I'm surprised by the explanation of the 8 in the "real de a a ocho" because "traders counted gold doubloons on their fingers, skipping their thumbs." (and the link to investopedia has a similar explanation).
> Spanish American gold coins were minted in one-half, one, two, four, and eight escudo denominations, with each escudo worth around two Spanish dollars or $2. The two-escudo (or $4 coin) was the "doubloon" or "pistole", and the large eight-escudo (or $16) was a "quadruple pistole"
I think it makes more sense that some time ago it was possible to split some coins in half and quarters, so someone decide to continue the tradition and use base 2 to move up.
Ever tried cutting a cake? It’s a lot easier to visually judge half of a circle segment. You’d need a compass to get accurate tenths (or fifths) and I imagine it is generally frowned upon if some tenths are a lot smaller than others (happens a lot with cake)
I committed to the GnuCash codebase pretty regularly in the 1999-2002 era... I think maybe I actually implemented the fractional representation that the article discusses? Not sure, it was a long time ago! I definitely remember receiving some very heated emails about how this was total nonsense and there was no reason to do anything other than a decimal representation. The phrase "a superhighway of abstraction, leading nowhere" has stuck with me for lo these many years :) good times
hm, I don't have any other ways to prove it. The thing is - I thought this is something LLM can't write about.
Just imagine a prompt: "Hey Claude, go ahead and come up with idea why GnuCash stores numbers as fractions and come up with an article for HN". I actually tried it and god damn thing came up with something very similar :D
I am a native English speaker, and I find accusations of LLM-writing exceedingly annoying – to the point where I sometimes intentionally write in that style, just so I can hit back with a profanity.
I used GNUCash years ago in Argentina while we had high inflation. Some operations were in local currency and other are Dollars. The currency exchange changing hourly. Tracking finance is a nightmare, since you basically need an exchange rate for every operation.
I would like to use finance tracking products like GNUCash. But I don't have the patience to download the csv for half a dozen every month (Products like plaid are a no go from a basic security perspective). I am in Canada, and there seems to be no hope that I will have API access to my bank accounts anytime soon.
Also, did I mention how much it annoys me that the transaction description differs between the CSV and the PDF statement for pretty much all banks I use.
I'm using self-hosted Sure.am and also using SimpleFin to connect to Canadian banks. It works, but barely, since it effectively scrapes with no real API access. I have to login daily to update 2-FA on various accounts, and have suffered account lockouts a couple of times, due to "suspicious activity".
But it still beats downloading multiple exports from the bank and importing it manually...
There needs to be a lot of investment in training and safe defaults though. Most people are not ready to automate even a little of their banking like that.
I would even prefer banks had the option to push data to trusted feeds than having open APIs you could call on your own.
Your title promises a story about the spanish traders. It does not deliver, but talks about 3 other different topics that I'm not interested in and jumping between them to test my nerves. If it is about the architecture of whatever you are doing, put the architecture in the title and don't make your title an offhand comment somewhere there. Fucked up clickbite.
Yeah, I got an AI vibe off it too, and was surprised to find this. The problem is, to quote Orson Welles, "it's not as conversationally written. It's full if things that are only correct because they're grammatical, but it's tough on the ear, you see." AI writing bears a resemblance to the most insufferable marketing speak, so if you ape that kind of writing in an attempt to sound punchy or whatever, you're going to be accused of being a bot.
That would explain a lot. Grammarly went full LLM brainrot a few years back, so if you rely on its suggestions your writing will... sound a certain way.
I strongly disagree. I found the whole article interesting and enlightening – I certainly wasn't aware of the topic before, and I'm glad it was posted on HN.
Furthermore, it didn't feel LLM-generated to me. Quirky, yes; nothing wrong with that.
Yeah this one is my attempt to write without LLMs rewriting my thoughts.
P.S. Came to the decision after going through https://thebullshitmachines.com/
Side question:
I'm surprised by the explanation of the 8 in the "real de a a ocho" because "traders counted gold doubloons on their fingers, skipping their thumbs." (and the link to investopedia has a similar explanation).
But from https://en.wikipedia.org/wiki/Doubloon
> Spanish American gold coins were minted in one-half, one, two, four, and eight escudo denominations, with each escudo worth around two Spanish dollars or $2. The two-escudo (or $4 coin) was the "doubloon" or "pistole", and the large eight-escudo (or $16) was a "quadruple pistole"
I think it makes more sense that some time ago it was possible to split some coins in half and quarters, so someone decide to continue the tradition and use base 2 to move up.
Yeah, I thought about it when I first saw the coin cut into pieces - https://www.pirateglossary.com/glossary/pieces-of-eight
But then why didn't they cut it into 10 pieces - https://etc.usf.edu/clipart/40600/40610/pie_01-10a_40610.htm ?
> But then why didn't they cut it into 10 pieces - https://etc.usf.edu/clipart/40600/40610/pie_01-10a_40610.htm ?
Ever tried cutting a cake? It’s a lot easier to visually judge half of a circle segment. You’d need a compass to get accurate tenths (or fifths) and I imagine it is generally frowned upon if some tenths are a lot smaller than others (happens a lot with cake)
Easier to cut into halves?
