Scott Hanselman

In a world where JSON (JavaScript Object Notation) is everywhere it's long been somewhat frustrating that .NET didn't have built-in JSON support. JSON.NET is great and has served us well but it's remained a 3rd party dependency for basic stuff like an ASP.NET web site or a simple console app.

Back in 2018 plans were announced to move JSON into .NET Core 3.0 as an intrinsic supported feature, and while they're at it, get double the performance or more with Span<T> support and no memory allocations. ASP.NET in .NET Core 3.0 removes the JSON.NET dependency but still allows you to add it back in a single line if you'd like.

NOTE: This is all automatic and built in with .NET Core 3.0, but if you’re targeting .NET Standard or .NET Framework. Install the System.Text.Json NuGet package (make sure to include previews and install version 4.6.0-preview6.19303.8 or higher). In order to get the integration with ASP.NET Core, you must target .NET Core 3.0.

It's very clean as well. Here's a simple example.

using System;
using System.Text.Json;
using System.Text.Json.Serialization;

namespace verysmall
{
    class WeatherForecast
    {
        public DateTimeOffset Date { get; set; }
        public int TemperatureC { get; set; }
        public string Summary { get; set; }
    }
    
    class Program
    {
        static void Main(string[] args)
        {
            var w = new WeatherForecast() { Date = DateTime.Now, TemperatureC = 30, Summary = "Hot" };
            Console.WriteLine(JsonSerializer.Serialize<WeatherForecast>(w));
        }
    }
}

The default options result in minified JSON as well.

{"Date":"2019-07-27T00:58:17.9478427-07:00","TemperatureC":30,"Summary":"Hot"}      

Of course, when you're returning JSON from a Controller in ASP.NET it's all automatic and with .NET Core 3.0 it'll automatically use the new System.Text.Json unless you override it.

Here's an example where we pull out some fake Weather data (5 randomly created reports) and return the array.

[HttpGet]
public IEnumerable<WeatherForecast> Get()
{
    var rng = new Random();
    return Enumerable.Range(1, 5).Select(index => new WeatherForecast
    {
        Date = DateTime.Now.AddDays(index),
        TemperatureC = rng.Next(-20, 55),
        Summary = Summaries[rng.Next(Summaries.Length)]
    })
    .ToArray();
}

The application/json is used and JSON is returned by default. If the return type was just string, we'd get text/plain. Check out this YouTube video to learn more details about System.Text.Json works and how it was designed. I'm looking forward to working with it more!

Sponsor: Get the latest JetBrains Rider with WinForms designer, Edit & Continue, and an IL (Intermediate Language) viewer. Preliminary C# 8.0 support, rename refactoring for F#-defined symbols across your entire solution, and Custom Themes are all included.

About Scott

Scott Hanselman is a former professor, former Chief Architect in finance, now speaker, consultant, father, diabetic, and Microsoft employee. He is a failed stand-up comic, a cornrower, and a book author.

About   Newsletter

Hosting By

I'm a huge fan of .NET Core global tools. I've done a podcast on Global Tools. Just like Node and other platform have globally tools that can be easily and quickly installed and then used in build scripts, CI/CD (Continuous Integration/Continuous Deployment) systems, or just general command line utilities, .NET Global Tools are easily made (by you!) and distributed via NuGet.

Some cool examples (and there are hundreds) are the "Try .NET" Workshop runner and creator that can you can use to make interactive documentation, or coverlet for code coverage. There's a great and growing list of .NET Core Global Tools on GitHub.

If you've got the .NET SDK installed you can try out a global tool just like this.

dotnet tool install -g dotnetsay

Then run this example with "dotnetsay," it's fun.

stepping back a moment, you may be familiar with PowerShell. It's a scripting language and a command line shell like Bash or DOS or the Windows Command Prompt. You may think of PowerShell as a tool for maintaining and managing Windows Servers.

However in recent years, PowerShell has gone cross platform and runs most anywhere. It's lightweight and has .NET Core at its, ahem, core. You can use PowerShell for scripting systems on any platform and if you're a .NET developer the team has made installing and immediately using PowerShell in scripts a one liner - which is genius. It's PowerShell as a .NET Global Tool.

Here's an example output from my system running Ubuntu. I just "dotnet tool install --global PowerShell."

