m3047

The biggest harm today is people in training for "people interaction" specialties with a high degree of empathy / ability to read others: psychology, counseling, forensic interviewing. They pay a lot of money (or it's invested in them) to get trained and then have to do practical residency / internship: they need to do those interships NOW, and a significant proportion of the population that they'd otherwise interact with to do so is off faffing about with AI. The fact that anecdotally they can't seem to create "convincing" enough AIs to take the place of subjects is damning.

> behind the scenes

Asbestos was at times used in:

* cigarrette filters

* water filters

* hair dryers

* space heaters

* anti-scorch pads (stoves and bunsen burners)

* HVAC duct sealing

* boiler, pipe, and duct insulation (buildings, machinery, vehicles)

* brakes

Debugging co-routines / threading is a pain in any language. I'm a big believer in observability anyway so this isn't a leap. Nor is it the first time I did this very thing, but last time I had implemented a console and it was a console command. This is in a product written in Python which implements a DNS gateway for Redis:

    # dig @sophia.m3047 coroutines.redis.sophia.m3047 txt +short
    "write_control:1"
    "get:3"
    "statistics_report:1"
    "process_pending_queue:1"

The implementing code (not including turning it into a DNS response) is surprisingly simple: https://github.com/m3047/rkvdns/blob/f6fba9523a83aa5125378bb...

_The Pervert's (Lacanism) Guide to Computer Languages_: https://www.youtube.com/watch?v=mZyvIHYn2zk or if you're old school or your father was a shrink you could use the Leary Wheel from his _Interpersonal Diagnosis of Personality_.

That would change the dynamic. Some might object: "psychohistory!" What are we doing here anyway? And, what are we meant to be doing? Where is the "philosophy of computer science"?

Not a collab environment (or wifi firing range), more of a curated collection of mostly static resources. But it's a good idea.

Having watched the repeated immolation of blissful innocence since smart email clients would run whatever smart (OLE? Smart? I'm kidding.) document was delivered, this is going to be so much fun in a trainwreck kind of way.

Ground feature echo (mentioned in the article)... possible. Not mentioned here or in the article: thermocline in the atmosphere. Thermocline in the water is traditionally how submarines "hide" from surface ships.

I hear flanging from the planes incoming from quite a distance, and they're pretty low when the fly over where I live. More telling: I can hear the freeway and the busy arterial, "depending on how the wind blows". Sometimes it flanges, too.

So: ground reflection along with thermocline refraction seems a perfectly plausible explanation for one source of the phenomenon; could be several, probably all involving ground and atmospheric factors.

If AI could build tires then the quick lube could just come to your place of work and assemble the tires right on your car from raw materials, amiright?

Look, my last set of tires cost $1000 (and that was before COVID). They might last 50K miles. How much do you drive? There is no way that new tires plus replacement is going to be $7.50/tire. Even a unicycle isn't going to have a $30 tire.

Not the ur-chicken I'm referring to, but a great read... could be a post on HN:

https://www.audubon.org/magazine/november-december-2014/stal...

(I own chickens. There's an "Ur-chicken".)

The thing that jumped out at me in the article was this implied definition of "intuitive":

"In mammals, brain development follows an intuitive path: The cells in the embryo’s amygdala region at the start of development end up in the adult amygdala."

So let me see if I understand: Here is rock on hill. Rock stays on hill, or rock falls down hill. Because it's intuitive.

I'm not sure I agree that that's what "intuitive" means.

There is a transmissible dog tumor which originated with handwave wolves. Apparently there are possibly others. (TIL!)

https://en.wikipedia.org/wiki/Canine_transmissible_venereal_...

I put auto horns on my motorcycle(s). In the Seattle urban core (when I lived there, 15 years ago), if I honked then people looked for the car -- they weren't looking at me. Other than that, seems to have the intended effect.

If you're on Medicare and trying to figure out whether something is covered, CPT codes are the glue in a rat trap, the pin sticking bug to a board, you get the idea.

