Project Fission/iframe-crawler: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
(tweak)
No edit summary
 
Line 1: Line 1:
To help gather data on real Web content that might be helpful to the Fission teams, jwatt has a [https://github.com/jwatt/iframe-crawler simple Web crawler] (see [https://jwatt.org/iframe-crawler/display-crawl-results.html some crawl results]).
To help gather data on real Web content that might be helpful to the Fission teams, jwatt has a Web crawler.
 
* Code: https://github.com/jwatt/iframe-crawler
* Sample results: https://jwatt.org/iframe-crawler/display-crawl-results.html


== Ideas and requests ==
== Ideas and requests ==

Latest revision as of 19:22, 30 May 2019

To help gather data on real Web content that might be helpful to the Fission teams, jwatt has a Web crawler.

Ideas and requests

Outstanding

  • Do a crawl with Fission turned off, and a crawl with it turned on, taking screenshots of OOP-iframes to be compared to look for Fission rendering bugs.
  • What is the max depth of cross-origin subdocs (the amount of nested processes we'd need basically)?
  • How often are cross-origin iframes under 'transforms' or 'filters'?

Answered

  • Percentage of pages with visible cross-origin subdocs? Around 25% based on initial crawl.
  • Frequency of 'display: none' and/or zero sized iframes.