels pretty real, this version.

Also:

And everything is free! Or almost everything. How convenient.

This also reminds me of this very unclear and sad story: https://www.wired.com/story/lee-holloway-devastating-decline-brilliant-young-coder/

it’s a way to self-host your website from home, without revealing your home IP address.

Wouldn’t Tor hidden services be better for that?

Hmm, I think I’d like to explore that. Just for the sake of it.

Also:

Have you ever seen a hidden service submitted to lobste.rs?

If they were supported by mainstream browsers then yeah, they would be great for that. I particularly like that since they’re bound to a public key[1], it’s impossible e.g. for a scalper to take over the domain if you stop paying for it. You actually don’t have to pay rent for a domain at all - you can actually own an onion address, unlike normal domains.

[1] Sadly, right now this means you need to have quite a lot of trust in the server you host your website on. It needs access to the private keys, and if someone steals them, they can impersonate you - and AFAIK they aren’t revocable. The protocol does support offline keys, but tor doesn’t yet support them on the server side (tor#29054, not sure about Arti).

That being said I don’t think I’ve heard of this being an actual issue for someone.

Also:

One of the issues I have with discussions about Cloudflare is that Cloudflare has been good at cultivating a following with certain technical types to the point where it could be said that they have a small army of defenders.

Try to discuss the kinds of things in this post on Reddit or on the orange site, and more often than not things will quickly devolve in to emotional exchanges with downvotes that have nothing to do with engaging or facts. What’s the point in having discussions if fans who don’t care about facts can simply downvote discussions until they’re not seen? I don’t value popularity contests — I like facts.

Add to this that Cloudflare seem to use certain sites for marketing purposes, and a great case can be made for avoiding those certain sites in favor of ones where actual discussion matters more than fanboying over a wanna-be monopoly.

That’s not to say that it doesn’t happen here — I’ve played around with it here — but it happens a lot less. Thank you, lobste.rs people :)

Note that /bin/bash isn’t guaranteed to exist, even if Bash is installed. /usr/bin/env is guaranteed to exist on POSIX systems, so /usr/bin/env bash is the generally recommended way to invoke it in a shebang. /bin/sh is guaranteed to exist on a POSIX system, but /usr/bin/env sh can be used for consistency.

Yeah, that’s what I noticed.

And in a fun twist, we’ve confirmed that it’s the exact same cell found in Apple’s MagSafe battery pack. You can swap between them and the phone still boots up just fine. Like a rear-mounted spare tire on an SUV, an iPhone Air with a MagSafe battery pack is ready for an on-the-go swap, if you will. Granted it’ll take a bit more than a tire iron to make it happen.

— Source

The FLX1s from Furi Labs runs a fully optimized Linux system called FuriOS, packing a lightning fast user interface, 3 hardware switches for microphone, camera and modem/gps, and a privacy centric approach like no other.

FuriLabs is ready to protect your data and keep you connected and secure at all times. Long term support and an unmatched price point. Get your hands on one today.

Dimensions & Weights:
Phone: 170mm x 76mm x 8mm : 201g
Box: 180mm x 90mm x 28mm : 120g
Total: 180mm x 90mm x 28mm : 321g

See our shipping details.

Read some reviews, also in the below tabs, or some external reviews.

This cost us hours and hours of our lives, causing especially my wife a significant amount of stress, during an already very difficult time in our lives (which I won’t get into). I’m seething with rage towards Microsoft and its utter incompetence and maliciousness. Let me, for once, not mince words here: Windows 11 is a travesty, a loose collection of dark patterns and incompetence, run by people who have zero interest in lovingly crafting an operating system they can be proud of. Windows has become a vessel for subscriptions and ads, and cannot reasonably be considered anything other than a massive pile of user-hostile dark patterns designed to extract data, ad time, and subscription money from its users.

If you can switch away and ditch Windows, you should. The ship is burning, and there’s nobody left to put out the fires.

Couldn’t agree more.

Don’t trust Microsoft to make good decisions in your best interest.

I was charging my phone 35…65% all the time, I even have my special automations for that. Sometimes I needed some relaxed recharges during the day, it was super effortless. When I see the phone is at 35%, I put it on charge for like 15 to 30 minutes (a slow charge), and it’s closer to 50% or even more. And it’s usable for almost the entire day. I don’t use my iPhone that heavily. It’s mostly messenger, or me reading some web, or me working with my to-do system in Obsidian (so, reading or writing text). All that was perfectly good on my 1st Gen SE, yet it’s screen isn’t that good. Performance-wise, I could use an iPhone X from 2017 (mine one from 2020), which is now-obsolete. And probably I’d like to downgrade, but since my iPhone is perfectly good, except the battery, I’m not going to.

This phone felt like it is from the future. The infinite battery changes a lot. My battery was infinite, and when Nick was giving the phone to me, he was smirking that charging it four times a day is my problem now. I replied in all seriousness that I’m not going to, and that was the truth. I was chrging that phone twice a day with 15 minutes sessions of slow-charge, literally. The battery was infinite, and each use was a couple of percents, I couldn’t do 60% → 50% unless I was using it very heavily. Not to say 65% → 35% (I have notifications for both). That travel would take half a day for me.

Today? Yeah, I can take it with 65% and put it back on charge with 35% within one session, which in my case isn’t very long usually. Like, half an hour tops?

That’s iOS 26 with all its fanciness. I understand I should pay for this. But the thing is, I don’t even like all these glassy things. They are nice and fancy and all, but the previous version wasn’t bad either! It was more stylish than this. But nobody asked me. I’m not some luddite in that regard, and I’m very open to many things. And I’m open to this glassy super-distraction (from the interface alone!) too, even when I clearly see it’s too much for me, it just spams and DDOSes my attention muscle. Which isn’t that strong.

I have an impression that my Intel MacBook Pro 13" (Late 2014) with battery at 60% health is much more stable with its battery. It works very confidently for at least 3 hours of heavy usage. Last time I was in the office a couple of weeks back, I didn’t take my charger (expecting to leave the office after two hours tops), and it survived 3 hours easily. I didn’t use my laptop that heavily for maybe half a year, yet it survived this.

Now, I’m like Nick. I charge my phone 3 to 4 times a day, and that charge is 20% → 80% now. Sometimes even 20% → 100% now (I can simply forget to remove the charger). Now these two Belkin dock stations Nick gave me are of a use, finally! They were just a fancy paperweights on my tables, at farm and at home. Now, I actually use them to re-charge my phone!

Cool, ain’t?

Same goes with Solomia. She’s a super light phone user, and her phone surely survived a day or two. Now, I’m paying attention, I see her phone is constantly closer to 20%, so it triggers me to go put it on charge, when I see it. She’s having iPhone 11, which is regular size. So, that’s not some tiny-phone-tiny-battery-plebs issue, it’s the system. The system.

Now, I feel this teleportation back in time. Not even into my SE1 years, not even my 5S ones, but straight into my first iPhone 4S. Oh, remember that iOS 6 → iOS 7 update? That was it, that was all the same. This time, I expect I won’t be given this privilege to update to iOS 9 equivalent. I expect our phones won’t be supported next year, and when there are at least some performance patches, we won’t get it. As how would they nudge us into buying new iPhones? These are perfect for us already.

Only I started value this little size of mine much more. I started using it as my organiser with the collection of my (mostly to-do) notes. What a blessing to have this super-computer in my pocket. Sorry thing, it was perfect before the update, and now it’s just a fancy toy that’s not too reliable, when I need it.

So, for the future, it’s obviously only the biggest possible size, since its battery is also bigger. Or… maybe Apple will fix this, at some point.

Jacob Kauffmannjacobgkau

Jun 27

As mentioned over in Feedback on Collaboration for Mission-Critical Work, Framasoft is maintaining a soft fork called Mostlymatter that removes the arbitrary user limits.

• The downloads are here: https://packages.framasoft.org/projects/mostlymatter/
• The patch’s source code is here: https://framagit.org/framasoft/framateam/mostlymatter

System76 migrated to Mostlymatter in late February for the Pop!_OS community chat, and it seems to be working well. (As explained in that other thread, we do have far fewer simultaneous users than registered users, and while I expect many servers are like that, YMMV based on your own server loads.)

I was cancelling my annual slack premium last month and had to click to acknowledge that some of my members are using the AI features and they will lose access to them.

They then offered me a discount and if I refused there was another checkbox where I accepted that I was about to cause disruption for other staff.

I was tempted to take the deal until that point, but I’m the only member of the organisation and I absolutely do not use their AI

In my experience, software developers love automation. You can pry my CI/CD server from my cold, dead hands!

But what exactly do LLMs really automate? First of all, they are loosely specified and cannot be trusted to produce the same outputs given the same inputs. Second, they are expensive to run at scale, if you want to use them for anything beyond a demo or toy. Third, they are in constant flux. What works today may be broken tomorrow when the underlying model is updated.

For automation, I much prefer technology that is reproducible, well specified, cheap and efficient.

Yeah! Me too! Especially I love cheap for that, solely because that scales really well!

# Disable Power LED after boot
dtparam=pwr_led_activelow=off
# Disable SD card activity led
dtparam=act_led_trigger=none
dtparam=act_led_activelow=off
# Disable the ethernet LEDs - these are Pi4 specific values.
# Look in the docs for the values for other Pi boards.
dtparam=eth_led0=4
dtparam=eth_led1=4

to /boot/config.txt

I have been checking it all day long, but it isn’t there. Wasn’t there. Now it’s here, as I’m checking now.

Apparently, American employers woke up! Thanks guys, I’m going to test it right now.

I mean, I know that for quite a long time, a decade maybe. But it’s only now that I realised that. Particularly, this comment made me think over it just once again:

Given that both grep and find are weird/inconsistent between BSD/GNU versions, and I typically use them piped together for the same things anyway, I’ve found that ripgrep is a nice/faster/universal alternative that is pretty unproblematic to install in whatever environment I want: https://github.com/BurntSushi/ripgrep

Followed by the joke:

But isnt that what AI is for? Writing syntactically correct regexes?

Also, I realised that I don’t have to use built-in tools, on my servers I can install everything I want. I guess I would rewrite my scripts over time, over and over again.

Random people on the internet aren’t fans either:

From my experience in anger, it’s only really Debian (and for very different reasons, Windows, of course). Arch, Fedora, etc. don’t seem to have this much random breakage version to version that is ignored and NOTABUGged.

And then as the reply:

That’s because everyone else ships the software more or less as released and tested upstream. Debian messes with EVERYTHING, splits it all up, then has to do their own testing. It’s their packaging philosophy, it’s intentional, and it creates a lot of work for sysadmins. They do it in the name of ultimate system control but if that’s what you’re after you can probably get a better product out of arch or something without the constant breakage and hair-brained defaults.

I guess it’s hare-brained¹, and I’m agree so much here!

And the first commenter replies:

Yeah if one is after ultimate control then they should probably use, in this particular order, Guix SD, NixOS, Gentoo, Void, Arch. I’m just fine with Arch for my home messing around server, and RHEL at work. I just don’t get what’s the appeal of ubuntu/debian for pet servers.

Me, I am experienced with Arch for 6 years now, using it daily. And Debian for maybe 15…17 years, using it on and off, and it was the worst all the time. It’s only advantage was that for some mysterious to me reason, it supports some obscure hardware, like my Orange Pi Zeros. I’d love to have Arch there, or even FreeBSD, but there are no options like that.

• Your secret key can be stored in Yubikey, handled by a dedicated OpenPGP agent. This allows deriving a strong key from a weak one. Your password is basically a short PIN with max 3 tries. Every password retrieval can require a physical touch. This is convenient and secure!

Pass makes sense if you use it with a hardware key, with touch enabled. With this setup, it’s hard to beat its security.

• It uses public key cryptography, and comes with its advantages. You don’t need your master password to add/encrypt passwords. You only need that for decryption. Less exposure of master key, and more convenience.

  For that reason, it’s well suited to share passwords with other people or devices. You can encrypt to multiple public keys. This adds multi user and device support.

  You can easily add a backup offline public key (which you may print) if you lose your Yubikey.

• You can decrypt a single password without decrypting and exposing other passwords. The passwords are isolated, if you use Yubikey.

• Searching passwords is quick and transparent. You easily see what is in your store.

• You can use it programmatically, eg, your backup script can grab a password from the store.

• It’s a short bash script that you can verify, and delegates encryption to a dedicated well-audited cryptographic tool.

• PGP is a standard, and GPG and git are widely available. There is no database to break or migrate. You can read your passwords anywhere and in the future.

• The script is written by the creator of the acclaimed Wireguard!

There are also cons.

• Some people don’t like that it leaks metadata (filenames, and password tree), though there are versions of pass that fix it.

• Lately gpg is causing some troubles with Debian Trixie. GPG agent frequently locks the Yubikey and requires restarting pcscd (probably due to conflicts with pcscd). There is a similar tool Passage using Age, maybe that solves it.

• There are mobile apps, but they are not as frequently updated as something like Bitwarden apps (which has client for every OS, and frequently fixes bugs and adds functionality).

• I haven’t used and not sure how good browser support is.

Here is a post on a similar password management with GPG replaced with Age

https://words.filippo.io/passage/

When a phone costs north of 1000USD something is terribly wrong. 99% of “modern” phone use is basically (doom) scrolling, browsing the web, using core apps (like maps) for directions, taking pictures and finally communicating with whatsapp and making the odd/rare old school phone call. This should not require spending 1000-1500USD on a phone.

Im doing all of the above with a iPhone SE for what i paid like 300-350USD for.

Second hand phones are even cheaper, just change the battery and you are good to go.

I’m rocking 12 mini, and I’m happy with it.

And also, as a bonus:

It’s wild to see the HN crowd, bleeding-edge technologists, regularly bring up “lying flat on a table” as a critical feature for a supercomputer inside a camera that fits in your pocket. Somebody (many somebodies?) is rolling over in his grave.

That’s HN, my man. Don’t expect them to be any smart, huh.

It’s just ridiculous. I use smartphones for like my entire life, literally. And never ever I needed to use them when they lie flat on a table. What is wrong with these people? I have my phone in my hands, always. Otherwise it’s in my pocket. Or bag. That’s it.

I like new iPhones. I won’t buy them, they’re too expensive for me. Not only because of their price, I have no real use for them. That’s it. I’ll continue being second-hand guy when it comes to iPhones. I bought few new ones, and I just regret it. Money being wasted. They better be put into any work, money.

To securely access your Mattermost server from the Android and iOS apps, which do not support self-signed TLS certificates, you can setup a reverse web proxy.

The main benefits of a proxy are:

• SSL termination
• HTTP to HTTPS redirect
• Port mapping 80 to 8065
• Standard request logs

Proxying can be achieved with most web servers.

According to specs.

Hardware requirements for team deployments

Most small to medium Mattermost team deployments can be supported on a single server with the following specifications based on registered users:

• 1 - 1,000 users - 1 vCPU/cores, 2 GB RAM

• 1,000 - 2,000 users - 2 vCPUs/cores, 4 GB RAM

Campfire

According to their promo page:

So, I can make a conclusion that Mattermost is much more efficient. While being packed with much more features.

However, since I have an old Intel Atom server with 2 GB of ram and 2 CPUs, I can try both, considering both of my organisations are less than 250 users.

Niri v25.05 (the previous release) added a feature called dynamic screencast target where you can bind a key that changes the target of your screencast to the focused window or screen. It’s very slick and easy to setup.

Ha, sounds cool! I’m curious how it works and how would you use it.

Free and open-source encrypted proxy project, shadowsocks.org

Has Durov did the most typical Russian thing? Did he stole the logo for his Telegram?

The ecological and societal costs of data centers are hidden from the FAANG companies. It’s very important to be well informed about it so society can regulate it. This podcast series, “Data Vampires” is really informative about the subject: https://www.youtube.com/playlist?list=PLm-sqZXTqq9oIG_d0P7aT…

You can find it in your favorite podcast player. Everybody should listen to it.

Okay, I will.

If you want lots of nvme, lots of sata, lots of ram and lots of ability to power drives then you’re not looking at a minipc anymore.

Would suggest you do the same I did — climb onto ebay and look at the AM4 systems that gamers are dumping. AM4 is aging out of gaming space but for NAS/proxmox duty they’re aging like fine wine. Look for one that has is X570 and has bifuraction support on the X16

Ahh, I might want to explore that one day too.

I set up one to run Frigate [1] to detect motion over my several security cameras and send me notification emails with still images and video clips attached. It works well, and I hate the idea of sending my private videos to the cloud for processing with usual security camera setups.

[1] https://github.com/blakeblackshear/frigate

Oh, I like the idea. Might want make it real one day.

The clock problem is actually a Linux issue. Linux sets the hardware system clock to UTC and only applies time zones when displaying the time.

Windows sets the hardware clock to local time.

Every time one or the other updates the clock, it’s now in the wrong format for the other OS. The fix is to tell Linux to use local time. There are no side effects as far as I can tell.

Oh wow, okay. Hope I’ll never need this knowledge though.

For all the Stallman incarnates out there, I understand that Linux is a kernel and GNU corelibs and userspace all unite in some unholy ceremony to make a usable operating system.

For the bulk of the userbase though, this ultimately doesn’t matter. Language also has the neat capability to evolve based on its accepted understanding. Everyone knows what is meant when an operating system is referred to as “Linux” and that’s what matters.

That’s a good one, I’m going to save it.

• I am unable to get asleep, and that’s a bad news
• And I realised I don’t need it, which also is a bad news.

The interface is a wonder. But comparing to 5.9 MB iSH this thing is 224.3 MB.

I need to ssh into my servers once in a very long while. And most times it’s so typical that I can just create a shortcut, which would be much more convenient and efficient.

That makes me sad, and angry too. Also, I don’t like that it’s made by a corporation. It’s worse than being too useful. Really, I cannot imagine a scenario one needs such a useful ssh into servers tool. It’s super fancy, and a great example of a very modern interface. Outstanding job.

• web.archive.org https://archive.is/o2rJa

Someone says:

Huge red flag. What extraordinary thing did Canonical accomplish by the way? Making a Debian distro for first year uni students?

And this reply is quite good:

What extraordinary thing did Canonical accomplish by the way

I’m not sure you remember what the Linux landscape looked like when Ubuntu was first released, but it was a pretty huge deal at the time. So much more hardware and software (esp. media playback) worked out of the box than any other distro at the time. USB and wireless networking just worked. The desktop they provided was was really nice and well configured as well with most defaults working the way most people expecting them to. Plus the pretty amazing fact that they would send you an install CD, for free, basically anywhere in the world. Back when downloading 700 MB over the internet was an impossibility for most people, this was a seriously huge deal for getting Linux into the hands of people. You could take that free CD, pop it into almost any computers, just accept all the defaults and be pretty sure to have a working Linux desktop at the end. At the time, that was a huge accomplishment.

We can all have opinions about the choices Canonical has made over the years and what they have become, but let’s not pretend that they weren’t one of the biggest and most significant drivers of Linux on the desktop back in the day.

