4 ms·
That's some honeymoon! The wife is gonna be mad at me for taking the moon bit literally. It's been more than a decade now for Go. What honeymoon is this? When
by eternalban 6y ago
That's some honeymoon! The wife is gonna be mad at me for taking the moon bit literally. It's been more than a decade now for Go. What honeymoon is this?
When I first saw Rust 1.0 spec, I made a mental note that "this will end in tears". (Do tears rust metal?) I made a similar prediction during Scala's honeymoon days: "This will be the JVM's C++".
Go will have a Java-lite career trajectory: it has a pretty wide sweet spot (in various dimensions) that more than compensate for its shortcomings.
The actual issue in this industry has been, and remains, the fact that it requires workers with above average intelligence that are tasked with building repetitive glorified book keeping applications. And in every programmer with a heart-beat's career, there comes a time when job boredom unconsciously lends a seductive glimmer to the shiny new thing. Java, Ruby, and Go all were in fact adopted because of the ambitious, smart, but bored, geeks decided they needed a brand new toy. (So taking your honeymoon metaphor, these trends are like mid-life crisis for programmers.)
And, compounding the issue is the fact that those minding the programmers were and remain (in the best cases) mostly behind the curve -- they're in meetings all day, after all -- and are simply not in a position to authoritatively smack down "let's use X!" enthusiasms. (In the general case, they are entirely clueless.)
Rust, like Scala, and Haskell, add an additional hook in that they offer the programming world's version of 'intellectual virtue signaling'. Case in point being your bon mot of "optimized for building".
Seriously, that's the actual job description for 99% of worker bees in this field: building things!
So, I'll make the prediction that Rust will go the way of Scala and Haskell, and there will come a brand new shiny and intellectually satisfying new thing and we'll have HN (in a typically forgetful fashion) dissing Rust and heaping praise on the new comer. This too shall pass.
[p.s. I have a self-contract to only post 1 message in a 24 hour period on the entire internet and this was it for tonight. /over & out]
- mohanmcgeek 6y ago> And, compounding the issue is the fact that those minding the programmers were and remain (in the best cases) mostly behind the curve -- they're in meetings all day, after all -- and are simply not in a position to authoritatively smack down "let's use X!" enthusiasms. (In the general case, they are entirely clueless.) I don't think them smacking the "let's use x" is a good idea since it's the Devs who'll be writing the code. The org needs a framework that answers: 1. How easy is it to train someone to use this stack? How long would it take for someone to start writing good code 2. How much time would it save us when building something new? 3. How easy is it for someone completely new to the codebase to start contributing? 1 and 2 can be easily measured within the org You can use OSS as a proxy for 3