David A. Harding

Haven't heard the pod but I tried OpenClaw and was pretty disappointed. It wasn't that useful without giving it access to all my data and accounts, and I didn't want to do that for privacy, security, and trust reasons. I'm also sticking with my desktop ai tool for now and accessing it from my phone to poke projects along (I'm using tailscale, mosh, and tmux for that).

2026-03-09 03:51:17 UTC

- reply

Been using it for six months as my main ai tool. Top uses have been for research for articles, website customization, and hobby programing. Probably 10 hrs a week, about $300 in OpenRouter credits a month.
I've looked at several alternatives and seriously tried one but none of them so far have been better for me.

2026-03-01 21:24:54 UTC

- reply

Congrats! Thanks for helping build a genuinely useful tool.

2026-02-26 05:52:08 UTC

- reply

I've used fossil for my personal projects in the past, it's actually really nice for solo projects or small teams. I can't imagine an agent would have any difficulty even if they had to figure it out entirely from its built in help text.

2026-02-02 23:46:21 UTC

- reply

Yes. I also sometimes use phone zoom to read small text when I'm too lazy to get my glasses or bring the text to proper lighting.

2026-01-27 05:37:06 UTC

- reply

Yeah, after the surprise, I realized it was probably just a link preview issue and I clicked the link. I'm not sure where to file an issue. I was using Amethyst but I assume it was just fairhfully rendering maps dot google dot com. Google was just displaying its default, which is my current location. I'm guessing its just a quirk that I'll have to get used to.

2026-01-26 20:39:59 UTC

- reply

Was super creeped out for a moment. My nostr client showed me a Google maps preview for your link, but rather than showing me the location you pinned, the preview showed my current location. It looked like you were either doxing me or staying in the AirBnB next door.

2026-01-10 04:03:43 UTC

- reply

I used to have an air gapped master key with a hot subkey that I used for everything besides key attestation. In that setup, dealing with any sort of tooling was a nightmare, so I gave up on it.

2025-12-10 19:43:14 UTC

- reply

The filesystem is itself a single file.

2025-12-01 05:01:04 UTC

- reply

What's the name for the emotion when you ask it to do something and then come back later to discover it wrecked things and you forgot to git commit the previous improvements?
(That makes me wonder if I could set up an every minute cronjob to create btrfs snapshots or something.)

2025-11-22 19:28:36 UTC

- reply

What does "global liquidity valve" mean? Is it something like: when people have excess cash, they buy bitcoin; when they need cash, they sell bitcoin?

2025-11-22 07:46:09 UTC

- reply

Lol

2025-11-21 22:11:21 UTC

Finally had my first chat with someone on bitchat (i live in an area where "region" may have less than a million people). Worked great.

2025-11-11 05:34:24 UTC

- reply

Excellent choices! Older but still great is Star Trek II by the late James Horner at the beginning of his career; Battle in the Mutara Nebula remains my favorite spaceship fight song.
Gattaca by M. Nyman is an absolute masterpiece of subtle and complex layering; I saw on a behind the scenes that songs from it were also frequently used as the stand-in music during the filming and editing of the TV show Firefly. The track The Other Side is my most replayed song of all time.
Ennio Morricone is best known for his westerns, but I think his best soundtrack is The Missionary, in which music plays a key part of the plot. Gabriel's Oboe is a favorite song to play for a great many musicians; I've heard it on everything from sax to cello and played it myself (poorly!) on flute.

2025-11-08 02:12:21 UTC

- reply

I use Darktable almost every day, it's excellent. I think its devs are often frustrated by it being called a Lightroom replacement as Darkroom has pioneered multiple improvements in RAW processing, from a powerful (non-ai) masking system to significantly improved color science.

2025-11-07 20:31:00 UTC

- reply

Maybe a little bit. I think it's mostly because anti-social behavior online is often rewarded (e.g., through increased engagement), whereas offline it's often punished (e.g., through decreased engagement as people avoid assholes).

2025-11-07 20:11:41 UTC

The vast majority of anti-social behavior I've witnessed in my life has been on so-called social media.

2025-11-07 18:06:57 UTC

- reply

Four years to breakeven ROI seems good, and I'd rather write code than do rote work even if it takes moderately longer to write the code than to just do the work.