$ dotnet --version
2.1.502  
$ dotnet tool install --global PowerShell
You can invoke the tool using the following command:  pwsh
Tool 'powershell' (version '6.2.2') was successfully installed.
$ pwsh
PowerShell 6.1.1
https://aka.ms/pscore6-docs
Type 'help' to get help.      
PS /mnt/c/Users/Scott/Desktop>
exit                                                                                                 

Here I've checked that I have .NET 2.x or above, then I install PowerShell. I can run scripts or I can drop into the interactive shell. Note the PS prompt and my current directory above.

In fact, PowerShell is so useful as a scripting language when combined with .NET Core that PowerShell has been included as a global tool within the .NET Core 3.0 Preview Docker images since Preview 4. This means you can use PowerShell lines/scripts inside Docker images.

FROM mcr.microsoft.com/dotnet/core/sdk:3.0
RUN pwsh -c Get-Date
RUN pwsh -c "Get-Module -ListAvailable | Select-Object -Property Name, Path"

Being able to easily install PowerShell as a global tool means you can count on it in your scripts, CI/CDs systems, or docker containers. It's also nice to be able to be able to use existing PowerShell scripts cross platform.

I'm impressed with this idea - installing PowerShell itself as a .NET Global Tool. Very clever and useful.

Sponsor: Ossum unifies agile planning, version control, and continuous integration into a smart platform that saves 3x the time and effort so your team can focus on building their next great product. Sign up free.

About Scott

Scott Hanselman is a former professor, former Chief Architect in finance, now speaker, consultant, father, diabetic, and Microsoft employee. He is a failed stand-up comic, a cornrower, and a book author.

About   Newsletter

Hosting By

There's some interesting stuff quietly happening in the "Console App" world within open source .NET Core right now. Within the https://github.com/dotnet/command-line-api repository are three packages:

• System.CommandLine.Experimental
• System.CommandLine.DragonFruit
• System.CommandLine.Rendering

These are interesting experiments and directions that are exploring how to make Console apps easier to write, more compelling, and more useful.

The one I am the most infatuated with is DragonFruit.

Historically Console apps in classic C look like this:

#include <stdio.h>
 
int main(int argc, char *argv[])
{
    printf("Hello, World!\n");
    return 0;
}

That first argument argc is the count of the number of arguments you've passed in, and argv is an array of pointers to 'strings,' essentially. The actual parsing of the command line arguments and the semantic meaning of the args you've decided on are totally on you.

C# has done it this way, since always.

static void Main(string[] args)
{
    Console.WriteLine("Hello World!");
}

It's a pretty straight conceptual port from C to C#, right? It's an array of strings. Argc is gone because you can just args.Length.

If you want to make an app that does a bunch of different stuff, you've got a lot of string parsing before you get to DO the actual stuff you're app is supposed to do. In my experience, a simple console app with real proper command line arg validation can end up with half the code parsing crap and half doing stuff.

myapp.com someCommand --param:value --verbose

The larger question - one that DragonFruit tries to answer - is why doesn't .NET do the boring stuff for you in an easy and idiomatic way?

From their docs, what if you could declare a strongly-typed Main method? This was the question that led to the creation of the experimental app model called "DragonFruit", which allows you to create an entry point with multiple parameters of various types and using default values, like this:

static void Main(int intOption = 42, bool boolOption = false, FileInfo fileOption = null)
{
    Console.WriteLine($"The value of intOption is: {intOption}");
    Console.WriteLine($"The value of boolOption is: {boolOption}");
    Console.WriteLine($"The value of fileOption is: {fileOption?.FullName ?? "null"}");
}

In this concept, the Main method - the entry point - is an interface that can be used to infer options and apply defaults.

using System;

namespace DragonFruit
{
    class Program
    {
        /// <summary>
        /// DragonFruit simple example program
        /// </summary>
        /// <param name="verbose">Show verbose output</param>
        /// <param name="flavor">Which flavor to use</param>
        /// <param name="count">How many smoothies?</param>
        static int Main(
            bool verbose,
            string flavor = "chocolate",
            int count = 1)
        {
            if (verbose)
            {
                Console.WriteLine("Running in verbose mode");
            }
            Console.WriteLine($"Creating {count} banana {(count == 1 ? "smoothie" : "smoothies")} with {flavor}");
            return 0;
        }
    }
}

I can run it like this:

> dotnet run --flavor Vanilla --count 3   
Creating 3 banana smoothies with Vanilla

The way DragonFruit does this is super clever. During the build process, DragonFruit changes this public strongly typed Main to a private (so it's not seen from the outside - .NET won't consider it an entry point. It's then replaced with a Main like this, but you'll never see it as it's in the compiled/generated artifact.

public static async Task<int> Main(string[] args)
{
    return await CommandLine.ExecuteAssemblyAsync(typeof(AutoGeneratedProgram).Assembly, args, "");
}

So DragonFruit has swapped your Main for its smarter Main and the magic happens! You'll even get free auto-generated help!

