Blocking trackers is making some websites very slow

still_dreaming_1

If I block trackers or trackers and ads, some websites load very slowly. For example, steelseries.com is very slow. Even after it loads, clicking one of the main navigation links (Mice Keyboards, Mousepads, etc) takes about 4 and a half seconds to load. But if I set it to no blocking, these pages load very quickly, in less than a second.

Shouldn't blocking the trackers and ads make it load faster, not slower?

Catweazle

@still_dreaming_1 , I can't reproduce it at least the page you mention loads in about 1 second with ad and tracking blocker activated. Although it is true that adblockers can slow down especially commercial pages, since they are usually full of crawlers that need to be intercepted, but I at least have no lags on this page.

Ayespy

@still_dreaming_1 I see no slowing from the tracker-blocker. Do you use any other blocking software?

still_dreaming_1

Today is a new day, and my computer has been off all night, and I'm still having this issue on the steelseries website today. I am not using any blocking software. Keep in mind that as soon as I turn off the tracker blocking, the pages load fast for me, and as soon as I turn the blocking back on, they load slowly for me. I am able to reproduce it every time, so it is not an intermittent issue for me, at least on that website.

I am on Windows 10 Home (64 bit) Version 1909. I don't have any extensions installed. I guess I will try a few more things, like removing all my Chrome extensions (from Chrome, just in case they somehow interact?), completely disabling my firewall and virus scanner (I only have the default Windows ones), logging out of my Vivaldi account. Something must be different if it is slow for me and not for the rest of you.

Pathduck

@still_dreaming_1 I can confirm this if enabling the Vivaldi blocker (and disabling uBlock).

I have no idea why it happens though, probably the site has some dependence on some script that is blocked. I notice uBlock does not block all scripts, instead it does some "magic" local redirect stuff for Google Analytics scripts.

Best recommendation is to make an exception for such sites, make a list of them and report it to the Vivaldi team:
https://vivaldi.com/bugreport/
Not sure how much can be done as Vivaldi just relies on public ad/tracker lists but it might help.

still_dreaming_1

Thanks. I will report it with that bugreport link, just in case there is something they can do about it. I tried completely disabling my virus scanner, disabling my firewall, removing all Chrome extensions, deleting all my browser data for all time, logging out of my account within Vivalid (the sync thing), verifying I am not logged into the steelseries website, and rebooting again. It didn't make any difference. The only other thing I can think of trying is reinstalling Vivaldi. It is still strange that @Pathduck is reporting they can reproduce it, but the others said they can't.

Pathduck

So I did some digging, and found the culprits are the requests to Google Analytics.

The site does have a lot of trackers, so if you're concerned about whitelisting the entire site, you can add the following rules in your personal list for trackers:

@@||google-analytics.com/analytics.js$script,domain=steelseries.com
@@||google-analytics.com/gtm/*$script,domain=steelseries.com

rjackdaw

i recommend enabling easy privacy list
https://help.vivaldi.com/article/android-tracker-and-ad-blocker/
https://help.vivaldi.com/article/tracking-and-ad-blocking/

BoneTone

Some sites can become system destroying resource hogs when blocking is enabled. Generally speaking, blocking ads & trackers will reduce page load times and improve general browsing speed (when implemented correctly). The minimal increase in resource use to parse the page and determine what to block is usually more than offset by the savings from not accessing unnecessary resources and not executing unnecessary scripts. This is even true for extensions with notoriously bad implementations.

Some blockers have optimized the way they analyze & determine what to block by taking the resource hit at launch, parsing all the filters, removing not only literal but also functional duplicates, and saving the resultant rule set in its own optimized format. So at page load time, assessing which rules to apply is far more efficient and faster.

A well implemented blocker will never touch the server for blocked ads and trackers, the network request never even goes out on the wire. So these blockers are also reducing the amount of data you use while browsing, the images & videos are never downloaded of course, because your browser never sends the request in the first place.

This approach is probably the best protection one can have against third party tracking. Tracking servers cannot track that which never connects to the server. The best way to achieve that protection is with a default deny all - allow exceptionally policy. Pages load *very* fast with this policy applied, but they are also frequently broken. That's when you create exceptions to the default deny filter, allowing only those connections to resources which are required for the page to work as you need (leaving alone all the third party trackers and ads so they are still never touched).

All of this is great and it leads to a much faster experience browsing the web in close to every case. There are, however, horrendous exceptions. Here's one particularly nasty example. With all third party connections blocked, I accessed the Grammarly website. When a connection to the tracking server fails, it tries the next one (these servers have gibberish names that are numerical labels). That's certainly fine to do, what is unreasonable is to do it with such rapid succession that after a few minutes hundreds of thousands of network requests have been attempted (and diligently blocked by my blocker as they should have been). But this pounding of resource requests can bring even a high powered workstation to a crawl if it's maxing out 8 cores to process them. That evening, Grammarly was properly black listed on every list I maintain. That abuse of a user's machine demonstrates an utter disrespect for the user, while attempting to monetize off the user, and demands a holistic approach to denying them any access to every piece of data possible.

That's an extreme example, but there are certainly less obtrusive yet still detrimental websites out there whose performance will degrade when blocking is applied. In such cases, my recommendation is to seek an alternative source for the information. It's not always possible, but I'l accept a few seconds delay over a privacy breach always.