— Source

Them I’m going to just cite some things I found interesting:

SuSE and other mainstream distros worked “out-of-the box” just as well, or better, than the early Ubuntu versions. Fedora, Debian, etc didn’t not work because they couldn’t, but they were trying to make a point about free software and what that meant, so they took a strict stance on included licensed code. It was 2 minutes of adding RPMForge to your repos to fix, however.

The early versions of Ubuntu weren’t popular because they were easier, they were popular because they followed the AOL model and would send you an install disc for free. They leaned into making the install process easier much later (around the 6…7 versions).

That’s about CDs:

The one real reason they were popular and the one real accomplishment. No doubt.

I agree here. The CDs were a popularity hack. I have some of them, or maybe I had. They were in a foreign country I’m not living in, and have no plans to ever visit.

I can understand a hiring process like this if we’re talking about fighter pilots, astronauts, special forces operators, scientists going to antarctica, or whatever. But software engineering? C’mon. Delusions of grandeur.

Well.. have you met their CEO? The truth is once companies have a certain size and revenue, their software developers don’t matter really that much. Yes it can be hell to work there, everyone is overworked, things are low quality, etc. As long as the checks keep coming in ( and they will ) there is really no critical problem to be addressed. Now the funny part: once the alarm bells go off and they try to change course… not happens, the “ship” is unresponsive and in most cases it will sink. But depending on the size of the company that will take years and maybe even decades. So the best skill everyone needs to have (I wish really technical introverted software developers had this more ) is: Develop “street smarts” and tell these people to go fuck themselves and shove their “hiring process” up their asses.

— Source

• 1:45 am, when it’s the night
• 8:45 am, when it’s the morning, and I’m somewhere in the kindergarten with the kids
• 12:45 pm, when I’m in the Kdg to get my daughter
• 17:45, when I’m in the Kdg again, to get my son
• 23:45, when I’m more likely to be asleep, so the reboot won’t affect me, and just in case

It’s an OpenWrt router, so the commands are:

crontab -e

crontab -l
45 1,8,12,17,23 * * * sleep 120 && touch /etc/banner && reboot

service cron restart

https://www.tomkranz.com/blog1/a-case-study-in-bad-hiring-practice-and-how-to-fix-it

And it’s so fucking funny! It’s exactly my experience of applying for jobs these days.

I never liked Ubuntu, and I believe they are harm to Linux community. Their system is an utter piece of shit, and if it wasn’t, I might be attracted to full switch to Linux much earlier.

Fedora is just a fantastic jewel, and surely one of the best Linux distributions alive.

I’m not very sure about Arch (which I use), but it was rock solid for me (solider than any Debian ever was). Also, there are Gentoo, Nix, and Alpine, that I want to try out.

Reading this, it feels like they are absolute idiots. They stink with mediocrity, even their font is just, I cannot stand it, there’s something very unprofessional about it to me.

So, yeah. Not even surprised their hiring processes are like that.

Also this

Canonical: the recruitment process really is that long/complex/you want how much info about high school?!

Four minutes and a half, to be precise.

Turned out, it never had a time sync thing enabled. Weird, should I do that manually? I cannot recall me enabling this on some other machine. Yet, all other machines have it enabled.

timedatectl

               Local time: Wed 2025-08-13 10:05:08 EEST
           Universal time: Wed 2025-08-13 07:05:08 UTC
                 RTC time: Wed 2025-08-13 07:05:08
                Time zone: Europe/Kyiv (EEST, +0300)
System clock synchronized: no
              NTP service: inactive
          RTC in local TZ: no


systemctl status systemd-timesyncd

○ systemd-timesyncd.service - Network Time Synchronization
     Loaded: loaded (/usr/lib/systemd/system/systemd-timesyncd.service; disabled; preset: enabled)
     Active: inactive (dead)
       Docs: man:systemd-timesyncd.service(8)

All it took, is just:

sudo systemctl enable --now systemd-timesyncd

And now it lives in the present:

Created symlink '/etc/systemd/system/dbus-org.freedesktop.timesync1.service' → '/usr/lib/systemd/system/systemd-timesyncd.service'.
Created symlink '/etc/systemd/system/sysinit.target.wants/systemd-timesyncd.service' → '/usr/lib/systemd/system/systemd-timesyncd.service'.

timedatectl status
               Local time: Wed 2025-08-13 10:03:26 EEST
           Universal time: Wed 2025-08-13 07:03:26 UTC
                 RTC time: Wed 2025-08-13 07:03:26
                Time zone: Europe/Kyiv (EEST, +0300)
System clock synchronized: yes
              NTP service: active
          RTC in local TZ: no

I wonder whether it influenced anything. This server being living in the future.

E.g. how would it resolve syncthing conflicts? Its files are always newer, right?

Got me laughing, the irony of all this. Perhaps, that’s a good thing, the open sourcing the whole thing.

But honestly, are there people who actually use Windows 11 these days? I understand the percentage, but still, it feels like everyone is on macOS or Linux. Windows feels so obsolete, so clumsy.

I worked with it from my wife’s with issued laptop. And it’s so unbelievably bad!

• I renamed a file, and it never showed that in file explorer from where I renamed the file. Until I pressed the refresh button. Oh, this is very futuristic thing to have.
• Then, turned out, I cannot select files and make a directory with them.
• When I manually created a directory and moved the files there, turned out the change wasn’t reflected in the other window that, turned out, had the same directory opened. Mind blowing.

Really, I used the system for like 3 minutes and I faced so many absolutely incredible stupid things. It’s hard to believe people are actually using the system. It’s just miserable.

Always has been.

amber-lang is dope, it requires this and doesn’t have insane syntax for it and compiles down to bash. 

Lots of other nice features that make it easy to do things you should always be doing but aren’t

Oh wow! Today I learned. Thanks, dude!

There’s also quite a few good reviews on Zorin on YouTube.

People should also note Zorin sells a “pro” version for around $50. I’m sure most people could achieve the same features the pro edition has without much trouble, but it also helps them with development costs and everything else.

https://zorin.com/os/pro/

Since the context is systemd, one really should look to Unices in the 1980s rather than to Microsoft Windows in the 1980s. INI files were popularized by DOS+Windows, where there was no formal specification and whatever the Win16 API happened to do was gospel, but were largely nonexistent as a file format on contemporary Unices. Far from being “the” configuration file format, they were barely a blip on the radar and their introduction into Linux was somewhat akin to an invasive species as the Unix world had plenty of file formats and tooling for such formats before then, and the Linux-based operating system world had been busy cloning them all for years.

Things like rc files for shells and .mailrc for BSD Mail were simply scripts written in the program’s own command language. The .newsrc used by ReadNews was an example of a common colon-separated key+value format, also seen in Xresources and others. The sendmail.cf used by BSD Sendmail was its own very special thing, indistinguishable from modem line noise.

BSD also widely used the “capabilities” file format, which still exists in many parts of FreeBSD, NetBSD, and OpenBSD to this day for various things such /etc/login.conf and /etc/gettytab and termcap. (In many cases nowadays, cap files are compiled into Berkeley DB or other constant database files by tools such as cap_mkdb and tic.)

Many configuration files in the Unix world were simply flat file tables, with colons, whitespace, or TAB as the field separator. phones(5) is one such TAB-separated table that is in FreeBSD to this day.

The Sony VAIO SVE111 is a compact and lightweight laptop featuring an 11.6-inch display with a resolution of 1366 x 768. It is powered by an AMD E2-1800 dual-core processor, includes 4GB of RAM, and has integrated AMD Radeon HD 7340M graphics, making it suitable for basic tasks and multimedia use.

The Sony VAIO SVE111 is a part of the VAIO E series, designed to be a portable and stylish laptop suitable for everyday use. Here are some detailed specifications and features:

Design and Build

• Form Factor: Ultraportable laptop with a sleek design.
• Dimensions: Approximately 29.4 cm x 20.5 cm x 2.1 cm (11.6 inches x 8.1 inches x 0.83 inches).
• Weight: Around 1.3 kg (2.87 lbs), making it easy to carry.

Display

• Screen Size: 11.6 inches.
• Resolution: 1366 x 768 pixels, providing decent clarity for everyday tasks.
• Panel Type: LED-backlit display, which offers good brightness and color reproduction.

Performance

• Processor: AMD E2-1800 dual-core processor, which is suitable for basic computing tasks such as web browsing, document editing, and media playback.
• RAM: 4GB DDR3, which is adequate for multitasking with light applications.
• Storage: Typically comes with a 500GB HDD, providing ample space for files and applications.

Graphics

• Integrated Graphics: AMD Radeon HD 7340M, which can handle basic gaming and video playback but is not suitable for demanding graphics tasks.

Connectivity

• Ports:
  • USB 3.0 and USB 2.0 ports for connecting peripherals.
  • HDMI output for connecting to external displays.
  • Ethernet port for wired internet connections.
  • Headphone and microphone jacks.
• Wireless: Wi-Fi (802.11 b/g/n) and Bluetooth support for wireless connectivity.

Battery Life

• Battery: Typically offers around 4 to 6 hours of battery life, depending on usage, which is standard for laptops in this category.

Operating System

• OS: Originally shipped with Windows 8, but can be upgraded to later versions of Windows.

Additional Features

• Keyboard: Chiclet-style keyboard with a comfortable typing experience.
• Touchpad: Multi-touch capable touchpad for easy navigation.
• Audio: Integrated speakers with decent sound quality for casual listening.

Use Cases

The VAIO SVE111 is suitable for students, professionals, and casual users who need a lightweight laptop for tasks like web browsing, document editing, and media consumption. However, it may struggle with more demanding applications such as video editing or high-end gaming.

Overall, the Sony VAIO SVE111 is a solid choice for users looking for a portable and budget-friendly laptop for everyday tasks.

The Sony VAIO SVE111 is equipped with the AMD E2-1800 processor, which is part of AMD’s low-power, budget-friendly APU (Accelerated Processing Unit) lineup. Here are the key details about the AMD E2-1800 processor:

Specifications

• Architecture: Based on the Bobcat architecture, which is designed for low power consumption and efficiency.
• Cores: Dual-core processor, meaning it has two physical cores that can handle tasks simultaneously.
• Threads: Does not support simultaneous multithreading (SMT), so it can handle two threads at a time (one per core).
• Base Clock Speed: 1.7 GHz, which is the speed at which the processor operates under normal conditions.
• Turbo Boost: Can boost up to 2.4 GHz under certain conditions, providing a temporary performance increase for demanding tasks.
• Cache: 1 MB of L2 cache, which helps improve performance by storing frequently accessed data closer to the processor.

Graphics

• Integrated GPU: The E2-1800 features an integrated AMD Radeon HD 7340 graphics processor, which is suitable for basic graphics tasks, such as video playback and light gaming.
• DirectX Support: Supports DirectX 11, allowing for better graphics performance in compatible applications and games.

Power Consumption

• TDP (Thermal Design Power): The E2-1800 has a TDP of 18 watts, making it energy-efficient and suitable for use in ultraportable laptops. This low power consumption contributes to longer battery life.

Performance

• Target Use Cases: The E2-1800 is designed for entry-level laptops and is suitable for basic computing tasks such as web browsing, word processing, and media consumption. It is not intended for heavy multitasking or demanding applications like video editing or high-end gaming.
• Benchmark Performance: In benchmark tests, the E2-1800 performs adequately for its class, but it may struggle with more intensive applications compared to higher-end processors.

Conclusion

The AMD E2-1800 processor is a budget-friendly option that provides sufficient performance for everyday tasks in lightweight laptops like the Sony VAIO SVE111. While it may not be suitable for demanding workloads, it excels in energy efficiency and is a good choice for users who need a portable device for basic computing needs.

Looks like I’m obsessed with an ultra-portable laptop for basic tasks, and I don’t know how to fight that obsession. Because, honestly, I have everything I need right now. My MacBook Pro 13 weights 1.54 kg, and that’s tolerable. Yes, I have to take a charger with it, but that’s not really an issue. Yet, this thing — an ultra-portable and cheap laptop — haunts me.

I don’t even need to have it all the time on me, I don’t have any use case for that. Yet, here we are.

Is anyone else getting this VLC clash while updating to most recent bits?

Yeah, me too!

Just run

dnf update -x 'vlc*'

That would update everything except vlc.

Or:

Executing

dnf remove vlc-plugins-freeworld-3.0.21-4.fc42.x86_64

solved my transaction issue.

Also, there were multiple rsync processes from root. I couldn’t understand what’s up.

The temp was 75 ºC and I believe it was that low — and not 95 ºC as it was after a brief cpu-intensive session — only because I set the CPU to powersave from the default ondemand by

echo "powersave" | sudo tee /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor

This modifies current behaviour, which would return back to ondemand after a reboot. For permanent change, add echo "powersave" > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor to /etc/rc.local

I didn’t know how to mitigate the issue. In hindsight, I realised I should have killed these multiple rsync processes, and possibly that could solve the issue of CPU being at 400%. But anyway, reboot would be needed anyway. So I had to reboot the server.

The issue is, this board, Orange Pi Zero, it has some weird bug somewhere, it cannot reboot. It only turns off. It’s not a big deal, when your SBC isn’t far (I have my at home), so I’ll turn it on manually in a while, I’ll let it cool down first. But I have an Orange Pi Zero at a remote location, and if I’ll face the same issue there, there’s no chance of manually re-enabling the board. I’ll have to either do a lot of unnecessary moves, or I’ll have to wait till the place have its next power outage. Which happens sometimes.

But what if there’s a very remote location with no power outages at all?

As a one off just browsing the explore section on smth like liberapay can help.

GrapheneOS support for Fairphone is unlikely to ever happen. Their hardware is too insecure to satisfy GrapheneOS’s reasonable requirements and have stated that they aren’t interested in improving it. Software is also lacking and they’ve partnered with Murena, who has been slinging shit at GOS

On a more positive note, due to the AOSP/Pixel drama there now is a real possibility a different major OEM will be supported: https://grapheneos.social/@GrapheneOS/114711328082841462

Fairphone’s… phones are sold in the USA through Murena. They’re preloaded with /e/OS.

Our options were:

• Nvidia GTX 640 (Turned out to be GTX 650) 1 GB for 50 BYN ($17)
• Asus GT 630 1 GB for 35 BYN ($12)
• MSI R6670 2 GB for 40 BYN ($14)
• Gigabyte GTX 750 Ti for 90 BYN ($30)
• BIOSTAR Radeon RX 550 2GB GDDR5 VA5515R for 100 BYN ($33)
• Gigabyte GeForce GT 710 1GB DDR3 for 60 BYN ($20)

He’s planning to get a 40 BYN ($14) card — MSI R6670 2 GB — which is the best option, I think.

So, he bought that card for his Lenovo PC. The full price was 115 BYN ($37) for working PC.

• 40 BYN for GPU
• 75 BYN for the computer

He said he tired to insert + 8 GB ram, and it doesn’t work. My guess is that either the motherboard or CPU doesn’t support it.

I love Home Assistant.

Many years ago we gave our then-toddler an old digital camera to play with. Some time later, we looked at the pictures he took. We were horrified to find out that he took pictures of the outside of the house at night. As in, our toddler would unlock and open the front door, go outside (at night!), take pictures of the house, go back in, close and lock the door, and go back into his bed. I bought some wireless door sensors and created an automation where if the sensors are triggered between 10pm and 6am, the lights in our room would turn on to wake us up.

I expanded this later and today we have sensors on all doors/windows that kids can use to leave the house (we have 4 young kids). As it happens, these are the same doors/windows that burglars can use to enter the house, so this doubles as an alarm system (that we can activate when we leave the house and will notify us remotely if the sensors are triggered).

The best part is that with Home Assistant you are not locked into an app/ecosystem. Our door/window sensors are of a different brand than our lightbulbs, and we control everything from a single app.

1. On an off day, with no reason to require phone use, put your phone in a dresser drawer for the day and do not use or look at it.

2. on an off day, with no reason to require phone use, put your phone in a dresser drawer for the day and leave your residence for at least one hour.

3. Leave your phone at home when either meeting friends, getting lunch, or going to the grocery store.

4. Leave your phone at home when going into the office for one day.

5. Leave your phone in a dresser drawer for an entire weekend.

6. Leave your phone at home when traveling for more than a day (vacation, visiting family, etc.).

I take phone-free walks around the neighborhood, to the store, downtown for a festival. It feels weird at first, then it’s nice.

It was awesome just knowing there was no way a notification could ding, nobody could call me, no news headline could pop up and harsh my mellow. Even if those things didn’t actually happen constantly, simply existing in a state where they _could_was stressful, apparently, and turning the damn thing off was remarkably cathartic.

These are idle measurements mostly.

• June 20, Friday, at 17:10 — 55%
• June 21, Saturday, at 10:35 — still 55%

I did some work, and left it lying.

• June 21, Saturday, at 12:00 — 53%
• June 27, Friday, at 09:00 — 52%
• June 30, Monday, at 14:00 — 49%

I hate webp. Not for any legitimate technical reason like, but I often just want to download an image from the web for an image board or drop it in a diagram or ppt or for a joke and nothing works with that format. Nothing. Google image search is useless because of it.

Cmd+shift+4 is now the only way to grab an image out of a browser. Which is annoying.

It has made my life needlessly more complicated. I wish it would go away.
Maybe if browsers auto converted when you dragged an image out of the browser window I wouldn’t care, but when I see webp… I hate.

Those are the folks from the community that praises AI in each and every shape or form.

Another one:

Worst case you can open it up in Paint and save as JPEG.

I almost throw up reading this. What? Open Paint?

Those are the folks …

Wannabes.

Get the script:

#!/bin/bash
    dwebp "$1" -o  "${1%%.webp}".png

They usually just do a mea culpa:

Camera: https://www.bitdefender.com/en-us/blog/hotforsecurity/facebo…

Audio: https://news.ycombinator.com/item?id=41424016

Conversations: https://www.vice.com/en/article/facebook-said-it-wasnt-liste…

Mass surveillance:https://thehill.com/video/facebook-spying-on-users-new-repor…

Across the web: https://www.wired.com/story/ways-facebook-tracks-you-limit-i…

Beacon: https://www.wired.com/2007/12/facebook-ceo-apologizes-lets-u…

Apps: https://www.theguardian.com/news/2018/mar/17/cambridge-analy…

People who aren’t even on facebook: https://www.vox.com/2018/4/20/17254312/facebook-shadow-profi…

Others do it too, e.g. Amazon: https://www.bloomberg.com/news/articles/2019-04-10/is-anyone…

But Facebook has always been on a whole other level

https://www.theguardian.com/technology/2018/apr/17/facebook-…

reply

Surely self-hosting a chat platform is not that hard? Surely hosting, a chat platform, is not that hard for the Kubernetes folks?

It’s pretty ironic, actually.

I’ll just cite this comment:

Yes, there is, and it’s not as innocent as many commenters here seem to believe.

Have a look at my favorites [0], and you can see recent popular stories which I noticed were flagged (some were unflagged, some were just wiped off the page without being marked as flagged). A lot of those stories are supremely relevant to HN’s supposed remit; from good sources and generating high quality discussion… There are some definite themes there — Musk, DOGE, etc.

