In this keynote talk, David will be covering the past, present, and future of Visual F# as best as we can predict. :)

Past and Present of Visual F#:

• F# 4.0: What's new? A brief overview of F#'s evolution.

• .NET Core: What is it? Cross-platform. Open source. Modular. Modernized.

• Awesome things going on in the F# community

Future of Visual F#:

• The Microsoft team's short-term roadmap and guiding principles

• Areas where people can contribute

Machine learning and functional programming are two exciting programming trends of the recent years. To a newcomer, both of these may sound very complicated. If you attend this workshop, you'll see that this could not be further from the truth. We will look at solving real-world machine learning problems with the functional-first programming language F# and you'll see that this is a perfect match - functional programming makes it easy to express machine learning algorithms in a clear and easy to understand way.

In this workshop, we'll implement a machine learning algorithm for automatically detecting the language of text, using pages downloaded from Wikipedia (using an F# web crawler) as a training data set. Along the way you'll learn about machine learning concepts, the F# language and the functional approach to data processing.

What you can expect to learn:

• How to use F# to implement machine learning algorithms
• Writing simple code using functional style
• The core ideas behind artificial neural networks

No prior machine learning or F# knowledge is needed.

C# is a great language for developing Line of Business applications but F# is even better! If you want to write code that expresses your requirements clearly, ensures correctness and supports rapid and fun development then guess what, F# does all that and more! In this session we will explore how the F# type system helps us to create domain models that we can use in any .Net application. No prior experience of F# is necessary for this session but be warned, once you've seen what F# can do, you won't want to go back to C#!

Handling errors, asynchronous computations, and hiding the behind-the-scenes plumbing are just some of the use cases of computation expressions. They can be a little bit tricky to learn, specially at the start.

In this hands-on workshop we'll learn how to use some common computation expressions with varying levels of difficulty and then learn how to roll our own computation expressions from scratch.

Nerd Dinner is the canonical example of a MVC application and is very familiar to web devs who want to learn MVC the Microsoft way. For everything that Nerd Dinner is, it is not terribly smart, for example, there are no business rules outside of some basic input validation. The tutorial is going to take the Nerd Dinner forms over data and inject some analytics into the application. We will make the UI much more intelligent about who is using it and we will start predicting attendance of upcoming dinners. The session will use JavaScript, C# and F#. Basic statistics will be used.

You can see Nerd Dinner here and the walk-through here.

IMPORTANT

You will need to have Visual Studio in order to attend this workshop.

More info coming soon!

One of the myths about F# and functional programming is that it is only good for complex mathematics. This could not be further from truth. A web server is essentially a function from a request to a response and so functional programming is a perfect fit for it.

In this tutorial, you'll learn all you need to get started with building web applications in F# using the Suave.io server. We'll start with an walkthrough that introduces the basic F# concepts and constructs and uses them to build a simple interactive web site. By the end of the workshop, you'll write an application that accesses multiple external data sources asynchronously, presents data through a web page and can be easily deployed to Azure or Heroku.

You will learn:

· Domain modelling with F# - how can we model what data structures we're working with

· How to use F# type providers to access JSON-based REST services as well as XML and CSV files

· How to write web applications using Suave.io and how to deploy them to Heroku and Azure

Requirements:

· No prior knowledge of F# or Suave.io is required. You also don't need too much .NET knowledge, so people coming from other communities are more than welcome!

· You'll need a Mac, Windows or Linux laptop with F# installed. See www.fsharp.org for installation instructions.

Cloud computing services offer unlimited and on-demand computational power including large amounts of data storage. The combination of built in support for asynchronous workflows, event combinators and data processing make F# uniquely suited to develop scalable cloud solutions efficiently. In this session we will build and deploy a service “in the cloud”, more specifically in Azure, leveraging its infrastructure for distributed computations across clusters of machines using Microsoft Orleans and the “Orleankka” library. Moreover, we will develop a simple mobile application to interact with the API of the deployed service.

Microsoft Orleans is a framework for .NET platform built by the Microsoft Research to simplify building and deploying cloud services that can scale to thousands of servers and millions of users. Orleankka is a powerful library which provides a set of API for supporting Microsoft Orleans framework in F#.

Private Eye is a profiler designed to be scripted in F# interactive. This give developers low level control over profiling data. This gives developers greater flexibility in how they investigate performance data. It not only allows developers to visualize the performance data the way they want to, but also to use all available date, unlike most commercially available which often throw away metrics on all but the most called methods/allocations.

In this tutorial you'll the basics of Private Eye. You'll learn how to show text tables of the data gathered in the F# console. You'll also learn how to use FSharp.Charting to visualize the data. Finally you'll see how to use F#'s agent/mailboxes to aggregated data to make it easier to understand.

Find out more about Private Eye

Concurrent/distributed programming is a brave new world that opens new challenges and opportunities. The introduction of abstractions such as MapReduce and LINQ has made it easier to tame cloud resources, but they can often be restrictive in the types of problems that they solve. A standout exception are the powerful asynchronous workflows -F#'s most significant intellectual export- which enable the articulation of asynchronous, concurrent and parallel code just as if it were sequential.

MBrace is an open source project that extends asynchronous workflows to the cloud setting. MBrace can leverage commodity cloud computing services, allowing instant deployment of distributed code to hundreds of CPU cores straight from Visual Studio.

In this tutorial we will be offering a brief introduction to multicore programming using async, progressing to distributed code using MBrace on Azure. Examples of computations will be covered, progressing from the sequential, on to multicore, to the distributed context. In the process we will be exposing the differing qualities inherent in multicore and distributed code. This will be a hands-on tutorial, so an available Azure account will be required to work on the examples.

http://mbrace.io/