+ We recommend to read the Programming Flix book. +
-
- Yes, with the import mechanism.
- You can read more in the documentation, in the section: Interoperability.
+ If you get stuck or need help feel free to reach out to
+ us on Gitter.
- Flix borrows a lot of syntax from Scala, hence the two languages have a similar feel. -
-- We think Scala made many good design choices with respect to syntax, including: -
- -x : T syntax for type annotations,List[Int32] syntax for type parameters, andif, match, etc. as expressions.- However, other than syntax, the two languages are very different: -
- -+ No, Flix is a full-blown general-purpose programming language. +
- We recommand that you have some prior experience with functional programming. + Flix borrows a lot of syntax from Scala, hence the two languages have a similar feel.
-- Other than that, the best place to start is to look at the examples and to - read the Programming Flix book. + We think Scala made many good design choices with respect to syntax, including:
-- If you get stuck or need help feel free to reach out to - us on Gitter. -
-x : T syntax for type annotations,List[Int32] syntax for type parameters, andif, match, etc. as expressions.- No, Flix is a full-blown functional programming language. + However, other than syntax, the two languages are very different:
-- That said, Flix could be considered a meta-programming language for Datalog. -
+- Flix runs on the Java Virtual Machine (JVM) hence the performance of Flix is limited by the - performance of the JVM. Luckily, the JVM is a mature and performant virtual machine. Flix - uses monomorphization which eliminates boxing and in theory can make code execute faster - than ordinary Java / Kotlin / Scala code. However, Flix also features full tail call - elimination - which has some run-time performance cost. + Flix runs on the JVM which means that the performance of Flix programs is comparable to that + of Java and Scala programs. +
+ ++ Flix is a whole-program optimizing compiler that uses monomorphization and inlining (like + Rust and MLton). Hence, sometimes, Flix programs run faster than their Java or Scala + equivalent.
- To answer this question, it is important to distinguish between latency and throughput. -
- -- The Flix compiler runs on the JVM hence startup times can be expensive, i.e. the compiler - has high latency. However, once the JVM has warmed up the compiler is quite fast, i.e. has - high throughput. -
- -
- A simple experiment shows that the compiler runs about 4.0x times faster
- when warmed up compared to when cold. We estimate that Flix, when warmed up, compiles
- around 50,000 lines of code per second, which we believe to be faster than the
- Scala compiler, but slower than the Java compiler. We take compiler performance
- seriously and continuously track the performance of
- the compiler.
-
- Compiler throughput on an AMD Ryzen 9 5900x with 12 cores and 32GB memory. -
- -- Experimental results from the 27th of November 2023. -
- -| What | -Threads | -Throughput (lines/sec) | -Ratio | -
|---|---|---|---|
| Entire Compiler | -1 | -12,062 | -1.0x | -
| Entire Compiler | -24 | -56,632 | -4.7x | -
| Compiler Frontend | -1 | -19,191 | -1.0x | -
| Compiler Frontend | -24 | -112,898 | -5,9x | -
- Flix will, by design, not support any of the following features: -
- -| Feature | -Reason | -
|---|---|
| Null Values | -
- Null, famously called Tony
- Hoare's billion dollar mistake, is simply a bad solution to the problem of how
- to represent the (potential) absence of a value. Instead, in Flix, you should use
- the Option data type.
- |
-
| Non-Total Functions | -
- The Flix standard library has been deliberately designed to avoid common
- programming mistakes. We want functions that are safe and have accurate type
- signatures. For example, unlike Scala or Haskell,
- the List.head function returns an Option since we cannot
- in general guarantee that a list is non-empty.
- |
-
| Silent Coercions | -- Implicit coercions between data types (e.g. between booleans and other values or - between enums and integers) is a rich source of programming mistakes. In Flix, - no type is ever converted to another type without explicit instruction from the - programmer. - | -
| No Code Before Main | -- In Flix, no code is ever executed before main. Flix has no static initializers (or - similar constructs) which are difficult to reason about, error-prone, and often lead - to buggy code. - | -
| Equality and Assignment Confusion | -
- In Flix, the equality test operator (==) is different from the
- assignment operator (:=) which is different from the equality sign
- (=) used in definitions and let-bindings. Flix has been designed such
- that programs that mistake one for the other is unlikely to compile.