And this thread will be flagged too, if people start getting into the discussion. Discussing HN censorship is not “generally” allowed as its own post, ie, [1, 2 etc]. This helps explain why so many people think there isn’t censorship here at all — people don’t see the people complaining, and don’t see the flagged/disappeared posts, so they don’t think it’s happening.

It’s very easy for a few users to flag any story they choose, which means that it’s very easy for mods to blame users for the stories getting censored — “the community chooses, not us”. Discussing how that works is not up for debate though, and posts about HN’s flagging system are removed whenever they come up.

While this is going on a long time here for certain topics, ie, US torturers [3], it has ramped up wildly here in the last few months.

When you look at Garry Tan or Paul Graham’s recent Twitter feeds, and read the weird support for the DOGE team, it’s hard to imagine that there isn’t top-down suppression of the stuff you’ll see in [0] (which is just a list of stories I’ve personally noticed being censored - there’s certainly far more).

0 - https://news.ycombinator.com/favorites?id=mandmandam

1 - https://news.ycombinator.com/item?id=43054368
\ 2 - https://news.ycombinator.com/item?id=42781604

3 - https://news.ycombinator.com/item?id=16576569

reply

Of course there is censorship. Why won’t they censor stuff they might not like? It’s a private website of a private entity monetising all of their users. The more they post there, the more value they create for YCombinator.

The company tries to balance their censorship to be less obvious, especially to newcomers, but since there’s no moderation log, things can be hid on a whim of someone, even a regular user, who can be a bot (as in a Russian propaganda troll factory bot).

They are biased, even despite them claiming otherwise. Why won’t they claim it, after all?

I dislike them. I don’t like Paul Graham, I don’t like Garry Tan too.

By sending manufacturing jobs overseas for decades, America sold out blue collar families so that the rich could get richer. After if/when AI will do the same to middle-class white collar workers, maybe people will realize that it’s long overdue that we eat the rich.

They’ll be living in bunkers in New Zealand, secured by armed guards wearing shock collars (so that the guards don’t eat them).

If the collars are IoT then the hackers will mount a coup.

GPT-4o may display inaccurate or offensive information.

Yeah, me too, your fucker!

I don’t really mind this. It’s clear that GNOME is trying to make a complete graphical environment in a way, not just something which neatly slots into the old school “desktop environment” shaped hole in a traditional UNIX system. It has a dependency on using GDM as your “display manager”, because GDM plays the role of a login/lock screen in the GNOME system. It depends on systemd because a graphical system needs to be able to spawn and monitor background processes, and systemd user sessions plays that role in the GNOME system. Same with seat management etc.
\ There are many great alternatives which do slot neatly into the “desktop environment” hole, most notably KDE and XFCE. These are great projects and using one of them is probably a much better idea than to try to extract only the desktop environment-like part of GNOME to get it to run outside of the intended GNOME system. Personally, I’m excited to see what the GNOME folks are able to do when unconstrained by old traditions like “everything should be a self-contained puzzle piece which can slot together with any arbitrary set of other puzzle pieces”.

I don’t think they’re giving up on modularity or clean architecture. They have gnome-shell doing the shell stuff, GDM doing the display manager stuff, logind dealing with user logins, systemd dealing with process management, GTK for the actual widgets, Mutter for the compositing and window management, dbus for the IPC, userdb to look up user information, etc etc etc. It’s all highly modular.

In fact, adding a hard dependency on various services and utilities increases modularity; gnome-shell no longer needs to have its own built-in process manager for example, it can use systemd user sessions for that.

What it’s lacking is interchangeability. GNOME expects to talk to systemd for process management, not any process manager. It expects to be started by GDM, not any display manager. It expects the window manager to be Mutter, not any window manager. It expects applications to be written in GTK, not any GUI toolkit.

I do think it’s going too far in some places. I really don’t like the lack of server side decorations, which makes all non-GTK applications look incredibly out of place. Nor do I like that e.g the GNOME emoji picker only works in GTK apps. I think there should be a much clearer distinction between “the GNOME graphical system” and “the GTK GUI toolkit”. But all the other dependencies on other modules make sense to me. The GTK situation limits what you can run on GNOME, unlike the other stuff.

When you don’t have resources, it makes sense to concentrate things instead of trying to expand in every 10 directions and burn out.
\ This decision makes sense to me.

TBF “distro” is just a term of art for an operating system that happens to be based on a particular kernel.

A nice catch.

the position of escape, Bill Joy was using an ADM-3A when he wrote vi and csh. Its keyboard layout is where the vi hjkl motion commands came from, and the shell abbreviation ~ for . It has ctrl in the correct location next to A, and esc next to Q. I don’t know of any classic keyboards with esc next to A.

Oh, wow! I never thought of that! Studying history is fascinating. It can explain quite many things. Especially the ones you think of as weird.

Back when mine worked and wasn’t broken, I was super happy with it. Like an iPad and a typewriter for basic tasks. I don’t need it right now, so I’m holding off. It’s just a very convenient little toy. The price is so low because obviously it’s slow and doesn’t support new systems.

On Linux (because that’s what I use), it runs super quiet and super fast — though you can’t really do anything serious on it. But for example, for blogging, it covers 100% of my needs. Even image editing, as long as it’s not 3D.

I should add — it’s not like this is some rare ad. There are tons of them. That’s just what the prices are like now 😭

It’s the price of a battery. And there are even some units with more or less decent batteries. Basically free 😭

I could say it’s the price of a charger — but this specific one doesn’t even come with a charger. Full sets might be a little more expensive. The last time I bought something for that price was a Chromebook. That was the perfect little typing machine, but as of now, Linux is the only thing that really suits me. Or a Mac, but Linux is way better for my current needs.

The Chromebook I showed you — it’s nice, but it’s really frustrating how hard it is to install Linux on it. Very confusing. It can be complicated. On an old MacBook, it installs in 5 seconds — I can literally have everything I need up and running in 10 minutes.

But damn, for now I’m trying to talk myself out of things like this. Maybe if I decide to get something like that, it’ll be next year, for example.

They don’t want us to fight this administration for some reason, and it seems that the wealthy and powerful want this government for the USA right now, so they flag and remove everything that’s negative about it.

— A random downvoted dude at news.ycombinator.

I absolutely despise all these investor folks.

Ahahahahhahaha! 🤣

No! Fuck you, Google.

I’m not trusting you on software since Google Reader murder. I don’t invest into anything except Gmail, and I’d remove it as soon as I’d have my chance to delve into self-hosting my mail. If I’d buy a Pixel ever again, I’d do that only for the Lineage OS’ or Graphene OS’ sake.

First announced on May 20, 2024, Microsoft Recall takes screenshots of your apps every few seconds as you use your computer and then stores them in an easily searchable database.

Laughing in Linux.

Every time I praise WSL on hn I pay the karma tax but I will die on this hill. WSL is more powerful than Linux because of how easy it is to run multiple OS on the same computer simultaneously. It’s as powerful as Linux with some janky custom local docker wrappers for device support, local storage mapping, and network mapping. Except it’s not janky at all. It’s an absolute delight to use, out of the box, on a desktop or laptop, with no configuration required. Edit: for clarity, by “multiple OS” I mean multiple Linux versions. Like if one project has a dependency on Ubuntu22 and another is easier with Ubuntu24. You don’t have to stress “do I update my OS?”

— and realised that the modern development (was it any different at any other time in history?) is mostly trying to combine the output from many not very competent people.

I’m in Linux since 2007, which is 18 years by now. Well, with some huge gaps. I’m in Linux every day since 2019. Which is 6 years by now. And I still don’t understand who are those people who run Ubuntu on servers. I still don’t understand what’s in the Ubuntu. It’s built on top of Debian.

I may understand Ubuntu as a desktop, but I see no point in it, as Fedora is much better, I’d say. Yep, still, folks may need choice, so I’m not saying I’m against it. But why Ubuntu on a server? I cannot grasp it. Either I’m over qualified for this topic, or I’m the opposite, not experienced enough to understand. Maybe one day I’m installing Ubuntu on some of my servers. These days they run Arch Linux (x64) and Armbian or DietPi (for Arm 32/64). Both Armbian and DietPi based on Debian.

I’m eyeing FreeBSD and NixOS, but it looks like it would take me a long time to delve into them properly. I see both as something worth investing my time into.

But Ubuntu? Right now I stand by the point that these people are simply incompetent, having their servers running Ubuntu.

Developing and testing software for Ubuntu Desktop, sure no issues with me.

Also, liked the reply to this original Windows lover:

“More powerful than Linux” is silly. It’s a VM. The most useful thing is that it does a bunch of convenience features for you. I am not suggesting that it is not extremely convenient, but it’s not somehow more powerful than just using Linux. You know what’s even more convenient than a VM? Not needing a VM and still having the exact same functionality. And you don’t need a bunch of janky wrapper scripts, there’s more than one tool that gives you essentially the same thing; I have used both Distrobox and toolbx to quickly drop into a Ubuntu or Fedora shell. It’s pretty handy on NixOS if I want to test building some software in a more typical Linux environment. As a bonus, you get working hardware acceleration, graphical applications work out of the box, there is no I/O tax for going over a 9p bridge because there is no 9p bridge, and there is no weird memory balloon issues to deal with because there is no VM and there is no guest kernel.

I get that WSL is revolutionary for Windows users, but I’m sorry, the reason why there’s no WSL is because on Linux we don’t need to use VMs to use Linux. It’s that simple…

The 10yo does have a lot of extra-curricular activities so having a cellphone would be a big help (to know where they are and give them an option to contact us if anything happens). But there’s just no way I’m giving them a smartphone anytime soon.

One option you might consider is a cellular Apple Watch.