Everything is done by CPT codes. But you can't go and reference them yourself, even on the Medicare web site (no, really you can't). Doctors will tell you something is covered, then you have to demand the CPT code from them and call Medicare to find out if it's covered. Or, you can trust the doctor... who will be correct until they aren't. Sometimes doctors don't even know what CPT code they'll use until they see you.

This just in... /s

Seriously though, Sullivan lost his appeal. You should have read up on this.

https://www.courthousenews.com/wp-content/uploads/2025/03/us...

Collect underpants... (LLM).... profit!

Using the recipe site cited, the only acknowledgement of profit is "but seriously fuck you if you have that many popup ads". Mkay...

"In 30 days, we launched a robust API & an RSS feed for top products. Within three months, traffic from AI-driven referrals jumped by 200%." yeah that's in the context that "I recently consulted for a mid-sized e-commerce firm".

A recipe site (with ad revenue) vs an ecommerce site: not the same revenue model.

Not a peep about sites which are expressly built for humans, or should be built for humans as the primary goal, with "profit" achieved by other means... maybe not profit at all.

The thesis that LLMs prefer / will prefer APIs is an interesting one, but unproven technically as far as I know as well as unproven as a viable economic model for content publishers.

Yup that's it.

By making its own "internet in a box": open wifi, DHCP and DNS, and Bob's your uncle. Anybody in range can play; everyone else is out of luck. Install NNTP so that news can move around by flooding, as nodes form new associations. Add a second wifi transceiver and some clever dynamic route discovery and you have a mesh network.

I made a wifi firing range / disconnected collab environment to this end a few years ago. Has BIND (which takes over the world), Apache, DAV, Etherpad. I carry a VM of it around on my laptop "against the day".

My eyes have always bugged me, but I've also had allergies. It's also become clear over time that I have pretty bad astigmatism. I've always been farsighted.

I first got "dimestore readers" specifically for computer use, to reduce eyestrain. When I put them on at the beginning of the day I'd notice some weirdness (phantom artifacts and double vision) but my brain would cancel it out and all was good. After about a decade though, I wasn't noticing those artifacts at the beginning of the day, but while driving home I'd see double. That led to getting prescription glasses, and learning lot about astigmatism.

Until recently, I always tested better than normal at infinity without glasses. At infinity, most of my correction is astigmatism. Corrected, I'm still 20/10. But here's the "weird trick": I've learned that it's something that my eyes / brain are doing combined with the astigmatism. I can still conjure "hot spots" with near perfect clarity even without glasses, but it strains my eyes (I don't get full FOV clarity without glasses anymore).

I can't stand progressive lenses, although I have a pair for super close-in work. All my task glasses are "single pane of glass", my "infinity" glasses are bifocals. These days my eyes are more comfortable wearing the infinity glasses than not. The truly noticeable difference with glasses was night driving: no more stars! Huge reduction in eyestrain and improved basic ability to see when challenged by oncoming headlights; I've combined that with some aggressive rose tint in a pair specifically for night driving (very 70's mod frames. woot!).

I'll never get Lasik because I've been told repeatedly that if I did so they wouldn't be able to correct me to 20/10 any longer.

On second thought, no I won't open an issue. I provide contact information here and on GitHub. Get in touch if you feel like it, you make it too hard to contact you.

Yeah, something like that I guess. "Debugging" implies for me a traffic volume which I don't intend to support. My thinking is based on properly segmented (DNS and DHCP properly implemented at the segment edge, not forwarded) networks, recognizing that diagnostics and evaluatives are distinct, doing the work to make evaluatives widely available in a controlled fashion, and remembering that diagnostic investigation will take place on the segment where complete logging is available. That sentence is way too long!