I committed to the GnuCash codebase pretty regularly in the 1999-2002 era... I think maybe I actually implemented the fractional representation that the article discusses? Not sure, it was a long time ago! I definitely remember receiving some very heated emails about how this was total nonsense and there was no reason to do anything other than a decimal representation. The phrase "a superhighway of abstraction, leading nowhere" has stuck with me for lo these many years :) good times
I wonder if Hackernews ranking algorithm has been updated to exclude comments toing and froing about whether or not the article is LLM generated!
haha, true.
Thing is - I'm not a English speaker. But I chat a lot with Claude/ChatGPT - i feel like I'm picking the style from them unintentionally.
Doesn't sound like reality to me. The article looks very much AI-generated. Nothing to do with not being an English native speaker.
hm, I don't have any other ways to prove it. The thing is - I thought this is something LLM can't write about.
Just imagine a prompt: "Hey Claude, go ahead and come up with idea why GnuCash stores numbers as fractions and come up with an article for HN". I actually tried it and god damn thing came up with something very similar :D
I am a native English speaker, and I find accusations of LLM-writing exceedingly annoying – to the point where I sometimes intentionally write in that style, just so I can hit back with a profanity.
Yeah, I'm just a developer mostly chatting with LLMs. I hope I'll be develop a distinct style some day.
I used GNUCash years ago in Argentina while we had high inflation. Some operations were in local currency and other are Dollars. The currency exchange changing hourly. Tracking finance is a nightmare, since you basically need an exchange rate for every operation.
How's the inflation now with Milei's changes? I know charts show its at a 10 year low but reality could feel different.
Lived through something like this after USSR collapsed and before Hryvnya was introduced. Boy I remember the bread cost - 10,000,000 :D
I would like to use finance tracking products like GNUCash. But I don't have the patience to download the csv for half a dozen every month (Products like plaid are a no go from a basic security perspective). I am in Canada, and there seems to be no hope that I will have API access to my bank accounts anytime soon.
Also, did I mention how much it annoys me that the transaction description differs between the CSV and the PDF statement for pretty much all banks I use.
I'm using self-hosted Sure.am and also using SimpleFin to connect to Canadian banks. It works, but barely, since it effectively scrapes with no real API access. I have to login daily to update 2-FA on various accounts, and have suffered account lockouts a couple of times, due to "suspicious activity".
But it still beats downloading multiple exports from the bank and importing it manually...
Agree. I wish banks have their APIs open. It's 2026 and we have OIDC.
But I used to actually pull the CSV once a week and feed it to GnuCash. It's pretty good at auto-categorization.
Also I simplified my finances to only a couple of checking accounts and only one credit account (for car rentals).
> I wish banks have their APIs open
There needs to be a lot of investment in training and safe defaults though. Most people are not ready to automate even a little of their banking like that.
I would even prefer banks had the option to push data to trusted feeds than having open APIs you could call on your own.
I simplified to just 2 accounts and I enter them manually twice a week. I keep a detailed budget so it helps to do it that frequently.
Do you do budget in GnuCash or somewhere else?
Feels ai generated and waste of time to read even though the topic could be interesting.
I think English is not the first language of the author ("we are counting fingers and thumbs"?).
There's also a `Surprisingly written by a human :)` at the bottom.
yep, I'm Ukrainian
> Surprisingly written by a human :)
Article ends with this
A true surprise if so, but still low in informational density and annoying to read.
huh, sorry I'm really bad at writing - this is one of my first attempts
I really enjoyed the read. Not everything needs to be some sort of utilitarian information density optimized reading piece.
Keep up the good work!
Thanks!
You're not bad at writing. You have your own style. Keep it up! This was an interesting article.
Your title promises a story about the spanish traders. It does not deliver, but talks about 3 other different topics that I'm not interested in and jumping between them to test my nerves. If it is about the architecture of whatever you are doing, put the architecture in the title and don't make your title an offhand comment somewhere there. Fucked up clickbite.
Yeah, sorry. It's kind of intentional. Did not want to give it a boring name like "Storing fractions vs decimals".
Clickbait is a thing and pisses people off.
Yeah, I got an AI vibe off it too, and was surprised to find this. The problem is, to quote Orson Welles, "it's not as conversationally written. It's full if things that are only correct because they're grammatical, but it's tough on the ear, you see." AI writing bears a resemblance to the most insufferable marketing speak, so if you ape that kind of writing in an attempt to sound punchy or whatever, you're going to be accused of being a bot.
I think it's because thought-process is done in Russian, translated into English and the checked for grammar mistakes with Grammarly.
That would explain a lot. Grammarly went full LLM brainrot a few years back, so if you rely on its suggestions your writing will... sound a certain way.
I strongly disagree. I found the whole article interesting and enlightening – I certainly wasn't aware of the topic before, and I'm glad it was posted on HN.
Furthermore, it didn't feel LLM-generated to me. Quirky, yes; nothing wrong with that.
Thanks! I believe blogs and indie products should have personality - otherwise it so easy to overlook them.
I think its not.
It is :) No AI slop. I'm intentionally trying to write without LLM (my previous blog posts clearly disclose it).
(you mean it's _not_ AI generated)
Yeah this one is my attempt to write without LLMs rewriting my thoughts. P.S. Came to the decision after going through https://thebullshitmachines.com/
Curious what made you think it's AI slop?