mirrorweb / pywb Goto Github PK

cc @themasonbanks

Example URLs
https://webarchive.nationalarchives.gov.uk/ukgwa/20031020010435/http://www.nationalarchives.gov.uk/news/stories/9.htm
https://tnaqa.mirrorweb.com/ukgwa/20211004151828/https://www.counterterrorism.police.uk/latest-news/page/20/

When trying to navigate to these pages we are served with an error like this - "The web page at https://tnaqa.mirrorweb.com/ukgwa/20211004151810mp_/https://www.counterterrorism.police.uk/latest-news/page/20/ might be temporarily down or it may have moved permanently to a new web address."

When viewing this page we are served with a PYWB error (there are multiple examples of this error throughout)
https://tnaqa.mirrorweb.com/ukgwa/20211004151800/https://www.police.uk/

When bypassing Cloudfront these pages load fine.

Describe the bug

When testing in a local pywb environment, the site is shown to work consistently, with the help of the developed content scripts. But inconsistent behaviour is shown when a crawl is done accessed. Sometimes the instance will show a blank screen, then when refreshed or the link is re-opened again, the site loads up. Links that were previously working may then decide not to work. Console.log's from my code are present in the dev tools, when locally running the site in pywb. But when looking for the same console.logs in the new marrionette instances, they are not present. Older instances show no effect and seem to be working, as expected. To note, only the most recent crawls seem to be affected. Possibly a cdx/indexing issue?

Steps to reproduce the bug

Access the most recent instance in marionette:

1. Open the archive to the latest instance
2. Navigate to News and Statements, via the News and Campaigns tab in the main navbar
3. Access any pages on the page.
4. Some pages will appear blank but accessible when refreshed. Others will load in as expected.

Expected behaviour

When a link is clicked the corresponding page is open.

Screenshots

Environment

• OS: macOs Monterey version 12.6 (21G115)
• Browser: Google Chrome
• Version: Version 108.0.5359.124 (Official Build) (arm64)

Additional context

Clients Profile Link: https://app.mirrorweb.com/management/web/16cac3a5-4135-4750-b8fe-657a3faa7a7e/profile/aec04c0d-4c98-4eb1-a491-2e10f306c99c/
Jira ticket: https://mirrorweb.atlassian.net/browse/CS-1326

Expected behavior

The page, when directed too, should replay as normal with all images being present and interactivity working as well.

What actually happened

The site is continually refreshing when the site is interactied with; when the site page is scrolled or any image in the image carousell or the dropdown menu is clicked. It makes site navigation impossible. When in the dev tools, and disable the javascript option, the refreshng stops. This suggests a javascript problem but at this point the solution has not been identiafiable.

Steps to replicate:

1. Go to 'https://webarchive.nationalarchives.gov.uk/ukgwa/20220601094243/https://stfccareers.co.uk/'
2. Click on the drop-down menu options, or scroll the page
3. The page will refresh, continously.

specs of tech used:
macbook m1 2021
macos os monterrey version 12.5.1
google chrome version: 105.0.5195.125 (Official Build) (arm64)