MadsTorgersen commented 9 years ago

C# Design Meeting Notes for Jan 21, 2015

Quotes of the day:

Live broadcast of design meetings: we could call it C#-SPAN

We've made it three hours without slippery slopes coming up!

Agenda

This is the first design meeting for the version of C# coming after C# 6. We shall colloquially refer to it as C# 7. The meeting focused on setting the stage for the design process and homing in on major themes and features.

Design process
Themes
Features

1. Design process

We have had great success sharing design notes publicly on CodePlex for the last year of C# 6 design. The ability of the community to see and respond to our thinking in real time has been much appreciated.

This time we want to increase the openness further:

we involve the community from the beginning of the design cycle (as per these notes!)
in addition to design notes (now issues on GitHub) we will maintain feature proposals (as checked-in Markdown documents) to reflect the current design of the feature
we will consider publishing recordings of the design meetings themselves, or even live streaming
we will consider adding non-Microsoft members to the design team.
Design team

The C# 7 design team currently consists of

Anders, as the chief language architect, has ultimate say, should that ever become necessary. Mads, as the language PM for C#, pulls together the agenda, runs the meetings and takes the notes. (Oooh, the power!)

To begin with, we meet 4 hours a week as we decide on the overall focus areas. There will not be a separate Visual Basic design meeting during this initial period, as many of the overall decisions are likely to apply to both and need to happen in concert.

Feature ideas

Anyone can put a feature idea up as an issue on GitHub. We'll keep an eye on those, and use them as input to language design.

A way to gauge interest in a feature is to put it up on UserVoice, where there's a voting system. This is important, because the set of people who hang out in our GitHub repo are not necessarily representative of our developer base at large.

Design notes

Design notes are point-in-time documents, so we will put them up as issues on GitHub. For a period of time, folks can comment on them and the reactions will feed into subsequent meetings.

Owners and proposals

If the design team decides to move on with a feature idea, we'll nominate an owner for it, typically among the design team members, who will drive the activities related to the design of that feature: gathering feedback, making progress between meetings, etc. Most importantly, the owner will be responsible for maintaining a proposal document that describes the current state of that feature, cross-linking with the design notes where it was discussed.

Since the proposals will evolve over time, they should be documents in the repo, with history tracked. When the proposal is first put up, and if there are major revisions, we will probably put up an issue too, as a place to gather comments. There can also be pull requests to the proposals.

We'll play with this process and find a balance.

Other ways of increasing openness

We are very interested in other ideas, such as publishing recordings (or even live streaming?) of the design meeting themselves, and inviting non-Microsoft luminaries, e.g., from major players in the industry, onto the design team itself. We are certainly open to have "guests" (physical or virtual) when someone has insights that we want to leverage.

However, these are things we can get to over time. We are not going to do them right out of the gate.

Decisions

It's important to note that the C# design team is still in charge of the language. This is not a democratic process. We derive immense value from comments and UserVoice votes, but in the end the governance model for C# is benevolent dictatorship. We think design in a small close-knit group where membership is long-term is the right model for ensuring that C# remains tasteful, consistent, not too big and generally not "designed by committee".

If we don't agree within the design team, that is typically a sign that there are offline activities that can lead to more insight. Usually, at the end of the day, we don't need to vote or have the Language Allfather make a final call.

Prototypes

Ideally we should prototype every feature we discuss, so as to get a good feel fro the feature and allow the best possible feedback from the community. That may note be realistic, but once we have a good candidate feature, we should try to fly it.

The cost of the prototyping is an issue. This may be feature dependent: Sometimes you want a quick throwaway prototype, sometimes it's more the first version of an actual implementation.

Could be done by a member of the design team, the product team or the community.

Agenda

It's usually up to Mads to decide what's ready to discuss. Generally, if a design team member wants something on the agenda, they get it. There's no guarantee that we end up following the plan in the meeting; the published notes will just show the agenda as a summary of what was actually discussed.

2. Themes

If a feature is great, we'll want to add it whether it fits in a theme or not. However, it's useful to have a number of categories that we can rally around, and that can help select features that work well together.

We discussed a number of likely themes to investigate for C# 7.

Working with data

Today’s programs are connected and trade in rich, structured data: it’s what’s on the wire, it’s what apps and services produce, manipulate and consume.

Traditional object-oriented modeling is good for many things, but in many ways it deals rather poorly with this setup: it bunches functionality strongly with the data (through encapsulation), and often relies heavily on mutation of that state. It is "behavior-centric" instead of "data-centric".

Functional programming languages are often better set up for this: data is immutable (representing information, not state), and is manipulated from the outside, using a freely growable and context-dependent set of functions, rather than a fixed set of built-in virtual methods. Let’s continue being inspired by functional languages, and in particular other languages – F#, Scala, Swift – that aim to mix functional and object-oriented concepts as smoothly as possible.

Here are some possible C# features that belong under this theme:

pattern matching
tuples
"denotable" anonymous types
"records" - compact ways of describing shapes
working with common data structures (List/Dictionary)
extension members
slicing
immutability
structural typing/shapes?