2025-11-07 10:33:10 UTC

- reply

About how many hours of coding in that two weeks (versus, e.g., collecting and entering documentation like you'd have to do anyway if you used a third party service)?
Taxes for me (USA/Hawaii) take me about 8 working hours a year using an online service that is sometimes more annoying than helpful and which has politically advocated for keeping taxes complicated, so I'd love to replace them even if it didn't save me a few hundred bucks. But I'm not sure it's worth more than an extra ~8 hours of my time.
However, even if I don't go all they way in like you, you've inspired me to revamp some of my personal accounting to take advantage of AI. I've been just doing the same thing I did five and ten years ago without thinking about how new tools could make my process less cumbersome. Thanks!

2025-11-05 03:22:51 UTC

- reply

Who gave him a grant?
Thanks for making your disclosure responsibly.

2025-11-04 01:45:35 UTC

- reply

"Argh, I'll just fix it myself!"
<Looks at code>
"This is shit code, I don't want to figure out how it works!"
<Asks chat again to fix it>

2025-11-04 01:41:37 UTC

I wish there was a way to split test my life. Today, in particular, I would like to meet the version of me who didn't try to vibe code an Arduino project and has now spent several hours thinking I was just one more chat query away from satisfaction.

2025-10-31 18:21:38 UTC

Unplanned shot. I was taking photos of surfers when I heard this plane coming, looked up, and noticed it would pass near the moon. I had about two seconds to zoom out, frame the moon, and set my focus to infinity before the plane arrived.
That it worked is pure serendipity.
https://image.nostr.build/6f8884ef7a5dbfd69a8dc007583d04d85a23fae0ba1b8796026cb861661c97e5.jpg

2025-09-29 01:23:19 UTC

- reply

Well, yeah. Duh.
Guessing the actuarial tables say I have about 40 years left. That's an estimated hourly cost of $10 in real terms. Alternatively, assuming I have to pay upfront and if I believed I could earn a wild 20% return per year over inflation on that $2,000 over 40 years, it's an hourly cost of about $147, which is more than I should pay given my financial standing but not crazy high.
My caveats would be that I'd have to have a reasonable belief that it actually would save the time, it'd have no significant downsides, and it'd not take away one of the mindless tasks I find relaxing and enjoyable.

2025-09-26 13:31:21 UTC

- reply

Is the implication meant to be that Bitcoin Core isn't conservative, well reviewed, and/or well maintained? If so, why do you think that?
Or is your desire to have additional implementations that are conservative, well reviewed, and well maintained? If so, do you think that's a better use of limited developer talent than focusing on making Bitcoin Core better?
When I read your post, I get the feeling that you think about "Core" and "Knots" as teams that each move with a single shared purpose, as if under the direction of an individual. That might be true for Knots, which effectively has a single developer, but it's not how I see the Bitcoin Core project, whose contributors are often in disagreement with each other (if usually only about the best way to achieve a particular goal). Maintaining that environment where independent contributors can easily collaborate, are free to express their differing opinions, and can create single-topic software forks (like PT's librerelay or the BIP148 activation client) if they feel out of alignment with the other devs seems to me like a better use of scarce talent than creating more implementations for the purpose of having more teams.

2025-09-24 20:11:59 UTC

- reply

Never interacted with him AFAIK but it was my understanding back in the day that Luke transferred ownership of Eligus to someone operating under the wizkid nym.
A possible explanation for those tweets is that Wizkid became salty about Bitcoin after Eligus became irrelevant.

2025-09-09 15:23:41 UTC

I once read a book about sociopaths and it said that the top sign that you're dealing with a sociopath is that you consistently leave discussions with them feeling confused and wondering if you're crazy.
The reason for that is simple: part of how we try to comprehend other people is by modeling them in our head. For earnest people with straightforward motivations that are similar to our own motivations, that modeling is easy and we feel comfortable satisfaction each time our mental model helps us predict their behavior. But for people who are deranged and have ulterior motivations, developing a model is difficult (or even impossible for those of us incapable of significant malace or deceit) and we become profoundly uncomfortable when we are unable to make predictions about their behavior.
Of course, its possible to sometimes leave conversations with good people feeling confusion and self doubt, and also a lack of those feelings doesn't prove someone isn't a sociopath, but I find it useful to privately contemplate whether someone might be a sociopath if they consistently make me feel like I might be crazy.

