Skip to content

"Same website on IPv6 and IPv4" test should not be required #485

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
peterjin-org opened this issue Dec 7, 2020 · 8 comments · Fixed by #996
Closed

"Same website on IPv6 and IPv4" test should not be required #485

peterjin-org opened this issue Dec 7, 2020 · 8 comments · Fixed by #996

Comments

@peterjin-org
Copy link

Currently, internet.nl gives an error if the IPv6 and IPv4 versions of a website are different. However, this is not useful for an IPv6-only website whose "IPv4 version" just says something like "You need to enable IPv6 to view this website" and has no actual content (e.g. earthite.com). Instead, this subtest should only be made a warning, in order to still catch IPv6-related misconfigurations, but not to disqualify websites that may have a legitimate reason to have different content on IPv6 and IPv4.

@baknu baknu added this to the v1.7 milestone Oct 21, 2022
@baknu baknu modified the milestones: v1.7, v1.8 Oct 21, 2022
@mxsasha
Copy link
Collaborator

mxsasha commented Jun 8, 2023

Easy test site: ip.tyk.nu

@baknu as part of this, I have added the use of detail web ipv6 web-ipv46 verdict info and we no longer use detail web ipv6 web-ipv46 verdict bad. Text can probably mostly be copied, maybe tweaked. Explanation may also need to be updated. This needs to be done in the content repo.

@baknu
Copy link
Contributor

baknu commented Jun 9, 2023

So this means that the website test gets a bit more relaxed (or less strict) in general. Although I believe it affects just a few websites, I can see the point. I would suggest to give a "warning" and not an "informational" instead of the current "bad", because in most cases the difference will not be intended and should be fixed by the website owner. Note that both a "warning" and an "informational" do not impact the %-score. Might be something to discuss with the steering group.

@mxsasha
Copy link
Collaborator

mxsasha commented Jun 13, 2023

Adjustments based on discussion:

    1. if HTTP (port 80) and/or HTTPS (port 443) over IPv4 are also available over IPv6;
    2. if HTTP headers (such as a redirect header) over IPv4 are also received over IPv6;
    3. if HTML content over IPv4 is the same over IPv6. After stripping eventual nonces we do a comparison of the recieved content. The content difference must not be higher than 10% to pass this subtest. The treshold makes sure that small differences (for example due to changing ads) don not lead to a fail of this subtest as well.

1 and 2 should be FAIL (❌), 3 should be NOTICE (⚠️)

@mxsasha mxsasha self-assigned this Jun 30, 2023
mxsasha added a commit that referenced this issue Jul 25, 2023
@mxsasha
Copy link
Collaborator

mxsasha commented Jul 25, 2023

The new behaviour is now implemented in #1039. We need to make content for the new label detail web ipv6 web-ipv46 verdict notice which is used for scenario 3 from the above list.

mxsasha added a commit that referenced this issue Jul 28, 2023
@mxsasha mxsasha removed their assignment Jul 28, 2023
@baknu
Copy link
Contributor

baknu commented Oct 26, 2023

@mxsasha Have you added this text label somehwere in the content repo? Or shall I add it to the content branch "version-1.8" under detail/web/ipv6/web-ipv46/verdict?

@baknu baknu added the question label Oct 26, 2023
@mxsasha
Copy link
Collaborator

mxsasha commented Oct 26, 2023

@mxsasha Have you added this text label somehwere in the content repo? Or shall I add it to the content branch "version-1.8" under detail/web/ipv6/web-ipv46/verdict?

I did not create it in branch yet, so yes, please create it.

@baknu
Copy link
Contributor

baknu commented Oct 26, 2023

@mxsasha Ok, done. Also edited the test explanation and the "bad" verdicts. Could you review? Thanks.

@mxsasha
Copy link
Collaborator

mxsasha commented Oct 26, 2023

@mxsasha Ok, done. Also edited the test explanation and the "bad" verdicts. Could you review? Thanks.

Looks good!

@baknu baknu closed this as completed Oct 27, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Development

Successfully merging a pull request may close this issue.

3 participants