I’m a fan of old tech, and I think it’s actually much better to have an old iPhone (the one with button) vs the new one (the one with Face ID) for a kid, or teenagers. That’s the device that can keep you in touch, but it’s not overpowered to succumb a person into mindless scrolling. Basically because all these services are abominations these days, and won’t work well.

But the Apple Watch idea is not bad, I actually like it.

Context: a comment to Why Are There So Many ‘Alternative Devices’ All of a Sudden?

1. Mango - Sweet and tropical, perfect for a vibrant server.
2. Kiwi - Unique and refreshing, a fun name for your tech.
3. Banana - A classic fruit name that’s easy to remember.
4. Pineapple - A bold and tropical choice that stands out.
5. Coconut - Suggests a sturdy and reliable server.
6. Berry - Simple and versatile, could refer to any type of berry.
7. Peach - Soft and sweet, a friendly name for your server.
8. Grapefruit - A bit tangy, adding a unique twist.
9. Fig - A distinctive fruit that’s both sweet and nutritious.
10. Lime - Zesty and fresh, a lively name for your setup.

These names should fit well with “Papaya” and create a fun fruit-themed naming convention for your servers!

Here are some more fruit-themed name options for your Raspberry Pi server:

1. Apple - A classic and widely recognized fruit.
2. Cherry - Sweet and vibrant, a delightful choice.
3. Plum - A rich and juicy option that stands out.
4. Apricot - A soft and sweet fruit with a unique flavor.
5. Pomegranate - Exotic and full of potential, perfect for a tech setup.
6. Tangerine - Bright and cheerful, a fun name for your server.
7. Cranberry - Tart and refreshing, a unique choice.
8. Paprika - A play on “Papaya,” adding a spicy twist.
9. Raspberry - A berry that’s both sweet and tangy, fitting for a tech theme.
10. Blackberry - A bold and rich fruit, great for a server name.

Feel free to mix and match or modify these suggestions to find the perfect fit for your server!

Here are additional fruit-themed name options for your Raspberry Pi server:

1. Cantaloupe - A sweet and refreshing melon choice.
2. Honeydew - A light and sweet melon, perfect for a friendly server.
3. Nectarine - A smooth and sweet fruit, similar to a peach.
4. Starfruit - Unique and visually striking, a standout name.
5. Dragonfruit - Exotic and colorful, a fun tech-inspired name.
6. Guava - A tropical fruit with a distinct flavor.
7. Lychee - Sweet and fragrant, an exotic option.
8. Passionfruit - Bold and flavorful, a name with character.
9. Mulberry - A lesser-known berry that adds uniqueness.
10. Clementine - Small, sweet, and easy to remember.

These options should give you even more variety to choose from for your server’s name!

Here are some more fruit-themed name ideas for your Raspberry Pi server:

1. Papaw - A lesser-known fruit that complements “Papaya.”
2. Soursop - An exotic fruit with a unique flavor profile.
3. Persimmon - A sweet and flavorful fruit with a distinctive look.
4. Kumquat - A small, tangy fruit that’s fun and unique.
5. Jackfruit - A large and tropical fruit, great for a bold name.
6. Quince - A unique fruit that’s both sweet and aromatic.
7. Tamarind - A tangy fruit with a rich flavor, perfect for a distinctive name.
8. Cranapple - A blend of cranberry and apple, adding a twist.
9. Blood Orange - A vibrant and visually striking fruit.
10. Acai - A trendy superfood berry, great for a modern touch.

These additional options should provide even more inspiration for naming your server!

Here are even more fruit-themed name ideas for your Raspberry Pi server:

1. Sapote - A sweet and creamy tropical fruit.
2. Coconut - A versatile fruit that adds a tropical vibe.
3. Bramble - Referring to bramble fruits like blackberries and raspberries.
4. Fuyu - A type of persimmon that’s sweet and crunchy.
5. Jujube - A unique fruit with a sweet flavor, also known as red date.
6. Rambutan - An exotic fruit with a hairy exterior and sweet flesh.
7. Star Apple - A tropical fruit with a star-shaped core.
8. Miracle Fruit - Known for its ability to make sour foods taste sweet.
9. Cranberry - A tart berry that’s great for a techy name.
10. Bing Cherry - A sweet and popular cherry variety.

These names should give you plenty of options to find the perfect fit for your server!

Here are more fruit-themed name ideas for your Raspberry Pi server:

1. Clementine - A sweet and easy-to-peel citrus fruit.
2. Tangelo - A hybrid of tangerine and pomelo, sweet and juicy.
3. Pawpaw - A tropical fruit with a custard-like texture.
4. Chili Pepper - A spicy fruit that adds a kick to your naming.
5. Zucchini - Technically a fruit, it’s a fun and quirky choice.
6. Bitter Melon - An interesting and unique fruit with a distinct flavor.
7. Satsuma - A seedless citrus fruit that’s sweet and easy to eat.
8. Cranberry - A tart berry that’s great for a techy name.
9. Figs - Sweet and chewy, a unique fruit choice.
10. Pineberry - A white strawberry variety with a pineapple flavor.

These additional options should provide even more inspiration for naming your server!

Here are some more fruit-themed name ideas for your Raspberry Pi server:

1. Cucumber - Often mistaken for a vegetable, it’s technically a fruit and adds a fun twist.
2. Bitter Orange - A unique citrus fruit with a distinct flavor.
3. Dragonfruit - An exotic fruit with a striking appearance.
4. Cran-Apple - A blend of cranberry and apple, perfect for a techy vibe.
5. Pistachio - A nut that’s technically a fruit, adding a quirky touch.
6. Tamarillo - A sweet and tangy fruit, also known as the tree tomato.
7. Kiwano - Also known as horned melon, it has a unique look and taste.
8. Cacao - The fruit from which chocolate is made, a sweet choice.
9. Buddha’s Hand - A unique citrus fruit known for its fragrant zest.
10. Salak - Also known as snake fruit, it has a unique texture and flavor.

These options should give you even more variety to choose from for your server’s name!

Five nines uptime is perfectly doable in a homelab environment, if you are willing to move the decimal point.

I got quickly that ‘five nines’ is 99.999% server uptime. Yeah, pretty huge.

‘But what was moving the decimal point about?’ I thought. ‘There’s nowhere to move!’

Yeah… before I realised there is 🫣

• You can also use the unattended installation system: github.com/memstechtips/UnattendedWinstall
• here’s another: https://schneegans.de/windows/unattend-generator/

Summary:

1. Format disks
2. Install system (packages)

Format Disks

mdadm --create --verbose /dev/md0 --level=1 --raid-devices=2 /dev/sda2 /dev/sdb2

Save the RAID Configuration:

mdadm --detail --scan >> /etc/mdadm.conf

Install System

pacstrap -K /mnt base base-devel linux linux-firmware intel-ucode mdadm neovim grub aria2 foot-terminfo fzf git lf ranger syncthing rsync smartmontools sudo fastfetch openssh lm_sensors bottom bat tailscale i2c-tools

Grub

grub-install --target=i386-pc /dev/sda
grub-install --target=i386-pc /dev/sdb
grub-mkconfig -o /boot/grub/grub.cfg

Packages I use right now:

7zip
aria2
atomicparsley
autoconf
automake
base
binutils
bison
btrfs-progs
cpupower
ddrescue
dhcpcd
duf
ethtool
fastfetch
flex
foot-terminfo
foremost
fzf
gcc
git
groff
grub
haveged
hdparm
intel-ucode
lf
libtool
linux
linux-firmware
m4
make
man-db
man-pages
minidlna
netctl
nfs-utils
nginx-mainline
nodejs
openssh
pacman-contrib
patch
php-fpm
php-sqlite
pkgconf
python-mutagen
python-pycryptodome
python-pycryptodomex
python-websockets
ranger
rsync
rtmpdump
smartmontools
sudo
syncthing
testdisk
texinfo
tldr
transmission-cli
ufw
vim
vsftpd
which
wipe
yt-dlp

AUR:

aic94xx-firmware cloudflared-bin wd719x-firmware

On 2023-12-24-1755:

aic94xx-firmware
aria2
atomicparsley
autoconf
automake
base
binutils
bison
btrfs-progs
cloudflared-bin
cpupower
ddrescue
dhcpcd
duf
ethtool
flex
foremost
fzf
gcc
git
groff
grub
haveged
hdparm
intel-ucode
lf
libtool
linux
linux-firmware
m4
make
man-db
man-pages
minidlna
neofetch
netctl
nfs-utils
nginx-mainline
nodejs
openssh
p7zip
pacman-contrib
patch
php-fpm
php-sqlite
pkgconf
python-mutagen
python-pycryptodome
python-pycryptodomex
python-websockets
ranger
rsync
rtmpdump
smartmontools
sudo
syncthing
testdisk
texinfo
tldr
transmission-cli
ufw
vim
vsftpd
wd719x-firmware
which
wipe
wireguard-tools
yt-dlp

On 2022-08-01:

aria2
atomicparsley
autoconf
automake
base
binutils
bison
btrfs-progs
cpupower
ddrescue
dhcpcd
flex
foremost
gcc
git
groff
grub
haveged
intel-ucode
libtool
linux
linux-firmware
m4
make
man-db
man-pages
minidlna
neofetch
netctl
nfs-utils
nginx-mainline
nodejs
openssh
p7zip
pacman-contrib
patch
php-fpm
php-sqlite
pkgconf
python-pycryptodome
ranger
rsync
rtmpdump
smartmontools
sudo
syncthing
testdisk
texinfo
tldr
transmission-cli
ufw
vim
vsftpd
which
wipe
wireguard-tools
youtube-dl

AUR: 

aic94xx-firmware
cloudflared-bin
pi-hole-ftl
pi-hole-server
wd719x-firmware

It’s really funny.