2025-09-08 05:49:51 UTC

https://image.nostr.build/922fded3a7381b0c42adc2a3a5b8c6cc8bfa369d30ca759e5b9410d9fa12723f.jpg

2025-09-07 06:58:42 UTC

- reply

Good suggestion. Message to her sent.
Related, I've never told you how impressed I've been by all of your work. It's been a joy to watch your Cashu work as a Bitcoiner but it's also been a pleasure to follow your other projects as a Nostrich and a general lover of freedom and privacy technology. Thank you!

2025-09-07 06:08:55 UTC

- reply

Thanks. I wondered if it was something like that.
It would be very unfortunate if they succeeded. Gloria has done a great job as both a dev and as a maintainer.

2025-09-07 01:03:40 UTC

- reply

I've not been on hiatus but I'm guessing this is some Twitter drama, so I also don't know what's going on.

2025-09-05 19:25:26 UTC

- reply

I've spent way more than that on home contractors who had similar or worse literacy problems, and I've generally been happy. A top evaluation criteria for me is the insightfulness and forethought of the questions someone asks. If they're actively thinking about potential problems, I wouldn't care much about their literacy. If they're just just telling you "ok boss", I wouldn't hire them even if they had a PhD.

2025-08-31 07:15:43 UTC

- reply

I think it's an advantage of BIP93 that it doesn't use a natural language encoding, as bech32 discourages people from trying to rely on memorization, but I figure someone will eventually make a transliteration scheme and it might as well be someone competent.

2025-08-28 02:56:38 UTC

- reply

Why do I need to trust the relay operator? With what am I entrusting them? Is there a document that explains at least the basics of what Amthest is doing / wants me to do?

2025-08-28 01:43:33 UTC

- reply

I would like help, yes, but not personalized help. I'd like a document I can read or a button I can push in Amthest or a set of term I can query an LLM about or something else along those lines.
Sorry my message was snarky. I gave into the dark side rather than composing with compassion. I'm very appreciative of all your work and I use Amthest because it's the best Nostr client I've found.

2025-08-28 01:37:41 UTC

Congrats to everyone who figured it out. Too bad that doesn't include me. No worries though, probably my lack of relay configuration means that nobody will see this message so Amethyst's upgrade will look like a success due to survivorship bias.
#nevent1q…6e0e

2025-08-26 03:54:12 UTC

- reply

For anyone considering this, Jimmy was the person who recommended me to O'Reilly for the third edition of Mastering Bitcoin. Everyone over there loved him and took his recommendations seriously.

2025-08-24 16:37:15 UTC

- reply

If the text you send is useful, why should anyone care who wrote it?

2025-08-22 02:30:08 UTC

- reply

What is a community? (I had the bug in the old version, so I guess I don't have a community. Now I have FOMO.)

2025-08-19 03:38:23 UTC

https://image.nostr.build/1587e5047fb1d33536e1a099201b6521c9b230224779ed61aa98234676711a77.jpg

2025-08-18 02:40:11 UTC

- reply

Rope sniper! (My neighbor is an arborist and former fire jumper.)

2025-08-17 09:20:10 UTC

https://image.nostr.build/e7bf9d29e983400b65cb73baa6da46ffedccee978f6c97a6928c151e19c9d88c.jpg

2025-08-05 00:11:11 UTC

- reply

This is true for me.

2025-08-02 19:12:49 UTC

https://image.nostr.build/1988d9442208bb26fe4263453852ec14cfea9348b72d9d3b63fce90509964267.jpg

2025-08-01 22:45:20 UTC

- reply

The law says they can't dump solid waste within three miles of shore. Whether or not they follow the law is unknown.

2025-08-01 18:12:09 UTC

https://image.nostr.build/ec8dd36533be7bc68282047976968f3299ba3ecd4af8c13788c66273f6166f85.jpg

2025-08-01 01:16:37 UTC

https://image.nostr.build/e18fe6c56de30d9aecf792ea92b12af43dfbd8848b477f64d61c7c47903ddbc6.jpg

2025-07-30 13:08:15 UTC

- reply

The tsunami hit about two hours after high tide. Happily, the sea state has been calm for a couple days so it didn't amplify the existing waves.

2025-07-30 02:50:36 UTC

- reply

Cops just evicted everyone from the surf beach in Hilo.

2025-07-30 02:12:40 UTC

- reply

Hawaii resident here, can confirm also on Big Island. Thanks for covering this!