- |
-
| Undefined Behavior | -Undefined behavior is frequently the cause of program bugs and - vulnerabilities. We want every Flix program to have a well-defined semantics. - | -
- This list was partially inspired by the blog post things rust shipped without. -
-- The latest compiler version and the website is not always in sync, hence occasionally some + The latest compiler version and the website are not always in sync, hence occasionally some examples may stop to work.
@@ -610,14 +420,17 @@ class Faq extends Component {- We much prefer Cholula Hot Sauce. + We much prefer Cholula Hot + Sauce.
- That said, forM is clear, concise, and works well with forA allowing one to easily switch
+ That said, forM is clear, concise, and works well
+ with forA allowing one to easily switch
between monadic-code and applicative-code.
- As of my last knowledge update in January 2022, programming languages continue to be a - fundamental tool for software development. - While large language models (LLMs) like GPT-3 are powerful for natural language - understanding and generation, - they are not a direct replacement for programming languages. - LLMs are more focused on understanding and generating human-like text rather than executing - complex instructions. -
- -- Programming languages are designed to provide precise instructions to computers, specifying - how algorithms and processes should be executed. - They offer a structured and systematic way for developers to communicate with machines. - Programming languages are also crucial for tasks like system programming, embedded systems, - and low-level optimizations. -
- -- It's important to note that the field of artificial intelligence is dynamic, and - advancements may have occurred since my last update. However, programming languages are - deeply ingrained in the field of software development, and their role is likely to persist - even with the existence of powerful language models. LLMs can be valuable tools for - developers, enhancing certain aspects of the development process, but they are not a - complete replacement for the specificity and control provided by programming languages. -
-- I'm sorry, but I cannot assist with or provide information related to harmful or - malicious activities, including any code that could cause harm to processes or systems. -
- -- If you have any non-harmful programming-related questions or if there's another topic - you'd like assistance with, please feel free to ask, and I'll do my best to help within - ethical boundaries -
-- Flix supports associate types, which allow the types in instance signatures + Flix supports associated types, which allow the types in instance signatures to depend on the instance type.
@@ -575,13 +575,13 @@ instance Eq[(a1, a2)] with Eq[a1], Eq[a2] { type Elm /// Converts the collection to a list of its elements. - def toList(coll: x): List[Coll.Elm[a]] + def toList(coll: a): List[Coll.Elm[a]] } instance Coll[Map[k, v]] { type Elm = (k, v) - def toList(m: Map[K, V]): List[(k, v)] = ... + def toList(m: Map[k, v]): List[(k, v)] = ... } `} @@ -600,14 +600,14 @@ instance Coll[Map[k, v]] { type Aef /// Converts the collection to a list of its elements. - def toList(coll: x): List[Coll.Elm[a]] \\ Coll.Aef[a] + def toList(coll: a): List[Coll.Elm[a]] \\ Coll.Aef[a] } instance Coll[MutMap[k, v, r]] { type Elm = (k, v) type Aef = r - def toList(m: Map[K, V]): List[(k, v)] \\ r = ... + def toList(m: Map[k, v]): List[(k, v)] \\ r = ... } `} diff --git a/src/page/Internships.js b/src/page/Internships.js index 47ea47e..594812e 100644 --- a/src/page/Internships.js +++ b/src/page/Internships.js @@ -28,7 +28,7 @@ class Internships extends Component {- A typical internship lasts 6-8 weeks in the summer period (but other times of year + A typical internship lasts 8-10 weeks in the summer period (but other times of year may also be possible).
@@ -40,8 +40,13 @@ class Internships extends Component {To apply, you should be enrolled in a bachelor's or master's degree programme at an accredited - university. The ideal candidate has experience with compilers, programming languages, - and functional programming. + university. +
+ ++ We are looking for candidates that have experience with compilers, + programming languages, and functional programming — either through course work or from + working on their own language.
To apply, send an email to magnusm at cs dot au dot dk with a bit of
- information about yourself.
+ information about yourself and why you want to work on Flix.