I need Windows for one minute, to re-connect a wireless mouse to another dongle. There’s this Logitech Connection Utility to do this. But it’s only for Windows.

I have just two neighbours on my storey. We’re on speaking terms, I could ask them for help. If only they didn’t have Macs too.

Nobody I know in a close proximity has Windows.

What a wonderful life.

sudo chmod 0755 /etc/sudoers.d

sudo ls -lah /etc/sudoers.d

total 20K
drwxr-xr-x  2 root root 4.0K Dec 19 09:28 .
drwxr-xr-x 93 root root 4.0K Dec 20 15:11 ..
-r--r-----  1 root root   42 Dec 18 14:14 walter
-r--r-----  1 root root   54 Dec  9 12:07 psd
-r--r-----  1 root root   93 Sep 16 10:27 wheel

sudo visudo -c
/etc/sudoers: parsed OK
/etc/sudoers.d/walter: parsed OK
/etc/sudoers.d/psd: parsed OK
/etc/sudoers.d/wheel: parsed OK

15:24 walter@hackswell:~ $ sudo ls -lah /etc/sudoers
-r--r----- 1 root root 5.0K Dec  9 12:11 /etc/sudoers

MacBook Pro 13

For some reason, it works on my MacBook Pro. I have no idea why, so I’d keep it for the record, to return to one day in the future. I really don’t get it, why doesn’t it work.

sudo ls -lah /etc/sudoers.d

total 20K
drwxr-x---   2 root root 4.0K Dec 17 16:40 .
drwxr-xr-x 103 root root 4.0K Dec 20 14:30 ..
-r--r-----   1 root root  179 Dec 17 16:40 walter
-r--r-----   1 root root   43 Nov 10 20:47 ddcutil
-r--r-----   1 root root   55 Nov 17 12:58 psd-overlay-helper

Solution

Almost two years later, I noticed that the directory permissions aren’t identical, and I need

sudo chmod g-x /etc/sudoers.d

to fix it.

These days, you can buy some super-cool SSD for almost 4 TB.

I was surprised by the price: only $250. By modern standards, this is overkill, because a 256 GB disk ($20) is often enough, if not 128 GB ($10-15), unless you’re doing 4K-8K rendering.

But as recently as (almost) 13 years ago (beginning of 2012) I was buying a 256 GB OCZ Vector SSD for $350 or something. And it died in a year or two. And it’s not that I somehow aggressively wore it down, it’s just that at that time these technologies were worse.

I see that there is a DRAM cache of 4 GB, that is (as I understand it) allows to significantly reduce writes. Well, it should be very durable, so that it’s not stupidly killed in everyday use.

I’m a little shocked. For example, if you take my Hackintosh, you can put 6 disks there. That is, for just 1.5K (which is not much in the context of aggressive video editing), you can get 6 disks with a volume of 23 TB and a faster speed (24 GBps). Well, I think it’s really important to haggle when ordering 6 pcs.

Not that you or I need it (right now), I’m just shocked at how accessible it is today.

In principle, it is only 2 times more expensive than a USB hard drive of the same volume (https://hard.rozetka.com.ua/ua/106473111/p106473111/).

But if anything, I do not advise you. HDD is much more reliable for data storage. SSD is more convenient only for a computer for work, when you constantly rush data back and forth, it is faster. I’m actually excited about this, but there are still very, very few use cases. I made 3 drives in RAID-0 (parallel operation, x3 speed) of 120 GB, the cheapest ones were the easiest ones — wow speed, 360 GB of space, I will have enough for years. It is easy to add 3 more disks, i.e. 2x more speed, and the same volume, +360 more, 720 GB in total. For this, I need only 1-1.5 thousand hryvnias. It could be done, but it is not easy to reinstall the system now, and in fact it is not really necessary, and everything is so super-fast.

But still, I can’t help but be happy about it, as simple as it is affordable! 😮

It is very exciting how cool and accessible modern technologies are. I want to do something about it. But the lack of understanding is very frustrating: what exactly? I looked at the mac mini. If you don’t buy M2 or M4, then the prices for other Intel products are simply ridiculous ($50-150 for a very cool configuration). So for me it’s super-overkill for everything: what’s a desktop, what’s more, a server. But even if I bought it, I don’t know how I would use it: for what? The same situation as with Shortcuts on the iPhone — well, it’s like the coolest thing, you can do so much. And what exactly to do, huh? And khz. And this is just the beginning, in the next 10 years it is difficult to imagine how much it will all advance.

Let’s say so that you understand. If I tell you, for the full picture. I can make a server on a mac mini, for example, with 16 GB / 1 TB, which can be bought for $50, well, maybe $70. Or with an SSD, it’s a little more expensive, well, at least it will be x2, for $100 - 150. This will allow hosting a website… well, I can’t even think of how much user load there should be for this server to be slow. Conditionally speaking, with a certain technological approach, onliner can be hosted (in the context of the volume of data).

Maybe there is more data, but even that is a very large resource to which information is added daily. And there is still a different arithmetic, because it earns a lot, that is, you can not count money on such equipment at all, and buy whatever you want.

If I wanted to host some private project on a mac mini, like everything I showed you, just… well, I don’t know, I’d start a super blog with videos and a billion pictures… well, I don’t know, travel -blog. I probably wouldn’t have been able to load it anyway. Only millions of simultaneous views would possibly download it, and even then I’m not sure.

How to add redundancy to take into account, for example, blackouts, and do load-balancing (when you go to the first server, and it is loaded - it sends you to the second, or to the third, etc.) - for example, to put 4 servers, mine, yours, in mother-in-law, in toi toi — this will give 4x for everything, i.e. if 1 server allows a million users, then 4x will be 4 million. And all this can cost $200! Or $400 if I’m an oligarch. Or if I am the richest person in the country, then you can buy an M4 from a store for $500, then it will be $2K (with a small difference in speed, but a big difference with light, then you will not notice at all that you have a mac mini in the house and is eating something there).

And taking into account that the main loading of a conditional travel blog is vidos (in 4K, the rest of the bullshit), which is still better to host on YouTube (not that it’s free, but you get paid for it), then in fact for a full-fledged cool site- one (or two in different locations, for reliability) mac-mini will be enough for the blog.

I really can’t imagine what needs to be done, what kind of project, so that it is necessary to invest a lot of money in it. Unless some enterprise-thing, well, like to make another Rozetka there. But even in another 10 years, I will not be surprised that such a site will be available at home on a Mac Mini.

In short, I am impressed (not only SSD, but globally)

🤯

For example, I thought there might be value in local online communities. If we take Lviv, or if we take a more complex project, it is a conditional resource where millions of people can communicate in the context of competition. That is, for example, anonymously. Ah, how it all happened in Telega, but not in Telega, but on some forum. Which is hosted on a mac mini. I think one or two would be able to simply take the whole load here. Even if the project is not for making money at all, but is entirely ideological - an investment of $300 🤨

So what do they work, well, they consume a maximum of 100 watts, that’s for us here for a month, well, I don’t know, 100 UAH maybe?

That’s what I understand. For example, I don’t understand (almost) 3D. But as I understood, even for your work a Mac Mini on M4 may be enough, unless it is not the cheapest one for 500, and there for conventionally 1.5K. But as I understood, this is enough to simply replace a gigantic computer. And it will last for many years. But how much easier it is for a beginner: it’s easy to come to the store, buy a ready-made computer, you can work easily within 10 minutes.

Well, there is also AI/ML, and there I understand the same situation, that such a computer is also enough for many things, except that the budget is a little bigger, but there the earnings are much higher, that is, in the relative budget, everything is a trifle.

‘Get yourself a Gentoo handbook, and never be afraid’ he continued.

I was afraid. Of breaking my system, of being unable to use my computer. I didn’t understand how to use Photoshop, after effects, illustrator, 3Ds Max (which I used only in a virtual machine!).

I gave up on Linux, and I went back to Windows for a couple of years, and bought myself a MacBook Pro, eat a virus bro.

I’ve been on macOS for almost a decade. But Linux haunted me, and after a couple of my Mac years, I’ve installed Debian. For my mom, some distant acquaintances, some obsolete computers here and there to play with it. I committed to supporting the systems. thinking Debian would be easy to support.

Debian was easy, but not simple. I didn’t like it, and it wasn’t my cup of it. It wasn’t something to be afraid either. (It’s one of the greatest distros, and one of the best. Just not my cup of tea.)

At some point later, I discovered Arch Linux. I was very afraid of even trying it, not for newbs. But I was slowly approaching it, been closing my eyes in fear and jumped.

And been falling. And the rabbit hole turned out to be immersive.

I’m still falling, five years later.

I was successful since my first attempt. But it’s only this year when I’ve become confident with Arch, even the installation process. I guess I can install it without the guide now, but I still consult one, just to be less doubt on myself.

But I always wanted Gentoo. It was my favourite, like, always. It’s that true love that isn’t mutual. Or it is, but you’re too afraid to even ask. No, not even ask, even speak. Even, come near. I’m not paralysed, but afraid of something, something I cannot explain.

I tired FreeBSD meanwhile, and I like it very much. It’s too good to use daily.

Because of the hardware support. Well, and the lack of my knowledge.

Gentoo inherited ports from there.

These days, I’m pretty sure I’ll have no issues with Gentoo, and I’m looking forward to that point in time. I’m just too consumed with my parenting adventure nowadays. I’ll come to it, when I’m ready.

The issue with Gentoo is, apart from the knowledge gap, is that I’m too sure I won’t be able to compile everything, I don’t have that powerful computers, I think. Or maybe I have just one, the one I’m working from. Laptops, I guess they’re in a different league.