It's probably simpler than you think it is. ;-)

  1: there's something producing diagnostic information which needs to be monitored
  2: identify and construct evaluatives
  3: put evaluatives in redis (e.g. rkvdns_examples/totalizers, but you can also write directly, see shodohflo/pcap_agent)
  4: put RKVDNS in front of that (this is where the control plane ends)
  5: put e.g. BIND in front of RKVDNS (BIND sits at the segment edge)
  6: lie to the world that the BIND instance is authoritative (optionally hack BIND to always return AA, but never mind, nothing cares)
  7: use the full set of features provided by BIND for access control, integrity, privacy

DNS as exfil / C&C can either bypass DNS infrastructure or embrace it; different mitigations / detection techniques accrue. (You just gave me an inspiration for an observable I should add to Rear View RPZ!) I'm not tunnelling protocols inside of DNS; our objectives are probably diametrically opposed, but ok! RKVDNS is meant to be a "data diode": you can get selected, curated data out but you can't put any data in.

DNSSEC is problematic for something like RKVDNS (as are zone transfers), doing negative proofs in particular. DNSSEC does not provide encryption; it provides proof of authenticity, and nonrepudiation / integrity. DNS can be used to publish public keys (DANE), and DNSSEC provides proof that the key belongs to someone / something.

I need to say an important thing here: DNS is a wire format, an algorithm, and an application layer protocol. That last one in particular seems to be something people want to wish away. Every (recursing) DNS server is like an HTTP proxy: what happens to proxy headers when you chain HTTP proxies?

Encryption is per-hop, each DNS server has access to the entire DNS question, and assembles its own response. DNSSEC provides identity across the entirety of the DNS; there is no encryption / privacy across the entirety of the DNS.

On a per-hop basis, encryption is provided by DoH and DoT. DoT is as simple as putting nginx in front of the nameserver (config templates used to be included with the BIND distro, I think Ondrej was responsible for that). You're more likely to have success implementing (key based) access management with DoT; if you don't need encryption, you can get access management and non-repudiation with TSIG (from BIND, RKVDNS doesn't support TSIG), on a per-hop basis.

The most common santa claus and easter bunny question I get is "why didn't my super special EDNS data make it to my authoritative server?": people truly wanna believe. That's because EDNS is per-hop; I don't care what Cisco or Microsoft say, ain't gonna happen. (That's another reason why you need true DNS at the segment edge, not forwarding.)

("My favorite programming language is the one with the least maintenance burden. If you find it, please let me know.": In my somewhat grizzled experience lack of dependency management is the biggest burden. I try to minimize dependencies and live off the land as much as possible.)

I'd love to see a demo. We should demo for each other. I will stalk you and attempt to send you an email. (Oh I see. I'll open an issue on GitHub then.)

I've been working on something I call "Poor Fred's SIEM" (a play on Poor Richard's Almanac) for several years, rethinking an approach to observability based on the functional difference between diagnostic and evaluative observables: cooking evaluatives at the edge, presuming that diagnostics will be investigated on the (well-implemented) network segment where they are recorded. I've also rethought PTR records based on local knowledge, dovetails quite nicely. Living off the land: redis, python + dnspython + dpkt, DNS as both observable and transport.

The core components are on GitHub:

   * https://github.com/m3047/shodohflo
   * https://github.com/m3047/rear_view_rpz
   * https://github.com/m3047/rkvdns

Not the first time I've referenced these on HN. I did just implement a Redis operator which Redis doesn't have, SHARDS:

   * https://github.com/m3047/rkvdns/blob/main/SHARDS_Command.md

No GUI, command line tools work for me. I've got a toolkit. I don't give it away because I don't want to support it but if you made friends and asked nicely I'd toss it your way. Here's a silly little two minutes of your life... done in one take, in Skype:

   * https://www.zipcon.net/~m3047/observability-dream.mp4

I actually eat this dog food to the point where I publish (some) telemetry data on the internet with this toolkit using DNS. Again, make friends and I'll tell you where to find it. I think it would be great if other people ran this toolkit and published telemetry data for other edge operators to see: things can be federated, they don't need to be centralized.