2025-07-25 17:42:51 UTC

https://image.nostr.build/539db51984501e39c6d4e510f60d881e95ea39d9ddbe784dd2d97c676fd52662.jpg

2025-07-23 13:13:22 UTC

- reply

I'd ask AI to write simple-to-understand code to virtually replicate how you rolled the dice a few billions of times and then calculate how many of the results matched what you saw. Then review the code to see if it seems to match what you expect and run it.
Even when I calculate odds myself, I often write a quick script to brute force an approximation for assurance. CPU cycles are cheap.

2025-07-23 12:59:54 UTC

https://image.nostr.build/5988c03ae578e21730f8c868161e0933bfc4845fa467a13040d9483fefdf316f.jpg

2025-07-21 16:16:58 UTC

https://image.nostr.build/7c756fcf9edd4ee602bc5f0a29618f41f35e0bcf35fcb1be76a643b944f85099.jpg

2025-07-17 02:35:16 UTC

- reply

In the back, your odds of surviving a flight stay go from 0.9999998 to 0.9999999 but your odds of getting off the plane within 5 minutes of arriving at the terminal go from 0.99 to 0.05.
NB: I made up these numbers.

2025-07-15 19:37:57 UTC

https://image.nostr.build/cec7f4270128ceaba4eea7650f25b72eb0840c49731d0e578d093014108eb011.jpg

2025-07-14 17:34:00 UTC

https://image.nostr.build/f37fbd41c7a10020fe4cb994dbd42f97e86ed2c34fc51744ee8cdf1ea22a4e13.jpg

2025-07-13 17:51:57 UTC

https://image.nostr.build/b58eb25c765db23c5028cbf16a57a804ceb6682c47081e57d4d4832249c64a36.jpg

2025-07-13 04:13:16 UTC

- reply

I'm the photographer.

2025-07-12 18:20:57 UTC

https://image.nostr.build/78b6b3c0e101f5cc8b369e964b3625c2d5af8f2cde955932c8fe211aea404040.jpg

2025-07-11 16:04:46 UTC

https://image.nostr.build/9aca4070d24b1f5fe46f2377893350c0b2c4e5b28eaeb84d5a3b0d69b9c8bdb4.jpg

2025-07-08 17:18:57 UTC

- reply

Dune. I wish I could read it the first time again twice. Once again like I did when I was a kid and took it at face value, being simply impressed by its depth and scope. Then once again like when I was older and realized that the book was really about the abuse of power, not just of the obvious villains but also of those whose seize power to defeat the villains but, in doing so, cause far more harm than the nominal villains ever did.

2025-07-07 15:18:49 UTC

- reply

Started a membership; you should get the referral bonus. Thanks for the recommendation!

2025-06-08 19:50:09 UTC

- reply

What did I say that you're debating? I'm guessing it was something about JM being serverless or P2P?

2025-06-02 19:34:05 UTC

- reply

Can't tellwhether this is an allusion to scifi or ancient fortifications.

2025-05-26 03:45:20 UTC

- reply

No, I'm talking about his Tunable Penalty Protocol. My summary here with links: https://bitcoinops.org/en/newsletters/2023/03/29/#preventing-stranded-capital-with-multiparty-channels-and-channel-factories

2025-05-25 22:05:35 UTC

- reply

Why not use BIPs? What's the value of standardizing them elsewhere?

2025-05-24 01:52:39 UTC

- reply

How does it compare to John Law's factory designs that also don't require soft forks?

2025-05-17 04:58:25 UTC

- reply

If Bitcoin is successful, that will have to happen eventually. There's only 300,000 sats per person in the world (less than that if we assume some BTC have been lost). People have already tried to address that perceived insufficiency through things like millisats. When we fix it at the consensus layer, which will practically require a hard fork, it would be nice to address the problem forever by allowing arbitrary precision, e.g. allowing output amounts to be defined as a fraction of MAX_MONEY. If that happens, the idea of base units goes away
If you assume that we'll never add inflation to Bitcoin, then the monetary constant is the supply, not the fraction of it that we currently transact in.

2025-05-16 22:18:54 UTC

- reply

Doesn't changing it once imply that it might be changed again? I don't want to spend the rest of my life listening to people arguing back and forth about whether it's time to redefine the basis again---that doesn't seem very clean to me.

2025-05-13 18:04:28 UTC

- reply

Additionally to the other responses in this thread, an increasing amount of Bitcoin P2P traffic should be encrypted, which makes it harder to analyze with network tcpdump style tools.