And when I’m arrived at that destination of accepting Gentoo as my potential distro.

Or accepting me, as someone who can run Gentoo, actually.

And when I’m arrived there, I see this Nix os. And I like it too. And I think of it, I’m too newb for it.

Or am I not?

• home-fedora.tail1gsaaf.ts.net
• rpi2b-01-video.tail1gsaaf.ts.net

I changed the addresses, just in case. However, I believe nobody can enter, since it’s effectively a VPN, you need credentials.

Yet, somehow I’m still afraid of uploading even photos, not to say videos.

I want to rethink this. I’m not thinking twice when sending heavy media over whatever chat app I do that. Yet, uploading it to my own hardware, feels wasteful. Actually, that’s just the opposite. You can buy a decently good laptop-grade small used HDD for like $10, and it’ll serve your personal needs for years. For half a year, maybe, if you’re bombing 4K videos of your kids non-stop. Still, very cheap.

Among Gen Z, there’s a problem of how a lot of people don’t even know how to use a computer in the workplace because they just used their smartphones for everything. With this phenomenon in retrospect, I was wondering if Linux could be a better introduction to an advanced computer.

Linux tends to be more secure against malware, can be made to resemble Windows superficially enough, and you can keep kids away from executing more dangerous commands with a sudo password. The compatibility issues with a lot of software can also be used against them trying to execute pirated games and the like, basically making it easier to control what they have access to.

I know there are stories of how if you apply child protection time limits or website blocks they end up adapting and circumventing it…but if they do that with Linux, then they are also forced to learn how the system works well enough to familiarize themselves with programming knowledge which could help them in the future.

• Hardware Acceleration With MPV

I do it like this:

vo=gpu-next
gpu-api=vulkan
hwdec=vulkan
gpu-context=waylandvk

You can replace hwdec=vulkan with hwdec=vaapi if Vulkan video doesn’t work.

Source

The tablet isn’t very useful (as a graphical tablet) with multiple displays setup with sway and or hyprland (I haven’t tested, but read it’s similar to sway). But it should be useful with Gnome and KDE, if needed (also, just read about that). Since I don’t really need any graphical tablet these days, I just use it as a nice scrolling pad thing, instead of Apple’s one. If you have one with touch, you may want to try it too. It works out of the box, iirc. Or maybe there’s libwacom for that. Or maybe libwacom is for just tuning it, I don’t remember. Anyway, it’s quick and easy to test. I plugged it in, scrolled, worked fine, let it be then.

To migrate my Thunderbird client from one PC to another, I just copied ~/.thunderbird directory. Cool! Interesting, what if I’d sync the directory across all my devices using Syncthing. I may add one more email account from one machine, and it should magically appear on another one. Not really useful, but still I’m amazed by the easiness of the process.

I killed my Hackintosh entirely and created one more Arch Linux machine. It looks like I moved most of the things I needed, including even Virtual Machines workflow. Excluding Hackintosh and Windows for design work, will do later.

Btw, I’m amazed how easy it is to do many complicated things with the help of ChatGPT. The most inspiring thing for me was disk management (I created quite a complicated set-up) and sway tuning (made most of my DE fantasies work!).

I was looking for a mouse-centric yet lightweight system for my netbook to keep it as a music server (it has a broken keyboard, so it’s either ssh or external keyboard). I wanted to use Wayland, and looks like there’s no DE for that. Either it’s X11 and LXQt, LXDE or similar projects. Or it’s keyboard-centric (sway, hyprland etc.). Or it’s full-blown Gnome or KDE. Ha-ha! May need to use it without display then or what. I don’t know. In the process I found a funny video, scroll through it. It has so funny ending that I laughed! You may like it, too: https://www.youtube.com/watch?v=T4puCAp1naY

Lowering resource usage with foot and systemd

For elder people, I think that’s an ideal device. Most people don’t need a desktop operating system these days. The 1st Gen. iPad Pro won’t get past iPadOS 16, while the current is 17. But I believe it won’t be a problem for the next 5 to 10 years, which is plenty considering the price tag.

However, with Linux it’s much better software-wise: if you have relatively recent hardware, your support would be literally indefinite. I use 15 years old hardware and I can believe it would serve another 15 easily.

This iPad… well, we’ll see!

Ask HN: How to manage phones and PCs for elderly parents?

I have a friend whose parents are getting up in years. They’re still functional enough that they want to be independent and surf the web, check their stocks, etc. He has some remote desktop software installed on their PCs, and antivirus installed and locked down. But they keep destroying their PCs with downloads, keep turning airplane mode on for their phones and can’t figure out how to disable it, and recently his dad got in a cleaning mood and deleted a bunch of system files and bricked his machine. He’s almost to the point of setting up an Intune tenant just to onboard and lock down his parents devices. Any suggestions on how to manage phones and PCs of the elderly to keep them functional but useful?

A rack unit (abbreviated U or RU) is a unit of measure defined as 1 3⁄4 inches (or 44.45 mm). It’s the unit of measurement for the height of 19-inch and 23-inch rack frames and the equipment’s height. The height of the frame/equipment is expressed as multiples of rack units. A typical full-size rack is 42U high. Racks and cabinets are then listed as 1U, 2U, 3U, 4U, and so on.

Home Lab Beginners guide (Hardware)

Will I ever be that popular? But the real question is, of course, will I ever want to be that popular?

That’s the reason I never follow all of these people. Just read them randomly, not regularly. Thank you very much, I know I can use shift in a browser, thank you for enlightening me, sir.

I mean. I’m totally fine with the author posting this. I do use my blogs as my notebooks too. My issue lies with the people sharing this actively, as if this is some very meaningful advice.

pacstrap /mnt base base-devel iwd git linux linux-firmware dhcpcd man-db man-pages texinfo openssh vim sudo tlp intel-ucode lm_sensors rsync grub aria2 minidlna neofetch smartmontools syncthing transmission-cli ufw vsftpd yt-dlp ranger pacman-contrib

and also micro-code

• Intel: intel-ucode
• AMD: amd-ucode

On a laptop

pacstrap /mnt base base-devel iwd git linux linux-firmware broadcom-wl dhcpcd man-db man-pages texinfo openssh vim sudo tlp intel-ucode lm_sensors rsync wl-clipboard aria2 grub ranger neofetch smartmontools syncthing ufw yt-dlp sway swayidle swaylock bemenu foot firefox fzf telegram-desktop viber transmission-cli pacman-contrib

Bootloader for Apple Macintosh

Instead of grub I can use systemd-boot.

I need to create an extra file: /boot/loader/entries/arch.conf

title   Arch Linux
linux   /vmlinuz-linux
initrd  /intel-ucode.img
initrd  /initramfs-linux.img
options root=LABEL=arch_os rw

Microcode

Installed intel-ucode package as microcode to Arch Linux bootloader.

Added initrd /intel-ucode.img as first initrd in /boot/loader/entries/arch.conf

[SOLVED] In GRUB Shell, unable to boot into arch.

Boot from your installation media, mount your /mnt & /boot directories etc. the same as you did for installation e.g. mount /dev/sdax /mnt. Then you need to change root: arch chroot /mnt

Then go from here: https://wiki.archlinux.org/title/GRUB#Generate_the_main_configuration_file

# grub-mkconfig -o /boot/grub/grub.cfg

Sensors

sudo pacman -Syu lm_sensors

Packages (1) lm_sensors-1:3.6.0.r41.g31d1f125-1

Total Download Size:   0.13 MiB
Total Installed Size:  0.46 MiB

Optional dependencies for lm_sensors
    rrdtool: for logging with sensord
    perl: for sensor detection and configuration convert [installed]

After Installation

systemctl enable dhcpcd

No longer relevant, I use systemd-networkd and systemd-resolver now.

https://www.reddit.com/r/swaywm/comments/tn8a5t/bindsym_to_wlpaste/

For that, I downloaded the firmware from the official website, Techdata: TP-Link Archer C60 v2 search for Firmware OpenWrt Upgrade URL.

• direct link as of today.

I upgraded through web-interface.

• Visit web-interface
• Then find System → Backup/Flash Firmware
• Then find Flash image and find your latest image you downloaded from the website
• Double-check it’s -sysupgrade, and not -factory!
• You can skip renaming it.
• Upgrade is in the process.
• Router will not reboot on its own. So, wait a couple of minutes (or 10…15, just in case), and then turn it off the socket, and back on again.

That’s the software vendor we used for the plotter: https://www.silhouetteamerica.com/software

I’ve started using the ipfs add command to share files that are too big to email. The users can just download the file from a gateway like ipfs.io without installing anything, and the file is only hosted on my laptop.

• https://docs.ipfs.io/concepts/how-ipfs-works/
• https://docs.ipfs.io/how-to/command-line-quick-start/

• #836772 - pinentry fails when used from dedicated account via su -
• and this bug looks similar to mine, with the exception I use Linux and am able to do many things to cover the bug.

• If you have Broadwell or earlier, and have USB 3 support, then you likely have EHCI and XHCI

Like… THE FUCK?

Update: Aha, it’s SystemD lingering, whatever that thing is.

Hi from Singapore, by the way.

• The work one is for work tasks
• The private one is for private tasks

One iPhone and one Android is even better, when you need to cover various scenarious and test things. Why didn’t I think of that earlier?

Here I mean Samsung, Xiaomi, Huawei and others, mostly Chinese companies.

I think I’ve got it. They want us to get used to their version of the system, so we would just buy their smartphones, as we already used to them.

To some degree it’s the same for me, as I’ve been buying only the Google-branded smartphones, as to my liking all the others are just shit. Well, and iPhones. I keep buying iPhones, because it’s stays mostly the same, keeping the UX consistent, and I’m buying that.