15,472 karma · joined January 4, 2011
> I also always believed that if you ever started a project that is valuable for companies, they would support you in return
But this just seems incredibly unrealistic. It's not remotely the norm for small open source libraries.
Omar Cornut, developer of Dear ImGui which is used by pretty much every AAA game studio, said he recently lost his biggest source of income. If he's having trouble getting stable sponsorship, I mean, good luck.
Here's an easy one: nearly any pun is literally impossible to translate while still being a pun. You can explain the pun by talking about the original language, but it can't be adequately translated.
If they were just releasing more good video (people loved that Alicia Keys thing), I think there'd be more excitement among Apple fans at least. As it is, it's hard to escape the conclusion that Apple really does think of this as more of a dev kit than a device even for enthusiast users.
I'm sure there are good use cases, but I just haven't seen anything life-changing in my experiments with various LLM products. The only thing I use day to day is IntelliCode, which saves a bit of typing here and there.
Gurman lists various reasons, but I feel like the biggest one by far is that the pace of smartphone development was incredibly exciting from 2007 to like 2017 (the iPhone X). Maybe a little later, but at some point you've got a fantastic camera and a very fast device with a great screen, so why would a normal person upgrade?
I think the cheap iPhone is one valid path, but you could also take the completely opposite direction, go all-in with on-device AI, with tons of RAM, a massively upgraded Neural Engine, etc. Find another big leap forward to offer.
If something's slowing it down, it's probably an autcomplete feature which can be configured to trigger less frequently.
Server: fritz.box Address: fd00::[etc]
Non-authoritative answer: Name: google.com Addresses: 2a00:1450:4001:80b::200e 172.217.19.78
I think if we had perfect data, there'd be a strong correlation between sunlight hours and lower depression rates. It's one of the few real factors for alleviating mild-moderate depression. Stuff like living standards has very little relation to actual clinical depression.
AltStore or Epic complying with Apple's current dubious terms makes no real difference.
This definitely isn't intended for general-purpose sandboxing. It's an interesting tool for analysis and debugging.
That's your candidate for "worst ever"?
My theory is that bookmarks are a failed concept, putting something out of sight but not in a useful system, so it's guaranteed to be forgotten. If I really need to save an important link, it goes in a notebook which is far more flexible.
I'm sure Microsoft still wants to collect Windows license fees from enterprises and hardware manufacturers, but they're far more relaxed about Windows licensing than they used to be.
I would default to just using std::unique_ptr<Node> in a situation like this, especially since using std::list suggests performance isn't critical here, so the additional indirection probably doesn't matter.
It's a fine scripting language for hobby games or glue code. I would never do serious software development with it.
What I have done is listen to a foreign language audiobook while having the text in English, which I find extraordinarily helpful at an intermediate level where you understand most of the words but may not catch the full meaning.
I think even that very minimal structure helps mentally elevate it from a boring math problem to a "real" scenario.
Funny that they were aware of the mechanic that can be used to skip grapple beam entirely (a huge time save), but didn't incorporate it into the "speed run" walkthrough.
The earliest known recorded run is from 2003, by which time the mockball had been discovered, but not the zebetite skip, so there's far more ammo collection than you'd see even in a beginner's run today.
https://wiki.supermetroid.run/Record_Progression#Ancient_His...