I'm an enthusiastic Julia user but I got to say having a documentation committee would go a very long way for us. Quality and exhaustive doc is almost as important as quality packages for language adoption. Right now, it still feels we're in the "let a thousand flower bloom" phase without many style criteria and asking new users help improve the doc while learning.
The discussion on using what tools align with your values at 23:57 - 25:12 perfectly illustrated why I love Kotlin and Rust, and why I hate working with Javascript so much, where others love it. JS is just...antithetical to everything I hold dear.
I'm "Javascript developer" and I hate it. The mere fact the TS is so popular just show how pathetic the language is. Dynamic typing vs Static typing...why is that even a question? It's like asking: Go to war with a Butter vs Rifle. What computers do great is keep track of things in the way we just can't. I want that a program runs through my code and tell me that it will be fine. It's a sort of testing, practically the very first layer, before unit testing.
In my career over the last 10+ years, JavaScript had gone from one of the worst languages I've had to work with to a language that's a delight to work with. I find it fascinating that the "flavor" of it changed so much that my view changed like that.
@@brokula1312 lol. The truth is TS is also a dynamically typed language... No static language can superset a dynamic language. The new trend in dynamically typed arc is to have the gradual and optional typing. Conversely, in static typing, there are stronger and stronger type inference and typeclass-ish features. As long as the code has a loose coupling, dynamically-typed languages are actually quite comfortable to use. And statically typed languages are weak at metaprogramming... Both with pros and cons
Around the 10ish minute mark Stephen talk about win-win vs. zero-sum interactions. One of the key teachings of economics is that the world is full of win-win opportunities, and we collectively do ourselves a favor by maximizing our utilization of those opportunities. Mostly we carry out that maximization by making interactions between people based on mutual consent: if you want to interact with me in a way I think I won't like I can say no, take my ball and go home; you can do the same to me. If we both say yes, that must mean we both believe we'll benefit from the interaction-that is, we believe it's win-win. There are corner case exceptions to this rule, but they are just that: exceptional corner cases. The fundamental idea in economics is that win-win is possible and good, and there's a system for achieving that which is fairly simple (although the fully detailed argument is a lot less simple than my sketch of the main point).
Sorry to be so offtopic but does someone know a tool to get back into an instagram account?? I stupidly forgot my login password. I would appreciate any help you can give me
Rust has one of the best documentations I've seen since I learned Java (to me its better than that and more organized as well) thanks if you had a part in that.
Throughput vs latency can be a fundamental tradeoff in the specific context of choosing packet sizes. With a fixed link speed larger packets allow more throughput because the ratio of payload to header is increased, but it aldo takes longer to transmit a complete packet. This came up in the design of the ATM protocol which was widely used for telephone and data networks in the 1980s and 1990s. The US wanted 64-byte payloads, but France wanted 32 bytes because this would have kept the latency low enough for them to build an ATM voice network across the whole of France without using echo-cancellation. They compromised and made everyone unhappy with 48-byte frames. with 5 bytes of header, making 53 bytes total. Seriously.
This is true enough, but again, as Steve was saying, it's not a _fundamental_ trade-off - it doesn't exist in all systems and scales and contexts, but rather you have different amounts that one must affect the others in different situations
cargo relies on the internet to function. Supporting old hardware is going to be next to impossible anyways. ... which is especially painful for FreeDOS development ... which has 2's complement notation
@Alex A company that can't handle it's employees having opinions on system designs and the way their jobs are supposed to be done, is not an employer I'd want for sure.
I like his presentations but 10:00 - none of them are inherently against each other. I see no reason to single out throughput/latency. How choices are against each other depends on how bounds are defined - those may be rather fuzzy sometimes. If you inline everyting - binary gets larger - you get more cache misses, so "everything gets faster" is not also always true.
This guy: "guys value different things then me, and prefer different languages, and that's cool" Other rust core devs: "The government should forcefully regulate the use for rust in applications!"
I've seen like 3 vids of this guy in the last day and I've seen his hair go through its own anime training arc
😂
I'm an enthusiastic Julia user but I got to say having a documentation committee would go a very long way for us. Quality and exhaustive doc is almost as important as quality packages for language adoption. Right now, it still feels we're in the "let a thousand flower bloom" phase without many style criteria and asking new users help improve the doc while learning.
