Mike Treit

@CoolAssPuppy @michaelfreedman Wow, ok that was perfect

@michaelfreedman Load bearing is a statement about structure and it’s doing real work here. Your framing points at the seam between vocabulary and intent. It’s not imitation. It’s incentive.

How did models get trained to call everything "load-bearing" instead of simply "important" or "relevant"? What massive text corpus is saturated with that phrase?

Do you DIG the PDP-10? We’re hacking 36bits today at icm.museum #retrocomputing #vintagecomputing

@DaQuackers @mika_tuomaala @DJSnM Are you a junior software developer? Because I do not see how it's even remotely possible not to get fantastic results from AI tools if you are a senior or above.

AI coding tools don't help me achieve shit. In fact I keep inheriting messes to clean up when other people try to use them and the AI fucks up. Each time it's come down to "well this took 30% longer to fix than it would have to write correctly from scratch... wish you'd just given me the ticket to begin with" Biggest offender keeps being Claude. Head of engineering wants to justify the seats he paid for and it just keeps making a mess and not accomplishing shit every time.

Someone 'accused' me of vibe coding the LFRR simulator. I've been coding since 1982, I've hand assembled machine code, injected fixes into running processes, worked on software that's made it to Mars and worked as a software developer for decades. OF COURSE I USED AI CODE TOOLS

I don't know who this mathematician you're quoting is. But I used to be a mathematician myself, and my reaction is: get over yourself! I see you all right. I see your pathetic whining and sense of entitlement. And I have no sympathy at all. The human pursuit of mathematics doesn't exist so snowflakes can feel special, it exists because we have good reasons for seeking the truth. I want from mathematics to software engineering. I've spent 50 years learning how to get good at it. You will never, ever hear me whinging because the frontier models can now write better code than I can. If I prompt them properly. And that if "I prompt them properly" is important. It will tell you what the way forward is, if you're not too busy drowning yourself in self-pity.

@marcgravell @neuecc I have the same question.

@neuecc Genuine question from one library dev to another: ns2.1: what's the point? ns2.0, sure - awkward, but I get it. But ns2.1: what's the genuine use-case where net10.0 etc isn't objectively preferable?

週刊進捗出してきます、ということでnetstandard2.0, netstandard2.1, net10.0をターゲットフレームワークにしましたの仕組みを作りこんできました。また、量産一歩手前なので、unsafeやPinを除去したり境界チェック入れたりで安全性への考慮点を増やしました。 github.com/MessagePack-CS…

@neuecc What possible reason would you have for targeting netstandard 2.1 if you are already targeting netstandard 2.0?

NT SUR was my third project at Microsoft (after MS-DOS and Windows NT Daytona). We took ALL the UI code from Win95, then spent months hand-tweaking and merging. Remember, there was no Git, so this was all 'cat' and 'diff' and 'grep' and so on. Then manual review of every line. Sheer grunt work. Characters were no longer 8 bits. The CPU was not an x86. Alignment mattered. Unicode mattered. Networking was TCP/IP. None of the code supported any of this. Over the course of several months, we had the core shell DLLs ported to NT. BobDay and I worked on shell32 while others worked on comctl32, comdlg32, and so on. Eventually, we got it to compile on one momentous day. We took bets on how far the process would run. Bob said it would crash in Init. I was way more optimistic; I thought we'd make it to the desktop. We placed our bets, typed 'explorer' and hit Enter. The shell chewed away, and a few seconds later, the seafoam green desktop and some blank icons came up. Then it crashed hard. But I had won the bet, and that dollar stayed pinned to my corkboard for my entire career. We built scripts and tools to overlay the new binaries on top of Daytona, and that became the Shell Update Release, or SUR. There was no setup, just a folder of binaries and a CMD file to copy them into place. Our T-Shirts read: SUR - We don't Need No Stinkin' Codename!

@DIYFractal @zherbert 😂

@zherbert this is, after all, their first time using the TRNG

I don’t even know what to say

never using Apple Maps again

Irresponsible. Against all development and security best practice. Impossible to overstate how egregiously bad this was. The developer, the processes, and, ultimately, the leadership are responsible.

@brian_trollz This is like the third post I have seen today where people don't know how to spell "throw" properly...🤔

DO NOT under ANY circumstances sell, destroy, through away, or give a compromised Coldcard to anyone else if you have lost funds. The odds are INCREDIBLY LOW, and you should NOT get your hopes up, but if any stolen funds are recovered this is going to be the only basis to potentially prove your legitimate ownership. (It has to be the device that GENERATED the seed, not just the one it is currently stored on).

@northernH0DL Consider not saying things that are completely mathematically incorrect. 2^256 is a very large number. It's smaller than the number of atoms in the universe by a factor of 1,000. Your statement is uninformed, made up gibberish.

Even if every atom in the observable universe were a supercomputer checking billions of seeds every second since the Big Bang, they wouldn’t make a meaningful dent in that search space.

A 24-word BIP-39 seed has 256 bits of entropy. That means there are 2²⁵⁶ possible seeds. That number is approximately 115,792,089,237,316,195,423,570,985,008,687,907,853,269,984,665,640,564,039,457,584,007,913,129,639,936 That’s unimaginably large. And the best part?

I'm a software engineer with an MSc in Computer Science and 25+ years in the industry. Here's why Coinkite's Coldcard bug isn't an oopsy, it's disqualifying. Coldcard's seed generation silently fell back to a non-cryptographic "random" number generator for 5 years. Not because crypto is hard, but because Coinkite violated the most basic rule of secure system design: fail closed, never open. When a security-critical path can't be verified, the system refuses to run and throws an error. It does not quietly substitute something weaker and carry on. This code should have refused to produce a seed. Instead it produced a predictable one and continue silently. ☠️ A company that lets its most critical code path go unverified for half a decade cannot be trusted with your keys. Throw away your Coldcards. Never buy one again.

Happy Birthday George Jetson he is 4 today… flying cars must be just around the corner. I can’t wait! 🎂

Here's the sound of a #QRP #PIXIECWTreansceiver for those who've never heard it. These inexpensive little kits are a great for an evening project. Great starter for kit building and code practice/sending. Not the greatest reception nor power but making a QSO is immeasurable.

@N4JAW I'm giving this a like because...I love CW. But...good lord that sidetone pitch. Is it not adjustable? 600 Hz sounds great, whatever that is is making me grimace. Although if you like it, good for you 🙂 Thanks for the content!