I've been reading up on the GUI situation for C# on linux and while it seems obvious that there is no support for Forms (or Windows Forms I guess) on Linux, there seems to be a lot of confusion for me regarding the GUI development. For context, I have just recently begun to learn C# and was looking to practice some simple commands while also trying to make a functional GUI with buttons that will execute some commands for me. I'm not exactly sure what next steps to take because any forums / wikis regarding this topic usually go like "Is it possible" while someone on the other end replies "Just use C++ lol".

Would really appreciate any help and especially some beginning point or links for me to go to so I can start learning all of this. Loving C# so far!

--

Réponses : 

• Avalonia - https://github.com/AvaloniaUI/Avalonia
• Eto - https://github.com/picoe/Eto
• Uno - https://platform.uno

Également proposés :

• Linux MAUI <<- version “communautaire” de MAUI, mais pas vraiment maintenu, et carrément pas prêt pour la prod
• Winforms + Mono <<- on oublie
• Razor <<- sérieux ? - pourquoi pas Electron tant qu'on y est, nan parce que quitte à faire de l'ASP.Net, autant faire du web)
• Blazor, mais quand ça sera intégré <<- ça ne l'est pas.
• Electron, “mais je vois pas comment intégrer avec C#” <<- Sérieux ?

You can host a .NET Core MVC app on Linux

On your local machine publish your project by running the following from the project folder:

$ dotnet publish -c Release -o ../_deploy
Create a website folder on the server:

$ mkdir -p /var/www/YOUR-WEBSITE

And copy the contents of ../_deploy from your machine to /var/www/YOUR-WEBSITE/ on the server.

Change the owner of your website’s directory so it would belong to NGINX’s user (www-data):

$ chown -R www-data:www-data /var/www/

systemd service

Create a systemd config for Kestrel instance:

$ nano /etc/systemd/system/kestrel-YOUR-WEBSITE.service

Edit it like that:

[Unit]
Description=YOUR-WEBSITE

[Service]
WorkingDirectory=/var/www/YOUR-WEBSITE/
ExecStart=/usr/bin/dotnet /var/www/YOUR-WEBSITE/YOUR-WEBSITE.dll
Restart=always
RestartSec=10
SyslogIdentifier=dotnet-YOUR-WEBSITE
User=www-data
Environment=ASPNETCORE_ENVIRONMENT=Production

[Install]
WantedBy=multi-user.target

Enable and start it:

$ systemctl enable kestrel-YOUR-WEBSITE.service
$ systemctl start kestrel-YOUR-WEBSITE.service

You can check its status:

$ systemctl status kestrel-YOUR-WEBSITE.service

If everything is okay, then it will show something like that:

● kestrel-YOUR-WEBSITE.service - YOUR-WEBSITE
   Loaded: loaded (/etc/systemd/system/kestrel-YOUR-WEBSITE.service; enabled; vendor preset: enabled)
   Active: active (running) since Thu 2017-08-10 11:30:09 UTC; 1s ago
 Main PID: 15628 (dotnet)
    Tasks: 14
   Memory: 25.4M
      CPU: 1.380s
   CGroup: /system.slice/kestrel-YOUR-WEBSITE.service
           └─15628 /usr/bin/dotnet /var/www/YOUR-WEBSITE/YOUR-WEBSITE.dll

NGINX

Install NGINX:

$ apt-get install nginx

Now, there are 2 options how NGINX can send requests to Kestrel: via TCP or via Unix socket.

TCP

Your Program.cs:

// ...

public static IWebHostBuilder CreateWebHostBuilder(string[] args) =>
    WebHost.CreateDefaultBuilder(args)
        .UseUrls("http://localhost:5000/")
        .UseStartup<Startup>();

// ...

NGINX config (/etc/nginx/sites-available/default):

# ...

server {
    listen 80;
    listen [::]:80;

    location / {
            proxy_pass http://localhost:5000;
            proxy_http_version 1.1;
            proxy_set_header Upgrade $http_upgrade;
            proxy_set_header Connection keep-alive;
            proxy_set_header Host $host;
            proxy_cache_bypass $http_upgrade;
            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
            proxy_set_header X-Forwarded-Proto $scheme;
    }
}

# ...

Unix socket

This should be better in terms of performance as there is no TCP overhead. Even if there was no performance impact, on a local machine “talking” through Unix socket simply makes more sense than communicating over TCP.

Your Program.cs:

// ...

public static IWebHostBuilder CreateWebHostBuilder(string[] args) =>
    WebHost.CreateDefaultBuilder(args)
        .UseLibuv()
        .ConfigureKestrel(
            (context, serverOptions) =>
            {
                // /var/www/YOUR-WEBSITE
                string root = Path.GetDirectoryName(
                    System.Reflection.Assembly.GetExecutingAssembly().Location
                );
                // if we are (behind NGINX) and on Linux, then can use sockets
                if (RuntimeInformation.IsOSPlatform(OSPlatform.Linux))
                {
                    // path to socket has to be absolute
                    string socket = Path.Combine(root, "kestrel.sock");
                    serverOptions.ListenUnixSocket(socket);
                }
                else
                {
                    serverOptions.Listen(IPAddress.Loopback, 5000);
                }
            }
        )
        .UseStartup<Startup>();

// ...

For that to work you need to add Transport.Libuv package, so you’ll have the following in your .csproj:

<PackageReference Include="Microsoft.AspNetCore.Server.Kestrel.Transport.Libuv" Version="2.2.0" />

NGINX config in this case:

# ...

location / {
    proxy_pass http://unix:/var/www/YOUR-WEBSITE/kestrel.sock;
    proxy_http_version 1.1;
    proxy_set_header Upgrade $http_upgrade;
    proxy_set_header Connection keep-alive;
    proxy_set_header Host $host;
    proxy_cache_bypass $http_upgrade;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header X-Forwarded-Proto $scheme;
}

# ...

Save it and restart NGINX:

$ systemctl restart nginx.service

Now open your web-browser and go to http://YOUR-SERVER-IP/. It fucking works!

This is the source code for the password generator page. The prounceable password generator is quite long in its initialisation arrays, so is a download. The alpha-numeric password generator is very simple and doesn't do anything complicated besides generate a random MD5 hash and truncate the output.