2025-05-11 18:48:51 UTC

- reply

Other nodes will continue following the most PoW valid chain but old Knots will reject that chain and instead accept backdated blocks on a lower PoW chain.

2025-05-11 18:44:04 UTC

- reply

The logic of that function seems pretty weird. Why use nTime instead of system time? Why reject blocks instead of shutting down (or simply nagging the user)? The way it's designed seems possible to abuse, whereas other easy choices would have much lower risk of problems.

2025-05-02 18:05:52 UTC

Thank you to everyone on nostr who has been patiently explaining the technical reasoning behind considering increasing or removing Bitcoin Core's default OP_RETURN size limit.

2025-04-19 21:33:52 UTC

- reply

Pretty sure LND uses it for both simple taproot channels and the latest version of their swap service, Lightning Loop.

2025-04-15 22:05:19 UTC

- reply

When I first started writing Bitcoin documentation, I asked my mentor at the time (Mike Hearn) whether merkle trees were important to understand. 🤦‍♂️

2025-04-15 22:01:45 UTC

- reply

Mine has an odometer, so technically it tracks me. But it has no network access, so I doubt it's reporting on me.

2025-04-07 21:46:04 UTC

- reply

Tl;dr, but an AI would read it, so I guess they're winning.

2025-04-07 03:12:02 UTC

- reply

No, when you buy something for $100, the merchant only receives $95 to $99. The bank keeps the difference.

2025-04-06 20:54:52 UTC

@nprofile…mh4h saw a post by you on Twitter talking about how credit cards don't charge for loans unless repayment is late so why should Bitcoin loan orgs? But CCs do charge 1-5% to the seller for the initial ~30 day loan period, which is equivalent to a 13% to 80% APR if you were to compound that period 12x in a year. That's about the same rate as Bitcoin loans.

2025-04-01 21:04:32 UTC

- reply

Also, wherever there's an argument, there's certain to be somebody linking to allegedly relevant material buried deep in a Twitter thread. 💩

2025-04-01 20:44:56 UTC

The hardest Optech newsletters to write are those where we summarize unresolved disagreements, e.g. for or against CTV+CSFS. When everyone ends up in agreement, at least mostly, we can skip describing any arguments and jump to the conclusion.
But when there's no resolution, we have to get into the details, try not to miss anything important, and be careful not to summarize in a way that misstates someone's key argument or attributes to them an opinion that they don't actually hold. On top off that, there are arguments that seem weak to me but which probably warrant describing in case I'm missing something. All that extra work is important but it's not my favorite.
Also, it makes for long and hard to read newsletters, so sorry about that for next week.

2025-04-01 00:40:18 UTC

- reply

Instead of having to whitelist people, you could probably have a simple naive Bayesian classifier in your client that you train on the type of comments you don't like. It would hide future similar comments in the future, with you able to tweak the threshold as your mood changed (e.g., you could risk seeing more junk on a day where you were bored). You could still whitelist your friends so they would pass through the filter.

2025-03-31 19:18:50 UTC

- reply

Yrah, but there's no social aspect to Twitter muting. What you described (as I understood it) is Alice's client influencing which posts Bob sees. In that case, once one person we like violates the protocol, it incentivizes other people in the social group to violate the protocol, quickly rendering it moot.

2025-03-31 18:50:33 UTC

- reply

Opt in ignoring in a social contexts always eventually degrades to nobody opting in. Imagine you have Alice and Bob, who like each other, and Unlikable Urkle. Alice initially ignores Urkle but Bob replies to one of his comments, stoking her curiosity, so she unignores him. This is why /ignore on IRC and similar systems don't work.

2025-03-23 18:52:07 UTC

- reply

What crucial role do they play? (Honest question.) For content creators, I think it's probably better to convert readers into subscribers (e.g. BOLT12 recurring invoices), in the patreon model. For content consumers, the patronage model is also probably better at intelligently funding the kind of stuff that's good for you rather than a dopaminergic ad hoc process of zapping.
For algorithms, my understanding is that zaps are just an unverifiable statement between two people, e.g. you and I could sign a zap saying that one of us paid the other 1M BTC without even a single sat changing hands. If they were verifiable, they'd seem to give algorithmic advantage to rich people and that doesn't particularly interest me.

