Firefox/Content Performance Program: Difference between revisions

Firefox/Content Performance Program (view source)

Revision as of 15:15, 3 August 2015

1,059 bytes added , 3 August 2015

no edit summary

ErinL ancaster

canmove, Confirmed users, Bureaucrats and Sysops emeriti

5,401

edits

@@ Line 16: / Line 16: @@
   | Mailing list || [https://mail.mozilla.org/listinfo/contentperf contentperf@mozilla.org] || devs, PMs
 |}
+== First Priorities (Q3 2015) ==
+;Evaluate top sites (Aaron, Avi, Vladan): For the initial set of sites (Twitter, Facebook, Yahoo search) + Google properties + ~5-10 other top-sites we choose ourselves from Alexa top 20, do the following:<br>
+*Agree on & write up the "most common" use-cases (i.e. user interactions) with these sites
+*Evaluate relative performance (wrt other browsers) for each of these uses-cases
+*Profile & determine why our performance is worse
+*This evaluation will need to be repeated for Fennec & mobile versions of sites, but with a smaller subset of sites
+;Gfx impact on scrolling (Avi): Re-do Aaron's scrolling measurement (using the scrolling bookmarklet) on the reference HP laptop with the 3 reference sites (Yahoo, Twitter, Facebook) using every relevant gfx configuration setting (see my blog post for list + consider Windows theme impact).
+;Windows 10 (Aaron): Compare content-perf on Windows 10 vs Windows 8
+;Figure out how to make dynamic websites reproducible (Wander)
+;New page-load test (Wander): Create a new cross-browser page-load test that measures how long it takes for the browser to become idle after loading a page (time-until-quiescent). This test could be a webpage
 == First Priorities (Q2 2015) ==
@@ Line 37: / Line 48: @@
 Again, focusing on the Windows 8.1 laptop/tablet and a particular Android phone.
-=== Implementation Strategy ===
-====Identify programs already in-flight (60fps program) and what problems those are designed to solve (elan)====
-*60 FPS:
 ====Existing Benchmarks and Measures====