bjorney

All this bill is telling industry is “if my product includes an uptime SLA, don’t build new data centers in Texas”, which is ironic, because that’s the exact customer Texas is trying to attract

Southern Australian springtime swimmers who properly update their Bayesian priors know that sharks are the true danger, NOT lightning strikes or plane crashes

“[…] In exchange for a waiver of fees accrued since 2023”

Sounds like Oracle got them with the good 'ol “buy an even bigger license or we’ll sue you”

Noooooooo

Newfoundland is “Newfoundland Time”, NOT “Atlantic Standard Time”

NL has a half hour time difference from the rest of Atlantic Canada

But AMD has been making leaps and bounds improving their GPU software

They are still largely shitting the bed here. Their ROCm installer won’t run on Ubuntu 25.04 last time I checked, and the 9070xt won’t work on OSs that ROCm DOES support because the kernel and graphics stack is too old.

ROCm has been “almost ready” to be a drop-in replacement for CUDA for almost a decade. I feel like it literally would take nvidia ceasing to exist to give them the critical mass to push it over the finish line

It’s a government institution that is set up like a normal corporation, but with the government as the shareholder. If that’s not an ass backwards way of providing an essential service I don’t know what is.

Counterpoint: by operating at arm’s length it can’t be steered on a whim by a sitting government, e.g. like DeJoy grinding USPS to a half during the 2020 election. Same can be said about CBC

Ah, yeah it is.

The “sign up to read this article” pop-up on mobile obscures literally everything except the article title and yesterday’s date for me, so I assumed it would have been about yesterday’s Eurovision final lol

Didn’t watch this year, but got the impression this was happening last year as well. Whenever there was ANY mention of Israel by the hosts, the audio became murky, then crystal clear all of a sudden because they had to cut the audience track out completely.

I feel like they must have had a separate room/section mic’d up specially for “Israel applause”

They broke the co-op to sell themselves to private equity

They were already under creditor protection when the sale happened. You can blame the old board for bankrupting the co-op, but the private equity acquisition was handled by the courts

> want to compile 50kb C++ console app on windows

> 6 GB MSVC installation

Lenovo apparently offers the choice on some models, with the windows license adding $140 to the price of the laptop.

We’re at a point where it’s no longer profitable for individual miners

We have been at that point since GPU mining stopped being feasible in 2014, it’s just gotten worse. ASICs made it so the only people who could profit off mining were people who could place a wholesale sized order of hardware from bitmain, etc. Anyone else who claimed to be mining profitably was likely someone who was:

1. buying old hardware 2nd hand (or new hardware at MSRP) and capitalizing on free electricity in their rental
2. not selling their Bitcoin immediately (they weren’t making money from mining, they were making it from speculating)
3. lived in Quebec and could double dip (North America’s cheapest grid + free heating for 8 months of the year)

unless there’s a radical change in bitcoin’s algorithm

The algorithm already does this though. Every 2016 blocks if it took more than 10 minutes per block, the difficulty of mining bitcoin goes down, not up. This is why every halving event you see a radical drop in difficulty, because at a given kWh you are producing half as many bitcoin - meaning people turned off their miners because it’s less profitable. The flipside is the rate of issuance goes down, so there is a lower inflationary effect, and the price of Bitcoin usually also skyrockets (which means eventually these miners re-enter, and difficulty eventually goes back to where it was). It can never get to a point where Bitcoin mining is completely unprofitable unless the price goes to zero, because there will always be a guy with a solar panel and fully paid-off hardware who can mine it for free. Granted, it can get to a point where a lot of people have to take a huge loss on capital expenditures if the price nosedives and never recovers

Miners like Riot Blockchain are operating at a loss

I’m not a finance wizard, but I peeked at their last SEC filing, and first 3 quarters of 2024 they posted a 35m operating loss, but added almost 900m worth of assets to their balance sheet (mostly Bitcoin), which to me tells a very different story

The quote is actually from the article this one paraphrased and linked to, while leaving out all of the actual, you know, information

New data tells us that mining a single Bitcoin or one BTC costs the largest public mining companies over $82,000 USD, which is nearly double the figure it did the previous quarter. Estimates for smaller organisations say you need to spend about $137,000 to get that single BTC in return. BTC is currently only valued at $94,703 USD, which seems to be a problem in the math department.

Bitcoin mining will always be profitable for the people with the cheapest electricity and largest economies of scale. There is a difficulty adjustment algorithm in the protocol that ensures this. When the price tanks people turn off thier miners, difficulty adjusts downwards, and then it takes less electricity to find a block.

tl;dr title is wrong

Title doesn’t match the text

one Bitcoin trades for around $94,000

$137,000 in electricity for small-scale operations to mine

optimal cost for mining a bitcoin at around $82,000

So really it should just say “mining bitcoin only profitable at scale in areas with cheap electricity” which has always been true to a degree

Gilmore girls isn’t mentioned in this article but it’s been measured before, it’s less than it’s always sunny

deleted by creator

It absolutely improves with practice, and once you have settled on an aesthetic you like you can simply reuse the code, e.g. store all your color/line properties in a variable and just update each figure with that variable

My thesis had something like 30 figures, and at multiple points I had to do things like “put these all on a log scale instead” or “whoops, data on row 143,827 looks like it was transcribed wrong, need to fix it”

While setting everything up in ggplot took a couple hours, making those changes to 30 figures in ggplot took seconds, whereas it would have taken a monumental amount of time to do manually in excel

Literally - you can pick out English longbowman bodies from the shape of their skeletons