foolswisdom

This seems to be like state machines, aside from

> As new ideas and requirements are discovered, we can forbid certain things from happening by simply adding new b-threads; without having to dig and figure out how earlier-written code works!

Which I'm not really sure is a step up from state machines?

Because there is nothing to make assumptions about. In the example code, both multiplication and division have a scalar on one side, there's no possible ambiguity of behavior. But there is the eternal question of terminology: do you specify dimensions by "axis" or "dim" and does your API actually use both terms in different places?

(that's what I think the GP meant, anyway).

B# == C.

I assume B# was intentional in that comment.

No "so what". I read the comment as purely informational.

It matters though because they didn't change their designation before acting differently, which would make them liable. Not sure to whom they'd be liable though, other than the IRS.

What if I almost never remember having dreamt once I woke up (and thus can't keep a journal), is it there any mechanism to make remembering dreams more likely?

What? 13*28=364 So, missing ~1.25 days every year (instead of ~0.25 days each year with a 365 day calendar). How does this solve leap years?

Is it not "a" business model, or not Boeing's business model? The latter is exactly what the FAA is saying :-).

Yes, because the merchants raise prices because people pay with credit cards (which take fees), you take back a little (though of course not all) of extra markup via points, assuming you use a 2% cashback card (at least) on your purchase (which assuredly many people don't).

What a wonderful license! https://github.com/Speykious/cve-rs/blob/main/LICENSE

The space where tabs go (when horizontal) isn't removed in vertical tabs mode. So using vertical tabs doesn't save any extra on vertical space, only uses additional horizontal space.

Currently, if a repo is on github, and you develop a patch on a fork on gitlab, you can't just click a button to open a pr, the way you could of both upstream and fork were using the same provider. Federated PRs would be the capability for one provider to "talk" to another to be able to pull your patch from the other provider.

I know edge at least supports vertical tabs? I don't like them though (I don't have enough horizontal screen space, plus it doesn't free up the top bar space). If I start to have many open tabs I start cutting them down anyway.

I wonder the same, I can definitely see the karma of users who have only double digit (and I think I have seen single digit) karma.

Not really, he's saying that the solution is to not have the entire process in a single model, it's better to have the model work on specific pieces that you broke down, rather than feeding the whole thing and expecting the model to be able to break it down and generate correctly by itself.

I really like this idea.

My initial impression of this is that it enables controlling just how much of the app you want to render for your test, while more traditional solutions force you towards either rendering the whole app or testing just a component. Is that accurate?

That's a P/E ratio based on the past four reported quarters, over which time earnings at Nvidia increased so much that the earnings for the last reported quarter was over 4 times that of four quarters ago. Assuming that future earnings for the next year are consistent with the last reported quarter, they're valued at 45x earnings. But of course such a valuation still only reddit makes sense assuming growth (and the projected earnings for the next quarterly report is indeed expected to be 12% higher, giving a 40x p/e valuation). Consider that Amazon's P/E (using the last reported quarter as the baseline for the future - Amazon has had consistently increasing earnings over the last year, let's assume it doesn't decrease from here) is 43x.

Is this true for turbotax specifically, or merely for intuit as a whole (which includes things like quickbooks)?

How is this able to pause work, given that the work is already executing on the main thread (and I assume you don't need to rewrite your code to explicity give up control at various points)?

Sounds like you're using the concept of statically served files vs dynamically generated, but that's a different conversation to whether content files are generated at all.

This article focuses on how money allocated (by universities) to non-study functions such as hospitals results in university management being much more docile and less principled. There's another interesting antitrust take on this, (the link mentions antitrust only in respect to health insurance, but the monopoly angle here is much broader than that): https://www.thebignewsletter.com/p/how-to-crush-the-ivy-leag...

Thanks for the links!

Well if you're not using unsafe, you would expect zero segfaults when using rust, no? Though I imagine it's impossible for anything with graphics to avoid unsafe code entirely.

Looks like humor to me.

This seems to be similar to https://github.com/cashapp/sqldelight, and I've always wanted a python equivalent!

In typescript, there are query builders (not talking about ORMs) that can basically do this within the type system, but that would be infeasible in python's type system. This approach (type/code generation is a good alternative, though I like using sqlalchemy / alembic to manage schemas/migrations.

One thing I'm curious about is how it knows the types of columns? I looked quickly at the Readme but didn't see it (probably a parameter somewhere I missed).

I don't know anything about that, but he was born in 1938, so he wasn't 19 years old.

No, it's impossible to do random access to retrieve a character, if you are dealing with code points, because code points do not have a fixed byte size. I thought this a good intro <https://tonsky.me/blog/unicode/>.

Code points are not bytes.

That's not exactly what happened though https://x.com/billackman/status/1725898682438979972?s=46&t=K...