They're pretty readable, I recommend that second one (though it leaves some questions I have unanswered). It's a good example of how latent issues with a large system can go unrecognised for a long time before a seeming unrelated change can trigger a cascading failure

One thing it doesn't make clear – the GPS card only supported the original L1 signal, which is what causes the GPS week rollover issue. The newer L2C signal has a much longer rollover period (157 years vs 19.6 years), which means no rollover until next century. If the GPS card had supported the newer L2C signal, then likely this would not have happened even if the other misconfigurations had still occurred.

Yeah I've seen plenty of kernels that start with a 2 with no forward path possible due to "we don't know how to get our driver working anymore"

Maybe the closest is 'join' similar to `",".join([...])`? If we could replace the string with an operator

I like fold too, but I can never remember foldl vs foldr, it's always backwards to what I expect somehow

https://news.ycombinator.com/item?id=49736033 may help :)

Thanks! That visual style approach works better for me, hopefully it sticks

Oh it's a bit like unordered-delete when using an arena. I guess I would have expected an ordered-delete instead

> How hard can it be?

> As it turns out: enormously hard.

This exactly tracks me learning everything

Jane Street released a new challenge so I'm considering picking it up. It hasn't fully captured my brain, but I can feel it in there from time to time and depending how it lands it might be my next project [0]. I'm not really a hardware person but I'm close enough that it's maybe within reach

[0] https://blog.janestreet.com/protocol-emulator-asic-competiti...

Uh oh. The last one was pretty tough for me and this one is much bigger. If anyone is interested to form a team let me know (probably east coast AU timezone ideally).

This is so cool, and so beautiful! You've done a fantastic job on this. This is the sort of site I wish I were capable of producing, it reminds me of the ciechanow site [0] in that it's both really educational and beautiful at the same time. It's answered questions I had like 'why is power always at the top and ground always at the bottom?'

And I love the way that you solved it by understanding really well what each section does. I worked out bits and pieces of each one, but I realise now that I was missing a lot of clues that were right in front of me (for example that it represented an 11x11 grid, the operation of what I called 'the blob' on the lower right etc). And so far as I can tell you're the only one to find `TWO''NOT TOUCH`

Edit to ask: How did you find the `TWO''NOT TOUCH` output? I assume you worked it out via some structure? I gave up after 4 messages not realising it was possible to get more.

[0] https://ciechanow.ski/ - I particularly liked the one about the mechanical watch mechanisms, but honestly pick any

Hi, thank you!

for the other output: 1) look at the inputs to the output block and simulate those, you'll get a condition you need to satisfy 2) check what those conditions mean by interpreting the other blocks 3) try to satisfy the conditions

You know how everyone thinks agents are bad at the thing they're good at and good at the thing they're bad at? It turns out I must be bad at testing, because I thought they do a reasonable job.

I thought so too but it's probably because I never review the tests. I made an exception recently and found tons of

    assert(CONSTANT_CONFIG == valueOfConfig)

or tests for keywords in prompts:

    assert(prompt.includes("repo url"))

Was your prompt, "make lazy ineffectual tests?" Tests are code as well, they take the same effort to create, if not more. You need to specify the properties you want to test.

Most developers are not good at testing. That's why the role of test engineer exists.

I've seen in multiple projects things like assertTrue(true).

I'm sure the agent is better in testing than average enterprise developer.

big tech is moving / has moved away from the role.

Unfortunately for the users, testing has been pushed to them. You can deploy whatever slop you want - nobody cares. More lines of code!

I don't think these results show that.

I think they show that additional prompting for testing approaches have wide differences in error rate (worst has twice the error rate of the best) but actually no extra prompting is pretty fine and most custom prompts are worse than no prompt.

Isn't that just a form of Dunning-Kruger effect related to the parts that person is bad in? It looks good because they don't have the depth of knowledge/expertise to see the deficiencies.