Having used compilers for a few new languages (Rust, Go, Dart, Kotlin, Swift). Swift is the only one I've had any issues with as well as Swift seems to be the only language to have adopted the "move fast and break things" philosophy of Silicon Valley. I dunno, I just don't see the argument.
LLVM is one of the most influential pieces of software of the past decade. Hero worship isn't good but credit where credit is due.
Additionally I'm sure we can all agree there is no substitute for maturation through time and usage in the field. Which frankly is an argument for more popular languages over obscure ones. None of the ones you mentioned are ready for safety-critical system development (including Swift 3), but which one is most likely to achieve widespread adoption and field testing in the long run?
I don't think Swift stands to gain wide spread traction outside of Apple orientated app development. Aside from a lack of stability, Apple is to well known for boxing its competitors out. I've used and loved their products my entire life and I know how annoying it is to go against Apple's grain.
It already is though, there are several Linux web frameworks etc. It's open source and community run so I'm not sure how they're planning to box out competitors from it.
When writing a server, I would take Go over Swift anyday. It out preforms it, uses less memory, its simpiler, oh and it uses a "tradiontal" GC.
That is very much _not_ the case according to the testing I have done recently.
Swift uses a lot less memory than Go unless the program uses only trivial amounts of memory in the first place. Using interfaces in Go data structures makes the difference even more pronounced.
On top of that, all runtimes that use a tracing GC require tons of spare memory at all times unless programs are very carefully written to avoid memory allocation.
That said, Swift has a few very weak spots when it comes to memory. Most notably the String type, which is terrible on all counts, but that is a whole different story.
My own code is unfortunately a bit messy and entangled with unrelated stuff. If I find the time I'm going to clean it up.
Only if the said language doesn't allow for stack or static globals.
Quite a few languages do allow it.
Not really, example in Active Oberon:
TYPE
point = RECORD x, y : INTEGER; END;
VAR
staticPoint : point; (* On the stack or global *)
gcPoint : POINTER TO point; (* GC pointer *)
noGCPoint : POINTER(UNTRACED) TO point; (* pointer not traced by the GC *)
Mesa/Cedar, Oberon, Oberon-2, Active Oberon, Component Pascal, Modula-2+, Modula-3, D, Oberon-07, Eiffel, BETA.
There are probably a few other ones.
That's fine for one point. How about N points where N varies at runtime?
If I allocate memory dynamically outside the GC's remit, I'm going to have to release that memory somehow.
On Active Oberon's case, those pointers are still safe. They can only point to valid memory regions, think of them as weak pointers that can also point to data on the stack or global memory.
This in safe code.
If the package imports SYSTEM, it becomes an unsafe package, and then just like e.g. Rust's unsafe, the rules are bended a bit and usage with SYSTEM.NEW() SYSTEM.DISPOSE() is allowed.
Just like any safe systems programming language, it is up to the programmer to ensure this pointer doesn't escape the unsafe package.
In a safe package it can only point to existing data, there isn't anything to release.
If the pointee is something that lives on the heap, it is similar to weak references. Points to GC data, but doesn't count as yet another GC root.
If the pointee is on the stack or global memory (data segment in C), then there is also nothing to release. Global memory only goes away when program dies, stack gets released on return. Memory that was allocated by the compiler due to VAR declarations, it is static.
Usually the idea is that you use untraced pointer to navigate statically allocated data structures, they are not to be exposed across modules.