A number of these features focus on the interplay between "kinds of types" and the ways they are used. It is worth thinking of this as a matrix, that lets you think about language support for e.g. denoting the types (type expressions), creating values of them (literals) and consuming them with matching (patterns) :

Type	Denote	Create	Match
General	`T`	`new T()`, `new T { x = e }`	`T x`, `var x`, `*`
Primitive	`int`, `double`, `bool`	`5`, `.234`, `false`	`5`, `.234`, `false`
String	`string`	`"Hello"`	`"Hello"`
Tuple	`(T1, T2)`	`(e1, e2)`	`(P1, P2)`
Record	`{ T1 x1, T2 x2 }`	`new { x1 = e1, x2 = e2 }`	`{ x1 is P1, x2 is P2 }`
Array	`T[]`	`new T[e]`, `{ e1, e2 }`	`{ P1, P2 }`, `P1 :: P2`
List	?	?	?
Dictionary	?	?	?
...

A lot of the matrix above is filled in with speculative syntax, just to give an idea of how it could be used.

We expect to give many of the features on the list above a lot of attention over the coming months: they have a lot of potential for synergy if they are designed together.

Performance and reliability (and interop)

C# and .NET has a heritage where it sometimes plays a bit fast and loose with both performance and reliability.

While (unlike, say, Java) it has structs and reified generics, there are still places where it is hard to get good performance. A top issue, for instance is the frequent need to copy, rather than reference. When devices are small and cloud compute cycles come with a cost, performance certainly starts to matter more than it used to.

On the reliability side, while (unlike, say, C and C++) C# is generally memory safe, there are certainly places where it is hard to control or trust exactly what is going on (e.g., destruction/finalization).

Many of these issues tend to show up in particular on the boundary to unmanaged code - i.e. when doing interop. Having coarse-grained interop isn't always an option, so the less it costs and the less risky it is to cross the boundary, the better.

Internally at Microsoft there have been research projects to investigate options here. Some of the outcomes are now ripe to feed into the design of C# itself, while others can affect the .NET Framework, result in useful Roslyn analyzers, etc.

Over the coming months we will take several of these problems and ideas and see if we can find great ways of putting them in the hands of C# developers.

Componentization

The once set-in-stone issue of how .NET programs are factored and combined is now under rapid evolution.

With generalized extension members as an exception, most work here may not fall in the language scope, but is more tooling-oriented:

generating reference assemblies
static linking instead of IL merge
determinism
NuGet support
versioning and adaptive light-up

This is a theme that shouldn't be driven primarily from the languages, but we should be open to support at the language level.

Distribution

There may be interesting things we can do specifically to help with the distributed nature of modern computing.

Async sequences: We introduced single-value asynchrony in C# 5, but do not yet have a satisfactory approach to asynchronous sequences or streams
Serialization: we may no longer be into directly providing built-in serialization, but we need to make sure we make it reasonable to custom-serialize data - even when it's immutable, and without requiring costly reflection.

Also, await in catch and finally probably didn't make it into VB 14. We should add those the next time around.

Metaprogramming

Metaprogramming has been around as a theme on the radar for a long time, and arguably Roslyn is a big metaprogramming project aimed at writing programs about programs. However, at the language level we continue not to have a particularly good handle on metaprogramming.

Extention methods and partial classes both feel like features that could grow into allowing generated parts of source code to merge smoothly with hand-written parts. But if generated parts are themselves the result of language syntax - e.g. attributes in source code, then things quickly get messy from a tooling perspective. A keystroke in file A may cause different code to be generated into file B by some custom program, which in turn may change the meaning of A. Not a feedback loop we're eager to have to handle in real time at 20 ms keystroke speed!

Oftentimes the eagerness to generate source comes from it being too hard to express your concept beautifully as a library or an abstraction. Increasing the power of abstraction mechanisms in the language itself, or just the syntax for applying them, might remove a lot of the motivation for generated boilerplate code.

Features that may reduce the need for boilerplate and codegen:

Virtual extension methods/default interface implementations
Improvements to generic constraints, e.g.:
- generic constructor constraints
- delegate and enum constraints
- operators or object shapes as constraints (or interfaces), e.g. similar to C++ concepts
mixins or traits
delegation
Null

With null-conditional operators such as x?.y C# 6 starts down a path of more null-tolerant operations. You could certainly imagine taking that further to allow e.g. awaiting or foreach'ing null, etc.

On top of that, there's a long-standing request for non-nullable reference types, where the type system helps you ensure that a value can't be null, and therefore is safe to access.

Importantly such a feature might go along well with proper safe nullable reference types, where you simply cannot access the members until you've checked for null. This would go great with pattern matching!

Of course that'd be a lot of new expressiveness, and we'd have to reconcile a lot of things to keep it compatible. In his blog, Eric Lippert mentions a number of reasons why non-nullable reference types would be next to impossible to fully guarantee. To be fully supported, they would also have to be known to the runtime; they couldn't just be handled by the compiler.

Of course we could try to settle for a less ambitious approach. Finding the right balance here is crucial.

Themeless in Seattle

Type providers: This is a whole different kind of language feature, currently known only from F#. We wouldn't be able to just grab F#'s model though; there'd be a whole lot of design work to get this one right!

Better better betterness: In C# we made some simplifications and generalizations to overload resolution, affectionately known as "better betterness". We could think of more ways to improve overload resolution; e.g. tie breaking on staticness or whether constraints match, instead of giving compiler errors when other candidates would work.

Scripting: The scripting dialect of C# includes features not currently allowed in C# "proper": statements and member declarations at the top level. We could consider adopting some of them.

params IEnumerable.

Binary literals and digit separators.

3. Features

The Matrix above represents a feature set that's strongly connected, and should probably be talked about together: we can add kinds of types (e.g. tuples, records), we can add syntax for representing those types or creating instances of them, and we can add ways to match them as part of a greater pattern matching scheme.