The discussion on using what tools align with your values at 23:57 - 25:12 perfectly illustrated why I love Kotlin and Rust, and why I hate working with Javascript so much, where others love it. JS is just...antithetical to everything I hold dear.
I'm "Javascript developer" and I hate it. The mere fact the TS is so popular just show how pathetic the language is.
Dynamic typing vs Static typing...why is that even a question? It's like asking: Go to war with a Butter vs Rifle.
What computers do great is keep track of things in the way we just can't. I want that a program runs through my code and tell me that it will be fine. It's a sort of testing, practically the very first layer, before unit testing.
In my career over the last 10+ years, JavaScript had gone from one of the worst languages I've had to work with to a language that's a delight to work with. I find it fascinating that the "flavor" of it changed so much that my view changed like that.
@@brokula1312 lol. The truth is TS is also a dynamically typed language... No static language can superset a dynamic language. The new trend in dynamically typed arc is to have the gradual and optional typing. Conversely, in static typing, there are stronger and stronger type inference and typeclass-ish features.
As long as the code has a loose coupling, dynamically-typed languages are actually quite comfortable to use. And statically typed languages are weak at metaprogramming... Both with pros and cons
Around the 10ish minute mark Stephen talk about win-win vs. zero-sum interactions. One of the key teachings of economics is that the world is full of win-win opportunities, and we collectively do ourselves a favor by maximizing our utilization of those opportunities.
Mostly we carry out that maximization by making interactions between people based on mutual consent: if you want to interact with me in a way I think I won't like I can say no, take my ball and go home; you can do the same to me. If we both say yes, that must mean we both believe we'll benefit from the interaction-that is, we believe it's win-win.
There are corner case exceptions to this rule, but they are just that: exceptional corner cases. The fundamental idea in economics is that win-win is possible and good, and there's a system for achieving that which is fairly simple (although the fully detailed argument is a lot less simple than my sketch of the main point).
I came from Haskell, and I am loving the Rust type system... it is just what was missing on C
Sorry to be so offtopic but does someone know a tool to get back into an instagram account??
I stupidly forgot my login password. I would appreciate any help you can give me
@@sageseth5483 you can do the forgot login thing and restore it with your email
Rust has one of the best documentations I've seen since I learned Java (to me its better than that and more organized as well) thanks if you had a part in that.
I'm going to name my child Rust
Throughput vs latency can be a fundamental tradeoff in the specific context of choosing packet sizes. With a fixed link speed larger packets allow more throughput because the ratio of payload to header is increased, but it aldo takes longer to transmit a complete packet. This came up in the design of the ATM protocol which was widely used for telephone and data networks in the 1980s and 1990s. The US wanted 64-byte payloads, but France wanted 32 bytes because this would have kept the latency low enough for them to build an ATM voice network across the whole of France without using echo-cancellation. They compromised and made everyone unhappy with 48-byte frames. with 5 bytes of header, making 53 bytes total.
Seriously.
This is true enough, but again, as Steve was saying, it's not a _fundamental_ trade-off - it doesn't exist in all systems and scales and contexts, but rather you have different amounts that one must affect the others in different situations
Nowadays, with increasing BDPs, we've got 9001-byte packets flying around!
I love how elucidating this talk is, I really like his style of communication mixed in with humor
cargo relies on the internet to function. Supporting old hardware is going to be next to impossible anyways.
... which is especially painful for FreeDOS development ... which has 2's complement notation
Absolutely amazing! Thank you!
Bending the curve lol
Well done! Great presentation.
Also I'm pretty sure why you are "between" jobs...
Maybe self employment is the best option for you!
Why?
@Alex A company that can't handle it's employees having opinions on system designs and the way their jobs are supposed to be done, is not an employer I'd want for sure.
Awesome talk, thank you!
Engaging speaker.
I like his presentations but
10:00 - none of them are inherently against each other. I see no reason to single out throughput/latency. How choices are against each other depends on how bounds are defined - those may be rather fuzzy sometimes.
If you inline everyting - binary gets larger - you get more cache misses, so "everything gets faster" is not also always true.
Talk starts at 15:26
This guy: "guys value different things then me, and prefer different languages, and that's cool"
Other rust core devs: "The government should forcefully regulate the use for rust in applications!"
Performant is a word in French.
Wow. Nice talk.
dark mozilla give us the erlang-like pre 1.0 rust
🤟🤠
the rust video starts after 16:00 only, SAVE your precious time