It’s one topological sort, implemented here. What could it cost? Ten lines?

This one idk, some of it could be more concise but it also has to build the graph first using that weird seemingly custom hashmap as the source. This function, however, is immensely funny

There’s a standard algorithm for new backends, NOLTIS

I think this makes it sound more cutting-edge and thus less scathing than it should, it’s an algorithm from 2008 and is used by LLVM. Claude not only trained on the paper but on all of LLVM as well.

We’re Still Early never dies

I wonder what actual experts in compilers think of this.

Anthropic doesn’t pay me and I’m not going to look over their pile of garbage for free, but just looking at the structure and READMEs it looks like a reasonable submission for an advanced student in a compiler’s course: lowering to IR, SSA representation, dominators, phi elimination, some passes like strength reduction. The register allocator is very bad though, I’d expect at least something based on colouring.

The READMEs are also really annoying to read. They are overlong and they don’t really explain what is going on in the module. There’s no high-level overview of the architecture of the compiler. A lot of it is just redundant. Like, what is this:

Ye dude, of course it doesn’t depend on the IR, because this is before IR is constructed. Are you just pretending to know how a compiler works? Wait, right, you are, you’re a bot. The last sentence is also hilarious, my brother in christ, what, why is this in the README.

Now this evaluation only makes sense if the compiler actually works - which it doesn’t. Looking at the filed issues there are glaring disqualifying problems (#177, #172, #171, #167, etc. etc. etc.). Like, those are not “oops, forgot something”, those are “the code responsible for this is broken”. Some of them look truly baffling, like how do you manage to get so many issues of the type “silently does something unexpected on error” when the code is IN RUST, which is explicitly designed to make those errors as hard as possible? Like I’m sorry, but the ones below? These are just “you did not even attempt to fulfill the assignment”.

It’s also not tested, it has no integration tests (even though the README says it does), which is plain unacceptable. And the unit tests that are there fail so lol, lmao.

It’s worse than existing industry compilers and it doesn’t offer anything interesting in terms of the implementation. If you’re introducing your own IR and passes you have to have a good enough reason to not just target LLVM. Cranelift is… not great, but they at least have interesting design choices and offer quick unoptimized compilation. This? The only reason you’d write this is you were indeed a student learning compilers, in which case it’d be a very good experience. You’d probably learn why testing is important for the rest of your life at least.

This could be regarded as a neat fun hack, if it wasn’t built by appropriating the entire world of open source software

This shouldn’t be left merely implied, the autoplag trained on GCC, clang, and every single poor undergrad who had to slap together a working C compiler for their compilers course and uploaded it to github, and “learnt” fuckall

Waiting for some promptfondler to complain that this kind of assignment is not really fair for an AI because it has actual requirements.

Here’s a fun exercise: go email the author of that blog (he’s very nice) and ask how much of it he still stands by.

Has someone already done it or should I send the email? Ludic would be delighted I’m sure

I guess my confusion is this – at what point does reality exert any influence on the lies-and-trash accounting? What in the physical reality is the reason they can’t just keep fapcircling the economy forever?

So we don’t have smoking gun evidence that Windows 11 is broken trash literally because of vibe coding. But Windows 11 feels like the most vibe coded thing ever. Nobody cared about Windows 11 working. Microsoft, where quality is job number 55 or so!

I think the timelines don’t quite match for that, when Windows 11 came out the LLMs couldn’t even output code that compiled more than half the time. However, the moment I’ve learnt the start menu was coded in React I knew the age of man Windows was over. I don’t know how good the guy behind it was in internal politics but everyone on the path from the idea to pushing it to main should be tried for treason and sent to hard labour (debugging the Azure web UI).

I think this was the good old artisinal vibe coding of the previous era, where you still had to not care about quality manually.

wasn’t audited (…) prompting the SEC to open their own investigation.

Ye, I guess I should’ve said “investigated” there, precision is important here.

There’s not only no stock-holding public to hold them to account, even if there was and they filed a lawsuit nothing would happen. SEC doesn’t exist.

I guess this is the part I don’t understand - can’t they all just play this game of pretend forever? There has to be something that corresponds to music stopping in your musical chairs analogy, but if everyone plays along forever what happens? Is there any load bearing part of this bubble where real reality can have an impact? If the stakes of calling the bluff are a complete collapse of the stock market then why would any NVidia stockholders call the bluff ever?

So you talk about Depression 2, but do you have any idea of what could possibly be the spark? You compare it to Enron, but Enron got had because they got audited. The AI bubble seems unparalleled in creating fake money from nothing, is there any limit to how long they do it? Tether still exists

Also “tokens per watt per dollar” by god where are we. It’s like when Elon started to talk about “userminutes per tweetparsec per death threat” or whatever when he took over Twitter

Huh, I hope Satya asks Copilot how to solve this or they’re in trouble!

Well, Satoshi’s whitepaper is called “Bitcoin: A Peer-to-Peer Electronic Cash System”, so it’s pretty obvious what the author(s) consider(s) its main contribution to be. Arguably, the concept of a blockchain wasn’t invented by Satoshi (the idea of a ledger using Merkle trees of hashes existed before), neither was proof-of-work (the idea of requiring computation as a cost of performing an action also existed, e.g. to prevent email spam). The novelty was combining all that to create a currency system and the implementation. And the motivation spilling throughout the paper is very much libertarian and goldbuggy.

I’m sorry but like why do you think Satoshi created Bitcoin, like what was the purpose of that paper?

Wait, does the Starlink internet suck as well? I mean, it has to have high latency, that much I’ve assumed, but other than that?

The market can remain irrational longer than a chud’s heart can remain functional

SpaceX must have discovered new physics that solves orbital heat management, because otherwise Musk and the stockholders are dumb.

Truly a conundrum worthy of the XXI century

and it will be trivially affordable to run them

ON WHAT HARDWARE BEN, DDR2 SCAVENGED FROM JUNKYARDS???

Harder than you think if you’re a cutting edge chud