Pattern matching

Core then is to have a pattern matching framework in the language: A way of asking if a piece of data has a particular shape, and if so, extracting pieces of it.

if (o is Point(var x, 5)) ...

There are probably at least two ways you want to use "patterns":

As part of an expression, where the result is a bool signaling whether the pattern matched a given value, and where variables in the pattern are in scope throughout the statement in which the pattern occurs.
As a case in a switch statement, where the case is picked if the pattern matches, and the variables in the pattern are in scope throughout the statements of that case.

A strong candidate syntax for the expression syntax is a generalization of the is expression: we consider the type in an is expression just a special case, and start allowing any pattern on the right hand side. Thus, the following would be valid is expressions:

if (o is Point(*, 5) p) Console.WriteLine(o.x);
if (o is Point p) Console.WriteLine(p.x);
if (p is (var x, 5) ...

Variable declarations in an expression would have the same scope questions as declaration expressions did.

A strong candidate for the switch syntax is to simply generalize current switch statements so that

the switch expression can be any type
the case labels can contain patterns, not just constants
the cases are checked in order of appearance, since they can now overlap

switch (o) {
case string s:
    Console.WriteLine(s);
    break;
case int i:
    Console.WriteLine($"Number {i}");
    break;
case Point(int x, int y):
    Console.WriteLine("({x},{y})");
    break;
case null:
    Console.WriteLine("<null>);
    break
}

Other syntaxes you can think of:

Expression-based switch: An expression form where you can have multiple cases, each producing a result value of the same type.

Unconditional deconstruction: It might be useful to separate the deconstruction functionality out from the checking, and be able to unconditionally extract parts from a value that you know the type of:

(var x, var y) = getPoint();

There is a potential issue here where the value could be null, and there's no check for it. It's probably ok to have a null reference exception in this case.

It would be a design goal to have symmetry between construction and deconstruction syntaxes.

Patterns at least have type testing, value comparison and deconstruction aspects to them.

There may be ways for a type to specify its deconstruction syntax.

In addition it is worth considering something along the lines of "active patterns", where a type can specify logic to determine whether a pattern applies to it or not.

Imagine positional deconstruction or active patterns could be expressed with certain methods:

class Point {
    public Point(int x, int y) {...}
    void Deconstruct(out int x, out int y) { ... }
    static bool Match(Point p, out int x, out int y) ...
    static bool Match(JObject json, out int x, out int y) ...
}

We could imagine separate syntax for specifying this.

One pattern that does not put new requirements on the type is matching against properties/fields:

if (o is Point { X is var x, Y is 0 }) ...

Open question: are the variables from patterns mutable?

This has a strong similarity to declaration expressions, and they could coexist, with shared scope rules.

Records

Let's not go deep on records now, but we are aware that we need to reconcile them with primary constructors, as well as with pattern matching.

Array Slices

One feature that could lead to a lot of efficiency would be the ability to have "windows" into arrays - or even onto unmanaged swaths of memory passed along through interop. The amount of copying that could be avoided in some scenarios is probably very significant.

Array slices represent an interesting design dilemma between performance and usability. There is nothing about an array slice that is functionally different from an array: You can get its length and access its elements. For all intents and purposes they are indistinguishable. So the best user experience would certainly be that slices just are arrays - that they share the same type. That way, all the existing code that operates on arrays can work on slices too, without modification.

Of course this would require quite a change to the runtime. The performance consequences of that could be negative even on the existing kind of arrays. As importantly, slices themselves would be more efficiently represented by a struct type, and for high-perf scenarios, having to allocate a heap object for them might be prohibitive.

One intermediate approach might be to have slices be a struct type Slice, but to let it implicitly convert to T[] in such a way that the underlying storage is still shared. That way you can use Slice for high performance slice manipulation (e.g. in recursive algorithms where you keep subdividing), but still make use of existing array-based APIs at the cost of a boxing-like conversion allocating a small object.

ref locals and ref returns

Just like the language today has ref parameters, we could allow locals and even return values to be by ref. This would be particularly useful for interop scenarios, but could in general help avoid copying. Essentially you could return a "safe pointer" e.g. to a slot in an array.

The runtime already fully allows this, so it would just be a matter of surfacing it in the language syntax. It may come with a significant conceptual burden, however. If a method call can return a variable as opposed to a value, does that mean you can now assign to it?:

m(x, y) = 5;

You can now imagine getter-only properties or indexers returning refs that can be assigned to. Would this be quite confusing?

There would probably need to be some pretty restrictive guidelines about how and why this is used.

readonly parameters and locals

Parameters and locals can be captured by lambdas and thereby accessed concurrently, but there's no way to protect them from shared-mutual-state issues: they can't be readonly.

In general, most parameters and many locals are never intended to be assigned to after they get their initial value. Allowing readonly on them would express that intent clearly.

One problem is that this feature might be an "attractive nuisance". Whereas the "right thing" to do would nearly always be to make parameters and locals readonly, it would clutter the code significantly to do so.

An idea to partly alleviate this is to allow the combination readonly var on a local variable to be contracted to val or something short like that. More generally we could try to simply think of a shorter keyword than the established readonly to express the readonly-ness.

Lambda capture lists

Lambda expressions can refer to enclosing variables:

var name = GetName();
var query = customers.Where(c => c.Name == name);

This has a number of consequences, all transparent to the developer:

the local variable is lifted to a field in a heap-allocated object
concurrent runs of the lambda may access and even modify the field at the same time
because of implementation tradeoffs the content of the variable may be kept live by the GC, sometimes even after lambdas directly using them cease to exist.