2025-03-23 18:34:45 UTC

- reply

Re: not using LNURL, it's mainly because I don't like hosting dynamic website code. It requires installing stuff on a server (requiring either a dedicated server or taking a risk with other stuff on the same server), breaks often in my experience, and is just not what I want to spend my time on. BIP353 looks promising because I'll just be able add a field to my existing DNS.
All that said, I haven't actually set up either LNURL or BIP353, so I could be confused about what they actually entail.

2025-03-23 18:18:08 UTC

- reply

Agreed about the possible use cases. And I also realize that I probably oversimplified things in my previous post. A normal part of my conversations, which are maybe not normal conversations for other people 😜, is suggesting that we make small wagers when we disagree about something easily provable either now or in the near future (my wife and I even use a website, fatebook.io , that let's us collaborate on questions and compare our Brier scores). Zaps would be great for resolving those small bets, rather than having to email LN invoices around like I did for https://x.com/jxpcsnmz/status/1839444666988757306

2025-03-23 17:58:01 UTC

- reply

Totally agree that there are a bunch of confounding variables that limit correlation. However, I think every other proxy we've had for personal key management has failed to achieve popular use: PGP, personal browser certificates, years of DID attempts, etc. So the default assumption at this point might need to be that everyday people can't or won't manage their own keys. Accepting that assumption would have a significant effect on either _how_ we build Bitcoin software or _who_ we build it for. If, instead, we want to continue to build for an anticipated highly self custodied future, it would be really nice to have Nostr to point to as a counterexample of lots of typical people controlling their own keys.

2025-03-23 03:40:42 UTC

- reply

AFAIK, zapping currently requires using LNURL, and I'm not interested in either self hosting that infrastructure or outsourcing it. When BOLT12 zapping (perhaps through BIP353 addresses) becomes a thing, I may set it up.
Tbh, though, mixing money and social interactions doesn't excite me. I'm not opposed to it, but to me the currency of a good conversation isn't money---its acknowledgement in the form of smiles, laughs, retorts, and the like.

2025-03-22 22:02:09 UTC

Controlling your own key on Bitcoin comes with significant costs, from the cost to create a UTXO, to the eventual cost to spend it, to the potentially high cost of losing your key and the money associated with it.
On nostr, controlling your own key has basically zero cost. It's one of my favorite things about nostr. If widespread control of personal keys can be made to work here, there's hope for widespread control of personal keys on Bitcoin. But if nostr fails, then I think it weakens the standard case for Bitcoin.
That may sound negative, but I think it's great to be able to test individual properties of Bitcoin separately from the heavy confounding factor of it being money.

2025-03-19 17:53:38 UTC

- reply

No, but I take a cheap unflavored nicotine lozenge each time I ride my indoor bike rollers for habit reinforcement.
I think it helps but I wonder if there's an easy way to blind test it (n=1). Maybe having my wife apply nicotine patches vs similar sized bandaids to my back?

2025-03-15 00:11:38 UTC

- reply

I've [thought](https://github.com/jamesob/assumeutxo-docs/blob/master/proposal/feedback-david-harding.txt#L172 ) since the start that the background validation added a lot of complication for not much benefit (maybe even negative benefit). It seems to me like it would be a lot simpler to just allow loading a UTXO snapshot or sync from genesis but not try to do both.

2025-03-11 18:40:11 UTC

- reply

Hawaii residents unaffected

2025-02-14 03:23:53 UTC

- reply

After much hand wringing and signing of forms, I was allowed to "withdraw" it as cash (no cash was actually touched) and made a cash "deposit" into the escrow account. Honestly scares me to be performing ad hoc rituals with a big chunk of our savings.

2025-02-14 02:04:58 UTC

- reply

You would not believe how normie this situation is. My wife and I are buying a home and the escrow company required the final deposit to be made by wire transfer. However, I use the same local bank as the escrow company and they don't allow internal wires.
Working around that situation has required multiple emails, multiple calls, an in person bank visit, and both my wife and I to sign a specially prepared document. Everyone has been really nice but also acting like they're doing me a favor. I've managed not to shout at any of the clerks who had nothing to do with failing to anticipate this obvious edge case, but I had to vent somewhere.

2025-02-14 00:42:30 UTC

I fucking hate dealing with banks.
Most days if you ask me why I work on Bitcoin, I'll probably say something about eliminating the temptation for the state to print money, with all of its disastrous effects.
But every time I have to deal with a bank and must grovel to use my own fucking money, I feel the urge to hoist the black flag and start slitting throats. On days like today my work writing about Bitcoin technology is fueled entirely by rage and spite.