OhMeadhbh

I can't say too much that Richard Murphy hasn't said already.

Here's his latest video: https://youtu.be/mKWfEQlcDoU

I encourage everyone to read about it if you're a person who absorbs info via text better.

Here's the Wikipedia entry: https://en.wikipedia.org/wiki/Modern_Monetary_Theory

It's worth a read. I wish there was a single volume "MMT for Dummies," but to get the full read on it, you currently have to read a bunch of different texts. I like the Wikipedia article because they include some of the critiques as well as the supporting arguments.

My take on economics is there are few theories that are universally applicable. Keynes is applicable to some situations as is Hayak and even E. F. Schumacher. The hard part is figuring out which theory is good for which situation (and how to tell if your assumption about applicability was wrong.)

What I find surprising about modern Americans is they really want to elect someone who has all the answers. FDR was somewhat famous for f*king up a couple times and changing course when the data didn't match the theory. We're so tribal these days we refuse to admit that even good leaders can get it wrong once and again and having an honest discussion about why the error occurred and how they're going to fix it might lead to better outcomes.

I am reminded of J. R. "Bob" Dobbs' quote "Act like a dumbsh*t and they'll treat you like an equal." Perhaps I should run for office on the platform of "I'm probably as dumb as you are, which makes me smarter than my opponents."

As David Graeber once said: "The ultimate hidden truth of the world is that it is something that we make, and could just as easily make differently."

I believe this, which I supposed makes me a Graeberite. Graeberian? Graeberista?

well... georgism is sort of orthogonal to capitalism and socialism. You could have a georgist tax regime in both capitalist and socialist utopia.

But... yes... it would be great if enough people knew about Henry George and Georgism that the Gallup people thought it was worthwhile to put it on a survey.

And allow me to bemoan the current state of affairs where the definitions of capitalism and socialism used by most Americans are defined by tribal communities whose (sub)social order depends on the in-crowd toeing a particular party line with respect to whom is allowed to extract rent, set prices, rent and tax rates. Which is to say, if you're comfortably making seven figures on your constellation of payday loan offices, you probably a fan of the status quo.

For my right-wing nut-job friends, remember, there are "socialist" countries like Sweden, Denmark and (arguably) Canada which are perfectly pleasant. Venezuela and Cuba's problems aren't due to their "socialist" nature, but to wide-spread corruption and oligarchy. Keynes (and even Marx) had interesting things to say, though doctrinaire adoption of political and economic philosophy from previous centuries is not likely a good idea.

For my left-wing pinko friends, remember, not all "capitalism" is regressive. Even places like Sweden, Denmark and Canada have a form of capitalism despite being mildly constrained by collective action. Hayak had interesting things to say, though doctrinaire adoption of political and economic philosophy from previous centuries is not likely a good idea.

Stop me before I mention Modern Monetary Theory.

Bonus points for wedging a Vignelli quote in your response.

I think artists are unhappy that something that does not fit all the criteria for "art" is being described as "art."

"Comfort implies art" is different from "art implies comfort."

Which is to say, not all art is comfortable, but not all things that are not comfortable are art.

It's perfectly fine for art to be comfortable. Stop gate-keeping.

Alas. Some people cannot use Argon2id due to regulatory or corporate policy reasons. NIST SP 800-132 hasn't received any love since 2010, so it's understandable it's not on the list of "approved" algorithms. The people involved in the "Password Hashing Competition" seem like serious people (I noticed Peter Gutmann and Matthew Green's names attached) but I suspect NIST really wants to see an "official" selection competition with plenty of NSA spooks given the opportunity to weigh in. The flip side is after the Dual_EC_DRBG kerfuffle, it's hard to imagine too many people wanting NSA in the room at such meetings.

Which is to say... yes... I would probably use Argon2id if I could, but I can't (I mean... not in all projects) so fixing this bug is appreciated.

Wasn't 'v' (vlang.io) once called 'volt'? Did they change their name because of a name conflict with another project?

I certainly agree with you, but this can't be news. It seems to have been slow and buggy for at least the last 10 years.

It's like an Apple Newton, but at a much different price/performance point.

If only that worked.