For these reasons, the recently introduced lambdas in C++ offer the possibility for a lambda to explicitly specify what can be captured (and how). We could consider a similar feature, e.g.:

var name = GetName();
var query = customers.Where([name]c => c.Name == name);

This ensures that the lambda only captures name and no other variable. In a way the most useful annotation would be the empty [], making sure that the lambda is never accidentally modified to capture anything.

One problem is that it frankly looks horrible. There are probably other syntaxes we could consider. Indeed we need to think about the possibility that we would ever add nested functions or class declarations: whatever capture specification syntax we come up with would have to also work for them.

C# always captures "by reference": the lambda can observe and effect changes to the original variable. An option with capture lists would be to allow other modes of capture, notable "by value", where the variable is copied rather than lifted:

var name = GetName();
var query = customers.Where([val name]c => c.Name == name);

This might not be too useful, as it has the same effect as introducing another local initialized to the value of the original one, and then capture that instead.

If we don't want capture list as a full-blown feature, we could consider allowing attributes on lambdas and then having a Roslyn analyzer check that the capture is as specified.

Method contracts

.NET already has a contract system, that allows annotation of methods with pre- and post-conditions. It grew out of the Spec# research project, and requires post-compile IL rewriting to take full effect. Because it has no language syntax, specifying the contracts can get pretty ugly.

It has often been proposed that we should add specific contract syntax:

public void Remove(string item)
    requires item != null
    ensures Count >= 0
{
   ...
}

One radical idea is for these contracts to be purely runtime enforced: they would simply turn into checks throwing exceptions (or FailFast'ing - an approach that would need further discussion, but seems very attractive).

When you think about how much code is currently occupied with arguments and result checking, this certainly seems like an attractive way to reduce code bloat and improve readability.

Furthermore, the contracts can produce metadata that can be picked up and displayed by tools.

You could imagine dedicated syntax for common cases - notably null checks. Maybe that is the way we get some non-nullability into the system?

svick commented 9 years ago

So, array slices are basically an improved version of ArraySegment<T>?

s-aida commented 9 years ago

While Pattern matching is awesome, I'm still disappointed to know that break; is mandatory. I'm pretty sure you've had lots of discussions in codeplex or elsewhere (which I've never read) so no need for rationale or history behind this, but that's about the only part of C# which I have to feel is lamely designed.

MadsTorgersen commented 9 years ago

@svick: Yeah I guess. With possible language support. @shunsukeaida: this is totally a straw man syntax - we aren't even close to settling on a specific syntax. However, if we want to fit pattern matching into switch, we have to kind of blend in well. One might argue that we could just get rid of break, and make it implicit and the end of a case block. That's certainly worth considering.

agocke commented 9 years ago

@MadsTorgersen Something not mentioned in relation to pattern matching: have discriminated unions been considered?

ashmind commented 9 years ago

@MadsTorgersen @shunsukeaida I proposed a better switch syntax on CodePlex (independently from destructuring) https://roslyn.codeplex.com/discussions/565550, though it might not work with labels.

I think a better switch syntax is rather important whether pattern matching gets into the language or not.

ashmind commented 9 years ago

Anyone can put a feature idea up as an issue on GitHub. We'll keep an eye on those, and use them as input to language design.

Is there some way to distinguish those from e.g. Roslyn bugs? E.g. some title prefix or so? Labels are good but I can't assign them when reporting the issue I believe.

And I think it would be cool to have some minimal format for it, e.g. at least require the description of the problem, a proposed solution, some use cases using the proposed solution. BCL team seems to be doing a good job there with speclets.

Also I think some certain yes/no on proposals would be great, so that any GitHub proposal is not only considered, but either rejected, accepted for prototyping, or sent to rework due to some design constraints. Again I think BCL team is doing great with their periodic review approach.

jods4 commented 9 years ago

Here's a theme suggestion.

With multi-core becoming commonplace and the nice Task and TPL apis, I see more and more concurrent programming, even if just a few (2-3) background threads.

It would be very nice to get some help from C# regarding correctness here. Immutable data structures are a big help, but for mutable data the main issue is that nothing formally indicates which thread has data ownership, or if data is shared in a read-only fashion, or if data must be accessed inside a lock. In complex programs and large teams this stuff is really hard to get right.

C# was a rockstar when it introduced async to help write asynchronous code. I know it's a very hard topic but I would love to see C# bring safe concurrent programming to mainstream.

daveaglick commented 9 years ago

I would love to see more native language support for metaprogramming. Mixins and/or traits would certainly help with code reuse challenges and I would personally love to see one of these make it in. While T4 templates work okay for code generation, they often feel clunky and hacky (for example, even though it's a tooling thing, the fact that VS addins are needed just to consistently evaluate them is a clue something is off). Some kind of language support or hook for code generation would be welcome. I would also love to see some sort of rudimentary support for aspects and weaving. PostSharp works well and has a nice API, but requires extra libraries and IL rewriting. Fody aspects work, but there's a really steep learning curve. I'm guessing full AOP support would be way out of scope for this language update, but incremental support or even hooks (perhaps in Roslyn) would be welcomed.