DragonFruit:
  DragonFruit simple example program

Usage:
  DragonFruit [options]

Options:
  --verbose            Show verbose output
  --flavor <flavor>    Which flavor to use
  --count <count>      How many smoothies?
  --version            Display version information

If you want less magic and more power, you can use the same APIs DragonFruit uses to make very sophisticated behaviors. Check out the Wiki and Repository for more and perhaps get involved in this open source project!

I really like this idea and I'd love to see it taken further! Have you used DragonFruit on a project? Or are you using another command line argument parser?

Sponsor: Ossum unifies agile planning, version control, and continuous integration into a smart platform that saves 3x the time and effort so your team can focus on building their next great product. Sign up free.

About Scott

Scott Hanselman is a former professor, former Chief Architect in finance, now speaker, consultant, father, diabetic, and Microsoft employee. He is a failed stand-up comic, a cornrower, and a book author.

About   Newsletter

Hosting By

As I've mentioned lately, I'm quietly moving my Website from a physical machine to a number of Cloud Services hosted in Azure. This is an attempt to not just modernize the system - no reason to change things just to change them - but to take advantage of a number of benefits that a straight web host sometimes doesn't have. I want to have multiple microsites (the main page, the podcast, the blog, etc) with regular backups, CI/CD pipeline (check in code, go straight to staging), production swaps, a global CDN for content, etc.

I'm breaking a single machine into a series of small sites BUT I want to still maintain ALL my existing URLs (for good or bad) and the most important one is hanselman.com/blog/ that I now want to point to hanselmanblog.azurewebsites.net.

That means that the Azure Front Door will be receiving all the traffic - it's the Front Door! - and then forward it on to the Azure Web App. That means:

• hanselman.com/blog/foo -> hanselmanblog.azurewebsites.net/foo
• hanselman.com/blog/bar -> hanselmanblog.azurewebsites.net/foo
• hanselman.com/blog/foo/bar/baz -> hanselmanblog.azurewebsites.net/foo/bar/baz

There's a few things to consider when dealing with reverse proxies like this and I've written about that in detail in this article on Dealing with Application Base URLs and Razor link generation while hosting ASP.NET web apps behind Reverse Proxies.

You can and should read in detail about Azure Front Door here.

It's worth considering a few things. Front Door MAY be overkill for what I'm doing because I have a small, modest site. Right now I've got several backends, but they aren't yet globally distributed. If I had a system with lots of regions and lots of App Services all over the world AND a lot of static content, Front Door would be a perfect fit. Right now I have just a few App Services (Backends in this context) and I'm using Front Door primarily to manage the hanselman.com top level domain and manage traffic with URL routing.

On the plus side, that might mean Azure Front Door was exactly what I needed, it was super easy to set up Front Door as there's a visual Front Door Designer. It was less than 20 minutes to get it all routed, and SSL certs too just a few hours more. You can see below that I associated staging.hanselman.com with two Backend Pools. This UI in the Azure Portal is (IMHO) far easier than the Azure Application Gateway. Additionally, Front Door is Global while App Gateway is Regional. If you were a massive global site, you might put Azure Front Door in ahem, front, and Azure App Gateway behind it, regionally.

Again, a little overkill as my Pools are pools are pools of one, but it gives me room to grow. I could easily balance traffic globally in the future.

CONFUSION: In the past with my little startup I've used Azure Traffic Manager to route traffic to several App Services hosted all over the global. When I heard of Front Door I was confused, but it seems like Traffic Manager is mostly global DNS load balancing for any network traffic, while Front Door is Layer 7 load balancing for HTTP traffic, and uses a variety of reasons to route traffic. Azure Front Door also can act as a CDN and cache all your content as well. There's lots of detail on Front Door's routing architecture details and traffic routing methods. Azure Front Door is definitely the most sophisticated and comprehensive system for fronting all my traffic. I'm still learning what's the right size app for it and I'm not sure a blog is the ideal example app.

