1950301 - shamir.com - Missing content from the page

Status: ASSIGNED

(Web Compatibility :: Site Reports, defect, P2)

Description

[Raul Bucata]

Environment:
Operating system: Windows 11
Firefox version: Firefox 135.0.1 (64-bit)

Steps to reproduce:

1. Navigate to: https://shamir.com/my/lenses_and_more/shamir-autograph-intelligence/
2. Observe the section

Expected Behavior:
Sections are populated with text and images

Actual Behavior:
Sections are blank

Notes:

• Reproduces regardless of the status of ETP
• Reproduces in firefox-nightly, and firefox-release
• Does not reproduce in chrome

Created from https://github.com/webcompat/web-bugs/issues/148749

Comment 1

[Raul Bucata]

Attachment: Chr vs FF

Comment 2

[Alice0775 White]

It works if add-ons Chrome Mask is enabled.

Comment 3

[Thomas Wisniewski [:twisniewski]]

This page behaves rather oddly. It doesn't show me its content on Chrome either after a page-refresh, until I do some sort of user-interaction (click, scroll, press a key). Until then, the content is hidden with opacity:0:

<div class="col-lg-6 text-left" data-aos-duration="800" data-delay="100" data-aos="fade-up" data-aos-anchor-placement="top-bottom">

It seems to be shown only after rocket-DOMContentLoaded is fired, but that event never seems to fire in Firefox for whatever reason:

            o.fn.ready = o.fn.init.prototype.ready = function (e) {
              return t.domReadyFired ? e.bind(document) (o) : document.addEventListener('rocket-DOMContentLoaded', (() => e.bind(document) (o))),
              o([])
            },

Playing around a bit, if I have the string Firefox in the UA string in RDM while Chrome Mask is running, then it breaks again, which lead be to this code in their RocketLazyLoadScripts stuff:

                if (
                  navigator.userAgent.indexOf('Firefox/') > 0 ||
                  '' === navigator.vendor ||
                  this.CSPIssue
                )

And indeed, if I make an intervention which changes Firefox to FireFox in the UA string, and sets navigator.vendor to Mozilla (or any value), it seems to consistently work, so this is likely the code which is causing the issue somehow. But I can't quite grok why, so for now I'll just ship that intervention, and we should probably reach out to them and confirm if this is just a problem with their site specifically, or the Rocket lazy loader itself.

Comment 4

[Thomas Wisniewski [:twisniewski]]

Attachment: Bug 1950301 - add a JS UA intervention for shamir.com; r?denschub

Comment 5

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/9d889bf0cdd4
add a JS UA intervention for shamir.com; r=denschub,webcompat-reviewers

Comment 6

[Serban Stanca [:SerbanS]]

https://hg.mozilla.org/mozilla-central/rev/9d889bf0cdd4

Comment 7

[Daniel Holbert [:dholbert]]

Removing the needs-diagnosis keyword -- it looks like this is already largely diagnosed as being several forms of UA-sniffing, and we've managed to sitepatch it, and we've got it flagged as needs-contact to see if the site can fix/remove their UA-specific codepaths. I don't think there's further diagnosis that'd make sense to do at this point.