There was also mention of improved overload resolution. This has been a particular pain point of mine, especially when dealing with generics and extension methods. For example, it would be awesome if generic constraints were taken into account when determining a match. I understand the reason why they're not considered (I.e., not part of the method signature) but it would be great if there were a way to overcome this.

jods4 commented 9 years ago

+1 AOP can have lots of useful applications, some of them have been requested by the community for a long time. A few examples:

Automatic INotifyPropertyChanged implementation.
Automatic check for null argument on public methods (I think the ASP.NET team does that for vNext).
Automatic implementation of a "IsDirty" flag on model entities.
Wrapping logging, security or error handlers around services. And the list goes on!

svick commented 9 years ago

@jods4 Of interest may be the fact that @jaredpar, one of the coauthors of the Uniqueness and Reference Immutability for Safe Parallelism paper, is now on the Roslyn team and said:

a part of my job will be trying to integrate some of the goodness that came out of our research back into the language

bojanrajkovic commented 9 years ago

I think let might be a useful convention for readonly var, or even const, though that's probably leading too much down the path of too many meanings for const that C/C++ fall prey to.

biboudis commented 9 years ago

@MadsTorgersen What do you mean by delegation? Following theories of incomplete objects that get composed by delegation with proper late-binding semantics maybe?

alexpolozov commented 9 years ago

Something not mentioned in relation to pattern matching: have discriminated unions been considered?

@agocke Yes, in the original proposal discriminated unions were implemented with record subclassing — similar to case classes in Scala.

@MadsTorgersen By the way, regarding the state of pattern matching. If you do decide to consider them for C# 7, will the first prototype be merged into the current Github source as a feature branch? Will the community be able to contribute to such prototypes? I think you've mentioned before that you want to simplify the PR process in a couple of months, so that external feature prototyping/bug fixing would be easier.

ghost commented 9 years ago

Non-nullable reference types might be better served with another approach. Assess the feasibility of CLR support for describing a non-null reference type. Even if C# cannot currently take advantage of non-nullable reference types there might be another .NET language that figures out how.

ghost commented 9 years ago

Not about new Visual Basic language version? Is Visual Basic coming to die or was placed in the background? I love VB and I'd love to hear about new features :smile: for it.

ghost commented 9 years ago

@HoracioFilho They mentioned that they are planning the features for both C# and VB right now.

To begin with, we meet 4 hours a week as we decide on the overall focus areas. There will not be a separate Visual Basic design meeting during this initial period, as many of the overall decisions are likely to apply to both and need to happen in concert.

damageboy commented 9 years ago

One thing HPC like devs could benefit from, is have the ability to "construct" primitive array types off of memory mapped file, like java's nio supports...

I realize this is mainly a runtime feature, rather than a c# feature, is there a "better" place to discuss those?

jvlppm commented 9 years ago

Something that I missed before, and reactive extensions almost gave it to me: Ability to combine await and yield, resulting in an asynchronous sequence.

async var GenerateNumbers()
{
    int i = 0;
    while (true)
    {
        await Task.Delay(TimeSpan.FromSeconds(1));
        yield return i++;
    }
}

foreach(await int item in GenerateNumbers())
{
    WriteLine("Item: \{item}");
}

jonpryor commented 9 years ago

readonly var could be abbreviated as let, which is already a contextual keyword within query comprehension expressions and semantically declares a readonly variable...

Cellivar commented 9 years ago

+1 to the Contracts system. There's already a similar concept with type constraints for generics.

class Foo<T> where T : IComparable { }

An extension to this pattern for functions would be intuitive to anyone familiar with constraints.

void Foo(string bar) where bar != null, Count > 0 { }

Would separate 'ensures' and 'requires' actually be necessary? Could the statement be inferred?

MadsTorgersen commented 9 years ago

@agocke: We are trying to come up with a general framework for pattern matching over classes. Ideally there wouldn't be a new "thing" in the language called algebraic datatypes or discriminated unions. Instead there may be compact syntax for describing hierarchies of simple value types (that would happen to be well suited for pattern matching) as a shorthand for class declarations.

@ashmind: yeah, maybe it's time to revamp switch. Worth a think.

@ashmind: I think we'll go through and apply the labels. We'll build up our processes as we go - the BCL team are ahead of us here, and we'll draw on their experiences.

@jods4: safe concurrency is an insanely hard problem. Not sure what we'd do.

@somedave: realistically I can't see us doing something about metaprogramming at the language level before we have a very clear idea of a) what would really move the needle for a lot of developers, and b) how to deal with it at a tooling level.

@jods4: AOP is notoriously hard to reason about for a developer. It's a very big hammer.

@bojanrajkovic, @jonpryor: let is certainly a candidate.

@biboudis: Full-blown delegation (with this-binding and everything) seems out of the question - that's a whole alternative mechanism to inheritance. But some languages have this nifty little implement-interface-by-redirection trick. Not sure if it's worthwhile or even sane.

@apskim: No clear decisions yet as to how we will share prototypes. For the time being, most of the team is heads down making the upcoming version great. This further-out blue sky stuff won't get priority until that's out the door. Yes, would be great to involve the community in prototyping.

@Romoku: I'd love to get non-nullable reference types into the CLR. It seems a tall order even there: what's inside an array of non-nullable reference type when it's first created? etc.

@HoracioFilho: Yeah, @Romoku has it right. It doesn't make sense to start separate efforts. Let's get the broad brushstrokes worked out and then split out to language-lawyer on the syntax. :-)

@damageboy: If I understand you correctly, yes that would be an awesome interop feature: Array-typed "windows" onto native data. Privileged (and unsafe!) methods would create them, but be able to share them with your code.

