Created attachment 8841166 [details] Web opening time performance comparison (Firefox Vs Chrome) 1. Platform Version (M): NOS(7.0) 2. Firefox version (M): Firefox 51.0.4 3. Reproducible on LG Ref. device (M) : G5/G6/Nexus device 4. Reproduce rate (M) : 100% 5. Issue Reported (M) : TD 6. Issue Description (Wrong behavior) (M) : Web opening time(Web loading time) is slow compared to Chrome browser. 7. Expected behavior (M) : Web opening time should not be slow compared to Chrome Browser. 8. Steps to reproduce (M) : 1) Connect to http://m.daum.net via Firefox browser 2) Firefox menu -> Bookmarks -> long click "Daum" -> Add to Home Screen 3) Click the home button -> Click the "Daum" in home screen 4) Firefox menu -> Settings -> Clear private data -> CLEAR DATA 5) Click the home button 6) Prepare a stop watch and measure from the time you click on the "Daum" on the home screen to the end of loading the "Daum" page. 9. Screen Capture (M) : Yes(attached video) 10. MORE INFORMATION (O): When it fixed, please share Engineering Report or Release info. for patches.
Created attachment 8841390 [details] Firefox's Web Opening Time is slow.png The results are measured by DIVA, and firefox is always inferior to chrome, although there may be differences depending on the network.
With those sorts of test need to startup Gecko each time. As I understand Android keeps parts of Blink in memory thus does not need to pay the startup costs each time.
The page opening time is slower than Chrome as below sites. www.yahoo.com mobile.nytimes.com
Starting in Firefox 51 we extract libraries to the cache directory on the first one. By clearing the data before each run, this step is repeated every time, which is not what would happen with normal usage. I recommend starting Firefox once without loading a page after clearing the data, then repeating the rest of your test.
You can use the 'Clear private data' button in the Firefox settings to clear the cache, if that's what you're trying to do.
You need to log in before you can comment on or make changes to this bug.