"Reject All" is not leveraged, leading to inefficient operation

[cherti]

When visiting certain sites, such as https://www.theguardian.com/, the Cookie acceptance system provides a "Reject All" option (that is also required by law if there is an "accept all", so it is both likely to be and factually quite prevalent). This option is, however, not leveraged by Consent-O-Matic, instead, according to observation, it seems to manually go through all options to tick. This is highly inefficient, leading to processing times for the cookie banner above the 10 second mark (about 13 seconds in particular), and even more on other websites that provide such a reject-all alternative.

If no accepted trackers are listed for a particular site, shortcutting processing via the "Reject All" option seems to be highly advantageous and would substantially improve Consent-O-Matic's processing times.

[dkgof]

There are several reasons to why we don't just use "Reject All" options when the specific options are available.

1. ConsentOMatic is ment to throw light on the problem that is the GDPR popups and how they function, this is why the latest version makes it very visual with all the clicking, in the PictureInPicture window. Most sites should let you scroll/read while this is happening though.

2. "Reject All" most of the time, does actually not reject legitimate interrests, so to make sure they are rejected aswell, we go through every option.

If you want to make ConsentOMatic work faster, then you can reenable the old behaviour of hiding the consent popup while handling it, which also removes the delay between each click, since it is not visible anyways. This makes most popups get handled almost instantly. If you go to the hidden debug tab, in settings and enable "hideInsteadOfPip".

Hope that helps a little.

[dkgof]

I can see that theguardian actually prevent scrolling intentionally while the popup is open, maybee we can look into fixing this.

[cherti]

Thanks, this does indeed solve the mentioned issue. I didn't think that reject all was so… liberally implemented, but with the immediate handling the problem in fact does not exist. Shedding light onto the issue is certainly a good point, though.

Shall I keep this open for the guardian-thing or shall I close because the actual question was solved?

[ShadowJonathan]

I had a feeling CoM was doing all options instead of reject-all, and it seems that feeling has been validated.