So... PLATO user, Notes user and Thunderbird user here with a couple of notes.

   a. Everything you think is new was created in the 1960s by Don
      Bitzer @ UIUC or Doug Engelbart @ SRI.  (and probably also
      Robert Creasy @ IBM.)

   b. Lotus Notes (and Groove) were products, not protocols.  If
      you wanted to use them, you used Windows and the server and
      client from the vendor.

   c. Lotus Notes only sort of worked over the intarwebs.  It had
      serious problems traversing firewalls, and technically worked
      over VPNs, but was nearly unusable as it required you to
      either replicate your AD server or use AD over the VPN (I do
      not recommend the latter option.)

   d. Like JIRA, in order to do anything interesting with Lotus
      Notes, you effectively had to have admin privileges.  This
      isn't strictly true, but it's effectively true.  Both JIRA
      and Lotus Notes were designed for organizations whose thinky
      bits were managed centrally.  Have a site or department
      specific process?  Better schedule time with your admin to
      talk about how they're going to model your work-flow because
      there's no way you're going to be able to build it yourself.

      Groove was a little better, but it was disturbingly easy for
      the brittle XML based DTDs used to model data to become out
      of sync between users.  Guess it's time to restore from
      backup.

   e. As much as I loved PLATO, it was a centralized / mainframe
      type system.  Of course it was easy to coordinate via a notes
      file, it was a single file on a centrally shared,
      non-replicated volume that (usually) had very loose write
      permissions.

      The reason notes (or it's bastard offspring Lotus Notes)
      didn't take over the world was it put all the thinky bits in
      one place, probably locked down by a BOFH.  There's little we
      learned from the micro-computer revolution from the 70s and
      80s, but what we did seem to get is we all hated being
      dependent on that one guy from accounting to write a few
      COBOL programs for us so we could see what our sales figures
      looked like this month.

   It seems weird we're comparing PLATO NOTES with Lotus Notes
   email.  PNOTES (Personal Notes) was the equivalent of email.
   NOTES was more the equivalent of Usenet News.

   And Notes was an application environment w/ data replication
   that had a shim layer on top of it that looked like an email
   system.  It was more like Hypercard with git-ish semantics.
   (Which now that I think about it, wouldn't be that bad of a
   product.)

   But... after more than 50 years in tech, I still use Thunderbird
   as my email client.  If I could figure out a way to get it to
   turn off HTML mail I would.  I'm on the vanguard of the orange
   catholic butlerian jihad.  Not only do I seek a prohibition of
   thinking machines, I'm also quite suspcious of richly formatted
   text.

   Maybe the reason (Lotus) Notes failed was that people kept
   thinking of it as an email client.  Raymond Lowey talked about
   the "Most Advanced Yet Acceptable" principle.  Meaning you can
   push product design only so far before people no longer
   understand how they're expected to integrate it into their
   lives.  Lotus Notes was the poster child for this concept.

Ah. This is why one of my co-workers referred to our boss' Tesla as a "Suicide Booth."

Me too.

Looks like it came back up after about 4(?) hours. I noticed there was a "special offer" to extend my subscription for about half the price. I think what I want is to pay more for something that looks vaguely like an SLA. Or maybe even a description of how they have two data centers and their fail-over process is tested frequently.

Why bother with reliability when you don't have SLAs?

Do you know if they have multiple data centers and can fail-over when one barfs? I'm willing to pay a little extra for a SLA. Also, cyberinsider.com thinks Fastmail is in bed w/ the feds cause they have data centers in the states. Is that accurate? meh. I dunno.

https://cyberinsider.com/email/reviews/fastmail/

Thx for the info, though.

The "funny" think is I came to HN to search for "proton" to see if anyone had any more info and hn.algolia.com gave me a 502. It's a tough day to be a vibe coded internet service. Fortunately, the 502 was temporary and I was able to find your comment.

FWIW, here's my b*tchy post... https://news.ycombinator.com/item?id=49523682

I would email a link to this post to Bart (Butler), but my email service isn't working at the moment.

"The truth is where the sculptor's chisel stripped away the lie." Kusama-san definitely embodied that statement.

