Using Gertrude

Unblocking Guide

Getting the websites and apps you need unblocked is one of the main tasks you'll need to perform when working with Gertrude. This guide aims to help you understand how and why this works the way it does, and give you some helpful tools for making this job easier.

Why multiple keys are needed

One of the core things you need to understand is that when you're trying to unblock a specific website, it often takes several keys to get the site fully working. It's not intuitive why—it seems like if I'm trying to unblock happyfish.com, I should be able to just create a key for happyfish.com, and everything should just work, right? Sadly, no. To explain why, I'll need to get just a little bit technical for a moment.

How websites work

Websites are made up of computer code written in 3 different languages: html, css, and javascript. When you try to visit happyfish.com in a web browser, what happens under the hood is that your computer makes first makes a network request to happyfish.com in order to get the html code for that site. Inside the html code are more web addresses, containing the location of the css and javascript code. So once it receives the html, the browser reads it, and then makes more network requests to get the rest of the code it needs to properly display the website.

Very often, the css and javascript code is located on a different domain from the one you're requesting. So, in our example, the html for happyfish.com will be located at happyfish.com, but the css might be located at static.happycdn.com, and the javascript might be at code.javascript-libraries.com.

Terminology: domain

A domain is the root part of a web address, before any forward slashes. For instance, the web address https://cupcakes.com/halloween-sale has the domain cupcakes.com.

When Gertrude is protecting a computer, it has to make a decision about whether to block or allow every new request to a new domain. So, if you create a key for just happyfish.com, Gertrude will allow the first request which fetches the html code. But when the browser reads that html and realizes it needs to go to static.happycdn.com to get the css, that request will be blocked by Gertrude. Sadly, there's no way for Gertrude's filter to know that the request for static.happycdn.com is being made because it is needed by a website that you already unlocked.

Fixing partially unblocked websites

So, for our example, imagine you created unblocking happyfish.com. If you tried to visit the site after creating this key, you'd probably notice the site loads, but doesn't look right. That's because web browsers try to do the best with what they've got. They don't show you an error message if they can't load some of the code they need, they just show a partially working, or funny looking website.

To make this more concrete, consider the website how-to-type.com, which we use as an example in our getting started tutorial. If we only create a key for the main how-to-type.com domain, the website loads, but looks a little funny, as shown below:

The site is loading, but if you look closely, all is not right...
The site is loading, but if you look closely, all is not right...

If we pull up the Gertrude network request window, we'll still see that multiple requests are being blocked when we try to visit the page:

After unblocking the domain, we still see all these blocked requests
After unblocking the domain, we still see all these blocked requests

There is a bit of an art to figuring out which of these requests should be unlocked, but you'll get the knack of it. If you look closely at the image above, several of the domains are repeated twice, so there are really fewer domains to think about than it appears. Also, there are a couple things we can ignore. The blocked request from the www.facebook.com domain can (and should) be ignored—it's just a little embedded "like" button that won't prevent the site from working correctly. And the one from google-analytics.com is just a tracking website, its not necessary to unlock.

Ignoring facebook and google analytics, there are 3 domains to be unblocked
Ignoring facebook and google analytics, there are 3 domains to be unblocked

You can create an unlock request for all three blocks at once:

You can create an unlock request for all three requests at once
You can create an unlock request for all three requests at once

Once these three domains are unlocked, the website loads!

Success! With all the keys created, the site loads fully
Success! With all the keys created, the site loads fully

This sounds hard!

This may seem daunting to you, and if that's the case, it's OK. The difficulty of unblocking websites is one of the reasons that most other safety solutions throw their hands up and only block parts of the internet, instead of blocking everything and requiring you to unlock specific sites. That approach, while easy, is inherently unsafe. By choosing to use Gertrude, you are signing up for more work than those solutions, but the payoff is immense: real internet safety. Not only is it really worth the effort, but there are several things that make this process easier:

  1. We have a growing list of publicly available keychains which have all the keys necessary for popular websites. Often, you'll be able to just select a public keychain and the site will immediately work correctly.
  2. The ability to send unlock requests from the macOS app makes this process much easier.
  3. As long as you don't unlock domains like google.com or facebook.com accidentally, there's usually very little danger in unlocking a few of the wrong domains while you're hunting for which domain or domains still needs to be unblocked for a site.
  4. You and your user/s will get good at this with a little practice.
  5. Once you have a site fully unlocked, with Gertrude's keychains feature you can easily share the full set of keys with all of the users you are protecting—you won't have to repeat the process again.
  6. You can always reach out to us if you're having trouble getting a site unlocked.
  7. We're exploring some ways to in the future try to automatically discover and unlock some or all of these other supporting domains, so it's very likely this process will keep getting easier over time.

Isolate network activity

Your computer is constantly making network requests to a myriad of websites and services. The vast majority of these don't cause you any noticeable problem, and don't need to be unblocked. But, all those requests can make it hard to find out which request DOES need to be unblocked. For this reason, probably the number one thing you can do to make it easier to find what to unblock is isolate network activity:

  • Close every other app except the one you're working with. This will greatly reduce the number of network requests you'll have to consider.
  • Close other tabs if you're working with a web browser, this also substantially eliminates noise and makes it easier to find the right request.
  • Be sure to clear and filter requests. Enter a search phrase into the filter field of Gertrude's network activity screen. Usually part of the name of a website, or the app you're using are helpful. Click the clear requests button and then refresh the page, or retry the activity that is not working.
Filter by a search phrase, and clear often
Filter by a search phrase, and clear often
Gertrude

The Gertrude app helps you protect your kids online with strict internet filtering that you can manage from your own computer or phone, plus remote monitoring of screenshots and keystrokes. $5/mo, with a 60 day free trial.

Start free trial →