@jvlppm: Yes! When I mentioned "async sequences" this is the kind of thing I had in mind.

@MrDoomBringer: Whatever the syntax, it would certainly occupy the same "region" of a method declaration as the constraints. Maybe it makes sense to merge them.

agocke commented 9 years ago

@apskim @MadsTorgersen

Technically, you're correct, but there are actually no new features there specific to discriminated unions. :) The method of generating discriminated unions in the draft specification has one (to me very significant) weakness -- it has no specification of completeness checking in switch statements right now. One way to add this would be to add specialized support just for discriminated unions, for example coming up with a new syntax for enum which desugers to a discriminated union behind the scenes.

Alternatively, I would like a spec addition that guarantees that the following data structure provides completeness checking in a switch statement.

public abstract class BinaryTree<T>
{
  private BinaryTree() {};
  public sealed class Leaf<U>() : BinaryTree<U>;
  public sealed class Node<U>(U item, U left, U right) : BinaryTree<U>;
}

It's a little more heavyweight than some de novo syntax, but it works. Obviously I would be volunteering to implement this warning. ;)

ghost commented 9 years ago

@MadsTorgersen From what I understand of how Eric Lippert and Spec# describe non-null references types the contents of the array cannot be null when observed during the program execution. This means that up to the point of observance the state can be undetermined, but there must be a value at the point of observance.

jamesbascle commented 9 years ago

@somedave Agreed on the AOP.

PostSharp is amazing, but It can be hard to keep it working right on a large team of developers, and it's pretty expensive for all the bells and whistles for a large team as well, even if only a couple of people are even writing the code.

Some language level support for generic argument and result manipulation on methods (like OnMethodBoundaryAspect) alone would be a huge step to meeting the need for metaprogramming and supplanting the wonkiness of PostSharp.

jamesbascle commented 9 years ago

@MadsTorgersen The PostSharp VisualStudio add-on seems to allow one to easily see which methods are manipulated by which aspects, both at the "This Method is manipulated by This Aspect" and "This Aspect manipulates These Methods".

We've saved ourselves literally thousands of lines of code, but it gets pretty expensive and it's a pretty obscure technical expense that is hard to justify to management types.

It would also be nice if C#/.NET were the first language/framework with anything approaching the kind of reach it has to REALLY have built-in first-class AOP facilities in a language. AspectJ and the things found in some of the dynamic languages or the LISPs are not exactly getting a lot of attention these days. I'd personally see it as a huge huge win for this new open design process from a developer perspective.

damageboy commented 9 years ago

@MadsTorgersen you understood me perfectly

ufcpp commented 9 years ago

How about Units of Measure (like F#)? From the perspective of performance and convenience, I prefer to use primitive values (int ,byte, etc.), but at the same time, from the perspective of reliability, I'd like to use user-defined values. This dilemma can be solved by Units of Measure.

chrisaut commented 9 years ago

+1 for Contracts from me.

The current solution seems half hearted and not really supported very well.

Non-nullness could IMO be left to contracts, IF there is a static analyzer that can enforce this.

If they are to become a language feature, common cases could be expressed much shorter. I'm thinking something like:

public int >= 0 M(string! s, double < 100 y, MyClass c)
    requires c.IsInitialized
{}

which would be the same as

public int M(string s, double y)
    requires s != null,
    requires y < 100,
    requires c..IsInitialized
    ensures result >= 0
{}

However it could looks a bit messy. Thoughts?

PS: the example

if (o is Point(*, 5) p) Console.WriteLine(o.x);

should that be

if (o is Point(*, 5) p) Console.WriteLine(p.x);

or am I misunderstanding?

tjdcs commented 9 years ago

Sorry to leave a non constructive comment here but I do have a question related to this process of specifying and designing C#7.

I am interested in watching and studying this process and caught the note about this being broadcast live. What is the best way to follow along and maybe watch these meetings?

I am an undergraduate student who just started a class on language specifications and design and it's definitely one of the most interesting topics I have studied thus far. I don't really use c# but I am really interested in studying the process of language design and that seems to be (partly or mostly) independent of the exact language being specified. I also think it's important, for my benefit, to note here that C# is obviously not a new language and so there is a lot of base work being built upon here.

Anyway, with that said I would love to follow along as closely as I can and would like some more information about how to do that.

Thanks, -TuckerD

RxDave commented 9 years ago

:+1: Method contracts! Yes please. Also really liked the idea of syntax for tuples and records. Pattern matching - even if just within switch case statements - would be fantastic as well.

jonschoning commented 9 years ago

If the 'pattern on the right hand side' allows matching on subclasses of the data on the left of this 'is' operator, then without having the capability of exhaustive checks, (as seems from what is described so far), imho the most important benefits of pattern matching are lost.

An 'exhaustive check' would mean that there would be a compiler warning if new subclasses were added to the data being matched on that didn't have cases for that new subclass in existing pattern matching code.

Actually, I'm thinking now it may be difficult to make exhaustive checks work for concrete classes because of ambiguity with the syntax allowing destructuring of the primary constructor.

Perhaps there could exhaustive checks only when matching on abstract classes if that causes less ambiguity.

Thinking about it some more, classes are not closed, so exhaustive checks would make no sense... This really is not what I would call pattern matching without discriminated unions. This should just be called "destructuring"

Plasma commented 9 years ago

Thanks @MadsTorgersen - a few various thoughts:

Introducing a "type safe basic type". Let me distinguish between: int customerId = 1; int billingId = 1;

Can we "mark" this integer (or Guid, or long, etc) of customerId (pulled from a database for example) as a "CustomerId" type?

public CustomerId : int { }

I want to prevent a method that takes:

public void Charge(int customerId)

... accidentally being passed in a billingId (also an INT).

Ideally my interface was:

public void Charge(CustomerId customerId)

... and passing in a billingId would result in a compilation error.

+1 to AOP integration (we use Castle Dynamic proxies right now). Really useful for authorization, caching, logging, ... cross cutting concerns. I think the developer experience is understandable when the debugger correctly shows the stack trace of calls being intercepted (which is what happens with Castle Windsor).
Ahead-of-time compilation of an entire ASP.NET website. I don't know if this is applicable to C# spec, but a way to "100% warm up, pre-load, be ready" a .NET website would be good. Right now even under Release mode, there is JIT, which I never want on a production site.
Let me use commas (or spaces for universal support?) in number definitions: const int BigNum = 1,000,000; // No need to count the zeros
I use Tuple<string, int> etc a fair bit; ideally there was a super nice way to return a tuple with named fields vs Item1/Item2 without making a new class

chrisaut commented 9 years ago

@plasma, regarding your "type safe basic type", what's wrong with:

public struct CustomerId {
     public int Id { get; }
     public CustomerId(int id) { this.Id = id; }
    // optional
    public static implicit int(CustomerId cId) => cId.Id;
}

If you need many of those, you can just create an abstract base class:

public abstract class IdBase {
     public int Id { get; }
     public IdBase (int id) { this.Id = id; }
    // optional
    public static implicit int(IdBase cId) => cId.Id;
}
public class CustomerId : Idbase {}

It's not a struct anymore, but it should hardly matter. It's a bit of typing, but with Record types and primary constructors it'll get so short it's essentially nothing.

public struct CustomerId(int id) {
     public int Id { get; } = id;
}

I believe the real feature you want is ValueType (pseudo?) - inheritance, a form a compiler enforced type aliasing. I don't think adding yet another form of "type" to the language is a good idea, it's getting kinda complicated as it is.

ryancerium commented 9 years ago

Apply the readonly modifier to classes, methods, and parameters to help support immutable data structures. Perhaps add the mutable keyword for readonly classes the way C++ uses it.

I've wanted static methods in an interface before, but generic constructor constraints might have been the underlying problem.

Could you implement constraints at compile time using a static analyzer? Like using an unassigned variable is a compiler warning (error?) now.

Discriminated unions like Rust has.

Interfaces with default implementations like Java just added (is thinking about adding?). Is this different than traits or mixins in large way?

What if you make the lambda capture list an extended portion of the parameter list?

var number = GetNumber();
var name = GetName();
var query = customers.Where((c, ref name, value number) => c.Name == name || c.Number == number);`

RichiCoder1 commented 9 years ago

@ryancerium :+1: for readonly class with mutable members.

What if you make the lambda capture list an extended portion of the parameter list?
var number = GetNumber();
var name = GetName();
var query = customers.Where((c, ref name, value number) => c.Name == name || c.Number == number);```

I'd be against this if only because of the cognative load of differentiating between captures and params. While [number, value name](c) => ... isn't the prettiest, it's at least a lot more immediately clear.

Fizzelen commented 9 years ago

How about adding private variables inside a property, to prevent side effects from modifying the the backing field directly

public int Id { int _id = 0; get { return _id; } set { _id = value; } }

leppie commented 9 years ago

The pattern matching construct should be an expression and not a statement like switch. My 2 cents.

RichiCoder1 commented 9 years ago

@Fizzelen wouldn't that just be a good old fashion property at this point? Or are you referring to doing something like:

public int Id
{
    int _id = 0;
    get { return _id * 2; }
    set { _id  = value / 2; }
}

tophallen commented 9 years ago

For Lambda Captures, perhaps something inline could be useful, such as:

var name = GetName();
var query = customers.Where(c => c.Name == [name]);

or

var name = GetName();
var query = customers.Where(c => c.Name == {name});

I would think something like this (perhaps not this exact syntax) could avoid the mess of having a capture list before the expression, and could allow it to capture sub properties from an object rather than the whole object as well, i.e

var person = GetPerson();
var query = customers.Where(c => c.Name == {person.Name});

So here, it would capture the value of person.name and not the whole person object.

This could also work fine with other capture options, such as by value, {val name} or {val person.Name} if this feature were desired, not sure I personally see the benefit of that either as we can create a variable to capture instead.

This might still get tricky with nested functions or nested class declarations, but if the capture was { person.Name.ToString()} - perhaps it would be aware and capture the property and call on it separately? or perhaps we would have to call it like so: {person.Name}.ToString() - and then calling {person.Name.ToString()} would actually capture the ToString() call.

RichiCoder1 commented 9 years ago

@tophallen Something like this maybe?:

var name = GetName();
var query = customers.Where(c => c.Name == |val name|);

Though I'm not sure what syntax would actually be happy point.

ashmind commented 9 years ago

@Fizzelen I suggest just having an auto-field called field instead, e.g.

public int Id {
    get { return field; }
    set { field = value; }
}

tpetrina commented 9 years ago

I would welcome any sort of "const"ness on method parameters. More powerful switch statement that can handle expression non-compile time constants would also help a lot.

jamesbascle commented 9 years ago

@tophallen, @RichiCoder1 IMO, the folks talking about lambda capture syntax seem to be missing the point by talking about in-lining this new syntax. We don't really get anything that way. The nice part about the way @MadsTorgersen originally typed it, is that it is "foward declared" before it is actually used, much like a method parameter or field or whatever, so to use it in the executing part of the lambda, you have to declare that you want to bring it into the closure before you actually use it.