Here's how I set up /blog to hit one Backend Pool. I have it accepting both HTTP and HTTPS. Originally I had a few extra Front Door rules, one for HTTP, one for HTTPs, and I set the HTTP one to redirect to HTTPS. However, Front door charges 3 cents an hour for the each of the first 5 routing rules (then about a penny an hour for each after 5) but I don't (personally) think I should pay for what I consider "best practice" rules. That means, forcing HTTPS (an internet standard, these days) as well as URL canonicalization with a trailing slash after paths. That means /blog should 301 to /blog/ etc. These are simple prescriptive things that everyone should be doing. If I was putting a legacy app behind a Front Door, then this power and flexibility in path control would be a boon that I'd be happy to pay for. But in these cases I may be able to have that redirection work done lower down in the app itself and save money every month. I'll update this post if the pricing changes.

After I set up Azure Front Door I noticed my staging blog was getting hit every few seconds, all day forever. I realized there are some health checks but since there's 80+ Azure Front Door locations and they are all checking the health of my app, it was adding up to a lot of traffic. For a large app, you need these health checks to make sure traffic fails over and you really know if you app is healthy. For my blog, less so.

There's a few ways to tell Front Door to chill. First, I don't need Azure Front Door doing a GET requests on /. I can instead ask it to check something lighter weight. With ASP.NET 2.2 it's as easy as adding HealthChecks. It's much easier, less traffic, and you can make the health check as comprehensive as you want.

app.UseHealthChecks("/healthcheck");

Next I turned the Interval WAY app so it wouldn't bug me every few seconds.

These two small changes made a huge difference in my traffic as I didn't have so much extra "pinging."

After setting up Azure Front Door, I also turned on Custom Domain HTTPs and pointing staging to it. It was very easy to set up and was included in the cost.

I haven't decided if I want to set up Front Door's caching or not, but it might mean an easier, more central way than using a CDN manually and changing the URLs for my sites static content and images. In fact, the POP (Point of Presense) locations for Front Door are the same as those for Azure CDN.

NOTE: I will have to at some point manage the Apex/Naked domain issue where hanselman.com and www.hanselman.com both resolve to my website. It seems this can be handled by either CNAME flattening or DNS chasing and I need to check with my DNS provider to see if this is supported. I suspect I can do it with an ALIAS record. Barring that, Azure also offers a Azure DNS hosting service.

There is another option I haven't explored yet called Azure Application Gateway that I may test out and see if it's cheaper for what I need. I primarily need SSL cert management and URL routing.

I'm continuing to explore as I build out this migration plan. Let me know your thoughts in the comments.

Sponsor: Develop Xamarin applications without difficulty with the latest JetBrains Rider: Xcode integration, JetBrains Xamarin SDK, and manage the required SDKs for Android development, all right from the IDE. Get it today

About Scott

Scott Hanselman is a former professor, former Chief Architect in finance, now speaker, consultant, father, diabetic, and Microsoft employee. He is a failed stand-up comic, a cornrower, and a book author.

About   Newsletter

Hosting By

I'm quietly moving my Website from a physical machine to a number of Cloud Services hosted in Azure. This is an attempt to not just modernize the system - no reason to change things just to change them - but to take advantage of a number of benefits that a straight web host sometimes doesn't have. I want to have multiple microsites (the main page, the podcast, the blog, etc) with regular backups, CI/CD pipeline (check in code, go straight to staging), production swaps, a global CDN for content, etc.

I'm also moving from an ASP.NET 4 (was ASP.NET 2 until recently) site to ASP.NET Core 2.x LTS and changing my URL structure. I am aiming to save money but I'm not doing this as a "spend basically nothing" project. Yes, I could convert my site to a static HTML generated blog using any number of great static site generators, or even a Headless CMS. Yes I could host it in Azure Storage fronted by a CMS, or even as a series of Azure Functions. But I have 17 years of content in DasBlog, I like DasBlog, and it's being actively updated to .NET Core and it's a fun app. I also have custom Razor sites in the form of my podcast site and they work great with a great workflow. I want to find a balance of cost effectiveness, features, ease of use, and reliability.  What I have now is a sinking feeling like my site is gonna die tomorrow and I'm not ready to deal with it. So, there you go.

Currently my sites live on a real machine with real folders and it's fronted by IIS on a Windows Server. There's an app (an IIS Application, to be clear) leaving at \ so that means hanselman.com/ hits / which is likely c:\inetpub\wwwroot full stop.

For historical reasons, when you hit hanselman.com/blog/ you're hitting the /blog IIS Application which could be at d:\whatever but may be at c:\inetpub\wwwroot\blog or even at c:\blog. Who knows. The Application and ASP.NET within it knows that the site is at hanselman.com/blog.