There is a continuum, and at the parasitic end you'll find Gastrodia Elata. It has basically no leaves at least in its herbally desirable expression where the tuber is called Tian Ma or colloquially "expensive potato".

The wild version is (should be!) unavailable because it's considered endangered, the cultivated version costs $80/lb. I know it well; I take small amounts of it in combination with Gou Teng and otherthings mumble.

https://en.wikipedia.org/wiki/Gastrodia_elata

Where I live the City helpfully^H^H^H^H^H^H^H^H^H performatively lets you see where the power is out, but provides no historical information. They pay VertiGIS to provide this service. At least today (I checked, it was something else last time I checked). Woot!

"Performatively": yes. If you can't see historical information, it's performative, it misses the historical case which is the important market. If you're there and the power's out, you know. If you can do something about it in realtime, you will; otherwise you will wait until the power comes back on and disaster recovery kicks in: From when to when? For how long? If you're not there, then go directly to disaster recovery. That's my opinion, and I'm not changing it.

The second important consideration for a democratically governed entity would be: are we equally served? If not, why not? Retrospective information for the entire stakeholder group is required.

Too much ado is made of "security", and that historical information is somehow a threat to operational security. If the power is out now, the live info is the impacted targets of opportunity right now (or the ones to phone / smish when the power comes back on); if the power was out in the past, what's the threat? Are adversaries going to pre-position in areas where outages are predictable due to some foreseeable conditions? Then maybe the City should preposition resources, too. I welcome other viewpoints on this aspect.

On a practical level I have systems which are always on and logging with sufficient granularity so I know when the power went out... and when it came back on. I would think that telemetry from locations within the service area would be ultimately a more reliable way of collecting information about outages, without relying on the utilities which can't be relied on (outside of a contractual arrangement). This telemetry could be active, or passive: geolocating web browser and other internet activity (even pinging or SYNs) would likely do an adequate job, I'm sure that stationary resources could be identified in the dataset. I'm sure this is colored by the fact that I'm an "internet plumber" and telemetry and observability is what I do for a living.

Trump killed the Trans Pacific Partnership.

(This was a surprise to me. I'm grateful that I'm not at risk of being extradited to rot in a chinese prison for reverse engineering some IoT POS.)

A discussion thread which won't die:

  https://github.com/orgs/community/discussions/135572

I don't regret posting my comment (on Sept 4, not shown by default) because it truly was a shit change, but finding a viable alternative to GitHib has definitely elevated in my "get around to it" pile as the preferable alternative to ever giving feedback again. It just exposes how much else interferes with getting things done.

How do I make the comments stop? It's noise at this point. My comment isn't in my own account page under "contribution activity". There's no way to search for the comment in the actual discussion thread and trying to search page source for e.g. the username doesn't do any good because not all comments are shown by default; changing the sort to "top" doesn't work because, even though I only got 2 upvotes (and 1 emoji), the initial display still hides the majority (221 out of 251) of items to display lower priority stuff at the bottom... and besides it's broken and threads aren't sorted strictly according to any interpretation of "top" I can muster.

I reported weird shit happening with SYN and PING and what I got was "how dare you insult my reports" from Paul Vixie; but I used to work for him. Ultimately I blocked all SYNs and ICMP ping inbound from Amazon addresses, spoofed or not. Problem solved. Boohoo soi disant "security researchers".

On a quarterly scale, history shows it typically has little to no effect on e.g. stock valuations.

Here's Renee Burton's (at Infoblox) comment on Philippe Caturegli's post on LinkedIn:

"When we contacted DNS providers about sitting ducks attacks ONGOING in their network via lame delegation... some responded with aggression and others with ambivalence. no criminals were disrupted and it was a waste of our resources even though it was the right thing to do."

And I can personally vouch that's mostly my experience and expectation as well, and not just for DNS issues.

I'm getting warnings on an old Macbook Air that the Firefox CA certs are going to expire... except the OS is too old to update to a newer version. Oh noes!

Do I really care? That would imply I trusted CAs in the first place... all of them.