Several years ago I was at a conference in Las Vegas and discovered that one of her "infinity mirror rooms" was on display at... can't remember the exact location... maybe the Bellagio, but one of the major casinos along the strip. I eventually found it, but it was SO out of place -- A moment of art in a sea of crude commerce. Still... I was happy to have seen it. Major props to my spouse who has an uncanny ability to find interesting things, even in Vegas.

I was a twitter beta tester in 2005 when it was a SMS to intarwebs gateway. IMHO, it stopped being relevant sometime around 2008.

I stopped virtually all social media (except commenting on YouTube videos and telling people they're wrong in HN comments) years ago. No twitter, no mastadon, no bsky, no facebook. I keep hearing myspace is going to re-launch and I would consider getting an account there for the purposes of irony.

But... instead of tweeting for an hour a day, I spend that hour reading long form dead tree format texts. I'm not saying this is what you should do, just offering it as an observation that it is an option.

I feel very sad. My AI knows essentially nothing about be. The only thing it said it knew about me was my location, which it got wrong. It completely missed the books, magazine articles, RFCs and web pages I've authored. I did ask it about a very old project I wrote as a test of the github integration and it threw substantial shade that I was using 2010 era idiomatic javascript on a project that was written in 2010 and suggested I should have used ECMAScript modules that were introduced in 2015.

I am constantly amazed by what AI gets right... but sadly... equally amazed by what I can't seem to reason well about.

Just to throw a little shade back at Claude... OpenCyc would never have made that mistake.

Sure enough... When I follow the instructions from Anthropic about how to upgrade from free to pro, it keeps applying the purchase to API Credits (and doesn't send me an invoice.) I now have $39.95 of API credits because I paid for the pro upgrade twice and used $0.05 of credits through the dashboard / playground to ask claude how I pay for claude code.

Sure, but the free plan doesn't require the use of API credits. Why would it bill me for that when I hit the "upgrade" button. And, of course, there's no way to check what Anthropic thinks I bought, the support page offers no insight as to how to figure this out and the [email protected] email alias is unstaffed.

I checked with a friend. When they bought credits, they got an email receipt. Seems like credits would show up in the UI somewhere. But I did ask claude how to talk to a human at anthropic and it gave me an answer, so maybe it applied the payment to credits and the UI just doesn't show they're there.

THX! I was trying to remember this reference.

Also... When we moved from having multiple users per machine to multiple machines per user, services like finger on a specific machine didn't make much sense. It seems you should have one well-advertised machine that maps `[email protected]` to a `plan.html` file somewhere. (If a user has multiple machines, which one do you query to get their daily .plan?)

Maybe the answer is to define a plan base URL as a DNS TXT resource on a particular DNS domain. For instance, I added the string `< plan https://bi6.us/PL/ >` as a TXT resource at _plan.hamrick.rocks. It should now be a trivial task to write a program called `finger-ng` that when I type `finger-ng [email protected]`, it queries the TXT record at `_plan.hamrick.rocks`, extracts the plan base URL and then retrieves and displays a text document found at `https://bi6.us/PL/meadhbh%40hamrick.rocks`. I guess you could serve up a text/plain, text/html or even image/avif if you wanted to.

This, of course, assumes `bi6.us` is set up to allow me to update that file, but this is also straight-forward. Anyway, check back in a couple days and I'll have some code and I'll write the RFC/ID next week.

EDIT: I wrestled with apache and got it to understand files under www.bi6.us/PL/ should be served as text/dsd+plain, so you can hit `HTTPS://BI6.US/PL/MEADHBH%40HAMRICK.ROCKS` and get something that looks legit. And sure enough, HTTPS://BI6.US/PL/MEADHBH%40HAMRICK.ROCKS encodes to a smaller QR code than https://bi6.us/PL/meadhbh%40hamrick.rocks. (Though it looks like the HN text beautifier doesn't automagically produce links for URLs beginning with upper-case HTTPS.)

EDIT EDIT: Hmm... still doesn't say anything about privacy. What if you wanted to have a different _plan file for friends, a _plan file for family and a _plan file for co-workers? Clearly deep thoughts must be thunk.