All you do by inlining it into the operative part of the lambda is introduce some more complicated syntax that doesn't gain us anything. I'm not saying the original syntax is pretty, or should be used, but that is a very clear benefit of it.

s-aida commented 9 years ago

As a less ambitious approach than non-nullable reference types, I think I'm okay with having a shorter approach to check that a variable neither refers to null nor has a value generally considered as blank (e.g. string is empty or array length is zero). String.IsNullOrEmpty(foo) is still a bit tiring even with stating using.

Maybe akin to ask if it's ever possible for null to call a method though.

tophallen commented 9 years ago

@jamesbascle true, but the compiler could look at the expression for pattern matching, and forward declare on build, and that would allow it to not capture the whole scope onto the heap, and you could pick sub properties, or call a method to capture a specific part of a class, or to capture a state, so while the syntax might get a bit cluttered (I see what you mean though, as this would especially be annoying when you use the same captured variable multiple times (i.e. c => c.StartDate <= [myDate] && c.EndDate >= [myDate]), you could get the benefits of explicit capture. Forward declaring would make for less work however.

To your point perhaps the suggestion by @ryancerium could be tweaked to keep the readability and look something like this?

    var name = GetName();
    var query = customers.Where(([name], c) => c.Name == name);

or

    var person = GetPerson();
    var query = customers.Where(([person.Name],[val person.Id], c) => 
                                   c.Name == person.Name && c.Id == person.Id);
    //or if we wanted captures to be one thing
    var query = customers.Where(([person.Name, val person.Id], c) => 
                                   c.Name == person.Name && c.Id == person.Id);

In this way it keeps the expression cleaner, but also makes it clear that it is a captured value and not coming from the expression itself, but is to be captured in the heap for the expression to use. The issue that would be left with this method would be, should [] be required to ensure that you capture nothing? Or could it be specified some other way?

der-daniel commented 9 years ago

ObservableCollection.AddRange()

jamesbascle commented 9 years ago

@tophallen I could be wrong, but I'm pretty sure that the compiler-generated lambda helper classes only bring "this" (whatever class it is in that context) and any referenced variables, so I think it already doesn't pull in the entire calling scope, just those it needs.

mariusGundersen commented 9 years ago

I'm glad you are thinking about virtual extension methods, it would be great to have dynamic binding outside of methods. For example something like

public string DoSomething(virtual ISomething s){
  return "this is the base method";
}
public string DoSomething(SomethingImplementation s){
  return "s is SomethingImplementation";
}
public string DoSomething(SomethingElse s){
  return "s is SomethingElse";
}
//...
ISomething s;
s = new SomethingImplementation();
DoSomething(s)//returns "s is SomethingImplementation";
s = new SomethingElse();
DoSomething(s)//returns "s is SomethingElse";

This is possible to do with dynamic, but dynamic isn't restrictive enough, as it will match any type. With virtual it would be possible to match only on subclasses/implementations of a type. In the above case the method marked as virtual would be the fallback in the case that a new type is introduced which doesn't have an implementation of DoSomething

mariusGundersen commented 9 years ago

C# really need a stricter version of enum, one that cannot be cast from int so the compiler can do static analysis on it, for example in switch/case:

enum Maybe{Yep, Nope};
//...
public bool IsYep(Maybe maybe){
  switch(maybe){
    case Yep: return true;
    case Nope: return false;
  }
}

The above code will not compile, since the compiler is unable to tell that all possible paths through the IsYep method has a return value. The way to fix this is to add a default case, which is really not needed. A stricter implementation of enum would have the above case compiling without error but not compiling in the following case (because not all the enum cases have been tested in the switch/case):

enum Color{Red, Green, Blue};
//...
public bool IsRed(Color color){
  switch(color){
    case Red: return true;
    case Green: return false;
  }
}

tophallen commented 9 years ago

@jamesbascle no, you are right, it was a mistake on my part - it does only capture what it think it needs.

But if say, you want to check if c => person.Name == c.Name - it is capturing the class containing the property Name (the person object) because it's really a method of the object (get_Name()) - if we could explicitly get the backing value being the value that is referenced in the expression (i.e captured explicitly via [person.Name]), that's where we would see the reduced footprint. That might not be possible or ideal though. However, if we can just gain better control the lifetime and explicitly capture objects or not, that could help quite a bit. I think the biggest piece we would gain from it is explicitly preventing capture on a self-contained expression via [].

It's a good feature suggestion, and there are quite a few different ways one could go about implementing it. I'm certainly not going to say mine is the best, and based on how the C# language has evolved since I've started learning and even before that, I can say fairly confidently that I trust the design team's decision on how to go about it, if it's something that is ready to be implemented in the next version at all.

dotnet / roslyn

C# Design Notes for Jan 21, 2015 #98

C# Design Meeting Notes for Jan 21, 2015

Agenda

1. Design process

Design team

Feature ideas

Design notes

Owners and proposals

Other ways of increasing openness

Decisions

Prototypes

Agenda

2. Themes

Working with data

Performance and reliability (and interop)

Componentization

Distribution

Metaprogramming

Null

Themeless in Seattle

3. Features

Pattern matching

Records

Array Slices

ref locals and ref returns

readonly parameters and locals

Lambda capture lists

Method contracts