304 karma · joined May 27, 2021
I think that's as 'boring' as it gets.
2. Career related stuff. It can be useful. But sometimes it too feels like echo chamber.
The person who asked 'is HN becoming an echo chamber?'
would've got their answer by now.
* Compiler support for function-level incremental may not be taken granted
* Editor shows a nice file tree (although you can do that with symbols too)
* Working with git is easier
* Reading code on site like GitHub is easier
What I think is there deserves to be one or two full pledged software engineering courses, teaching how to use testing, version control, software structure etc.. As always, you can start from basic commands / usage and reach to general concepts, it isn't that hard.
And actually I believe US and European academia, at least the Ivy league and second tier colleges, do a decent job at teaching many required concepts. Come to a country like India, and you will see a C programming course that doesn't cover malloc or function pointers, a data structures course without proper introduction to algorithm complexity or difference between abstract data type and implementation, a java course that never properly teaches subtyping. I think students in US graduate decently equipped to write real world software compared to us. At least, I learned a lot of things by reading notes and Ppt of US universities.
(most Indian colleges are woefully inadequate in teaching even the basic concepts for writing software, despite the CS courses being named CSE (comp sci and engineering), not just CS. Incompetence runs to the root in education system except few elite colleges. From that perspective I think CS depts saying that is a first world problem. Sorry for the rant.)
As far as I know
Rust has the same problem although to lesser extent. Monomorphization works well with judicious use. C++ STL is not written like that, they depend on 11111 layers of inlining to work well. Rust libraries aren't much better in this regard.
LTO removed some code bloat, but LTO itself takes more time. until thinLTO summary pass / equivalent pass in GCC WHOPR at least, middle end and early IR optimizations still have to happen, and Go wants to avoid that. I think that's a fine design choice. In Go's design, they have decided virtual calls aren't a cost they'd care anyway, pre 1.8 Go heavily used interfaces and that's not going to change.
> writing the code out customized for each given type should have identical performance to the template-generated version of the code
In theory yeah, but templates tend to generate more instantiations than strictly what you'd write by hand.
Also, obscure corner cases exist, but not big enough, thanks to those numerous man years spent on GCC and LLVM. https://travisdowns.github.io/blog/2020/01/20/zero.html
UFCS is better than C++ ranges. builtin sum types are better than std::variant, etc..
Monomorphization: * code bloat * slow compiles * debug builds may be slow (esp c++)
Dynamic dispatch & boxing (Usually both are needed): * not zero cost
Pick your poison
All in all I think the semantics debate is irrelevant. No one is going to use go for an OS only because someone on internet calls it a systems language.
People will still use Android or iOS even if you build a state of art mobile OS based on SeL4 completely written in Rust & ATS, with more features and customizability than Android or iOS. Because it doesn't come preinstalled on their devices, and does not have popular apps.
I am afraid walled gardens are logical conclusion of the field, for the things industry tends to optimize.
It's not about the quality of chromium at all, only ubiquity.
Web developers are cheap, its easy to port website's code and slap UI on it. That is the _main_ point.
The manager for whom electron is an option is least bothered by chrome being technically solid foundation. His motive is using cheap webshits and ticking a box.
Doesn't help other cross platform toolkits are either outdated, or immature, or in a weird position like Qt.
Our industry always optimized for some sort of local maxima. From a purely engineering perspective, scripting languages, or hacky UNIX/C glue, or weak typing, would not have made sense at all. But they sticked, because they fulfilled immediate commercial needs.
I don't know if this phenomenon is prevalent in other disciplines of engineering or there is a common name for it.
The place for generics in Go is data structure libraries, ORMs etc..
>> fast compiler
But yeah everything is fast at runtime as well when you coming from JS and python.
It takes 5-6 seconds to search for some files on my machine. Would'nt it be an easy win if `find` had spawned 4 threads?