That's important, since I may write a URL like ~/about when writing code. If I'm in the hanselman.com/blog app, then ~/about means hanselman.com/blog/about. If I write /about, that means hanselman.com/about. So the ~ is a shorthand for "starting at this App's base URL." This is great and useful and makes Link generation super easy, but it only works if your app knows what it's server-side base URL is.

To be clear, we are talking about the reality of the generated URL that's sent to and from the browser, not about any physical reality on the disk or server or app.

I've moved my world to three Azure App Services called hanselminutes, hanselman, and hanselmanblog. They have names like http://hanselman.azurewebsites.net for example.

ASIDE: You'll note that hitting hanselman.azurewebsites.com will hit an app that looks stopped. I don't want that site to serve traffic from there, I want it to be served from http://hanselman.com, right? Specifically only from Azure Front Door which I'll talk about in another post soon. So I'll use the Access Restrictions and Software Based Networking in Azure to deny all traffic to that site, except traffic from Azure - in this case, from the Azure Front Door Reverse Proxy I'll be using.

That looks like this in this Access Restrictions part of the Azure Portal.

Since the hanselman.com app will point to hanselman.azurewebsites.net (or one of its staging slots) there's no issue with URL generation. If I say / I mean /, the root of the site. If I generate a URL like "~/about" I'll get hanselman.com/about, right?

But with http://hanselmanblog.azurewebsites.net it's different.

I want hanselman.com/blog/ to point to hanselmanblog.azurewebsites.net.

That means that the Azure Front Door will be receiving traffic, then forward it on to the Azure Web App. That means:

• hanselman.com/blog/foo -> hanselmanblog.azurewebsites.net/foo
• hanselman.com/blog/bar -> hanselmanblog.azurewebsites.net/foo
• hanselman.com/blog/foo/bar/baz -> hanselmanblog.azurewebsites.net/foo/bar/baz

There's a few things to consider when dealing with reverse proxies like this.

Is part of the /path being removed or is a path being added?

In the case of DasBlog, we have a configuration setting so that the app knows where it LOOKS like it is, from the Browser URL's perspective.

My blog is at /blog so I add that in some middleware in my Startup.cs. Certainly YOU don't need to have this in config - do whatever works for you as long as context.Request.PathBase is set as the app should see it. I set this very early in my pipeline.

That if statement is there because most folks don't install their blog at /blog, so it doesn't add the middleware.

//if you've configured it at /blog or /whatever, set that pathbase so ~ will generate correctly
Uri rootUri = new Uri(dasBlogSettings.SiteConfiguration.Root);
string path = rootUri.AbsolutePath;

//Deal with path base and proxies that change the request path
if (path != "/")
{
    app.Use((context, next) =>
    {
        context.Request.PathBase = new PathString(path);
        return next.Invoke();
    });
}

Sometimes you want the OPPOSITE of this. That would mean that I wanted, perhaps hanselman.com to point to hanselman.azurewebsites.net/blog/. In that case I'd do this in my Startup.cs's ConfigureServices:

app.UsePathBase("/blog");

Be aware that If you're hosting ASP.NET Core apps behind Nginx or Apache or really anything, you'll also want ASP.NET Core to respect  X-Forwarded-For and other X-Forwarded standard headers. You'll also likely want the app to refuse to speak to anyone who isn't a certain list of proxies or configured URLs.

I configure these in Startup.cs's ConfigureServices from a semicolon delimited list in my config, but you can do this in a number of ways.

services.Configure<ForwardedHeadersOptions>(options =>
{
    options.ForwardedHeaders = ForwardedHeaders.All;
    options.AllowedHosts = Configuration.GetValue<string>("AllowedHosts")?.Split(';').ToList<string>();
});

Since Azure Front Door adds these headers as it forwards traffic, from my app's point of view it "just works" once I've added that above and then this in Configure()

app.UseForwardedHeaders();

There seems to be some confusion on hosting behind a reverse proxy in a few GitHub Issues. I'd like to see my scenario ( /foo -> / ) be a single line of code, as we see that the other scenario ( / -> /foo ) is a single line.

Have you had any issues with URL generation when hosting your Apps behind a reverse proxy?

---

Sponsor: Develop Xamarin applications without difficulty with the latest JetBrains Rider: Xcode integration, JetBrains Xamarin SDK, and manage the required SDKs for Android development, all right from the IDE. Get it today

About Scott

Scott Hanselman is a former professor, former Chief Architect in finance, now speaker, consultant, father, diabetic, and Microsoft employee. He is a failed stand-up comic, a cornrower, and a book author.

About   Newsletter

Hosting By

Comment on this post [4]

Share on BlueSky or use the Permalink and post anywhere!