> usually within 24 hours

I have never seen Apple approving within 24 hours

I've had Apple approve app submissions within 24 hours. It certainly hasn't been the norm (on average for me it takes more than 1 day, but less than 2), but I've had it happen occasionally over the past couple of years.

Not only that, if you have `trait Foo: Iterator<Item = u8>` and `trait Bar: Iterator<Item = i32>` how could `!` implement both of them? It would simply make the language incoherent.

Both are for for different niches, but their intersection is relatively big.

That page doesn't mention how joins between tables using different split keys work, or how indexes work.

Remember that correlation != causation.

They might also have unionized due to the bad conditions while developing GTA 6, which could also lead to it being a bad game. Who knows though.

I reached the midpoint of the podcast without seeing a proper argument against this then I gave up.

It tries to argue that higher reward rates are necessary to attrach customers that pay a lot (for credit card companies) and that lower income people are generally subsidized by taxes (obvious but unrelated), but at no point (until where I read) it seems to address the issue of merchants having to generally increase prices due to these cards.

> lower income people are generally subsidized by taxes

Pretty flimsy argument to begin with. Because the rich then argue to lower their taxes and/or simply not pay them and the whole system falls apart. Not to mention that lower income people paying more to keep afloat debt than taxes (which at worst is a much better interest deal for an installment plan) is a much worse model for society.

> credit card providers

The article did an awesome job explaining what are the parties involved and you choose to use a generic term instead.

> dealing with fraud and chargebacks

A lot of that is offloaded to the merchant, which instead has to pay them on top of what they already pay to the issuer bank.

> it’s a $14 fee per solar panel

Note that usually you mount many solar panels, not just one, so this will stack relatively quickly.

If you're building out, say 12kw of PV, that's about 24 panels for a modern install (500w bifacial n panels).

So a $336 premium, or roughly 1.3 additional panels in terms of cost.

While sure, that's not nothing, it's ~$6700 vs ~$7050, or a ~5% premium to the pre-tax cost of the panels.

Panels are much cheaper than that.

Here’s 31 445W panels for $4,650, full retail. This is literally the first google result I clicked.

Two and a half years ago I bought 18 405W panels for $3,312, full retail. They were cheaper wholesale.

https://cdnsolar.ca/products/canadian-solar-445w-all-black-p...

The vtable contains at least the following informations:

- the size of the pointed type - the alignment of the pointed type - a function pointer for its drop glue

That's already quite big to inline to save on indirection, and this is before adding more methods.

> for some reason

Just to expand on this: which vtables will be needed is not known before running codegen. At that point however multiple codegen units might generate the same vtable, and that might not get deduplicated by a later pass.

Moreover with dynamic linking (especially on Windows) it's impossible to guarantee that such vtables will be unique.