One thing that keeps amazing me about Visual Studio Code is how much it helps you automate annoying tasks when you write code. For example, if you add an img tag to the document you need to know the src and find out the width and height to avoid reflow issues when the image was loaded. You also need to add an alt attribute either to provide an alternative text or to prevent assistive technology to read out the src attribute instead.

You can do this by following these steps:

• Start the img tag
• As soon as you write src=” VS Code gives you an autocomplete for local images.
• Close the img tag (this is important)
• Use the Command Palette” (CMD + Shift + P on Mac or Ctrl + Shift + P on Linux / Windows) and type “size” to trigger the Emmet size script that automatically adds width and height.

You can see it in this screencast:

My setup also warns me when I have images without any alt attribute. It puts a wavy line under any element with a problem like the img tag and shows the error right next to it. I am using two extensions for this: webhint for VS Code and Error Lens.

When you want to save a GIF from Twitter, the interface plays a cruel joke on you. If you context-click the GIF you get a menu that states “Copy GIF address”.

All it does though is copy a link to the tweet with the GIF in it, which is pretty pointless. Under the hood Twitter also converts any GIF to MP4 because it is a much more effective file format.

Knowing this, the easiest way to save the GIF is to open DevTools, go to the Console and type/copy:

window.open(document.querySelector('video').src)

This opens the GIF as a video in a new tab and you can save it there as shown in this screencast:

Of course there are a myriad of web services, Twitter bots and other things that do the same for you and also in more advanced fashion, but I think it is always a very good idea to not trust any service that offers to save protected content from the web for you. I’ve seen far too many malware browser extensions in that arena.

The “expand/shrink” selection keyboard shortcut in Visual Studio Code ⌃⇧⌘→ / ← isn’t the easiest to remember, but ridiculously powerful. It recognises code block boundaries and selects accordingly!

I’m currently upgrading the Console documentation for Microsoft Edge and I needed to create a lot of screenshots for it. I wanted to make sure that they all are the same size, so I fixed a browser window to a certain size and instead of trying to screenshot a part of the screen, I thought it wiser to take full window screenshots. You do this by pressing CMD + Shift + 4, move your mouse on the window you want to take a screenshot of and press Space to save it to your Desktop.

MacOS by default ads a gorgeous drop shadow on any windows and puts some padding around it.

You can get rid of that on the command line or by selecting an even more complex keyboard combination.

However, I also wanted to get rid of the browser Chrome. The final result should be something like this.

So I tried to find a way to crop all the screenshots I have taken in a batch process using Automator. Turns out it can crop, but only from the centre. What I needed was a way to say “crop a rectangle of sx by sy from the image starting at x and y”. The solution was the Swiss army knife of images, ImageMagick. After installing it (via Homebrew), I was able to crop part of an image and create a new one on Terminal using the following syntax (with the bits in {} being the values):

convert sourceimage.png -crop {width}x{height}+{left}+{top} resultimage.png

I created a folder called “nochrome” on my Desktop and wrote this small shell script to batch convert all of them:

for filename in *.png; do\
  convert $filename -crop 1842x918+123+234 "nochrome/$filename"
  echo "$filename done"
done

There are probably easier ways, but that did the trick for me.

Using animations, transitions and smooth scrolling is fun, but they also represent an accessibility problem. Various groups of people have a hard time using your products when things move and change all the time. This is why operating systems have a “reduced motion” setting you can turn on. Our CSS animations should respect these settings and only apply themselves when the user wants to see animations. The best way to achieve this is to wrap them in a prefers-reduced-motion media query. You can use that in various ways as described in this excellent article but they all come with the problem that you need to repeat the settings. There is a simpler way. You can use a custom property:

@media (prefers-reduced-motion: reduce) {
  :root {
    --nomotion: none;
  }
}
html {
  scroll-behavior: var(--nomotion, smooth);
}
button {
  animation: var(--nomotion, rotate 1s infinite alternate);
}

This defines the CSS custom property nomotion as “none” when the user doesn’t want to see any animations. If the user wants to see animations, it isn’t defined and thus the CSS engine applies the fallback, which is your animation settings.

You can see this in action in this CodePen:

See the Pen
Conditional animations with CSS properties by Christian Heilmann (@codepo8)
on CodePen.

You can test it using the reduced motion emulation in the Browser Developer Tools. Here’s a screencast showing this in action: