Open Bug 1667248 Opened 4 years ago Updated 3 years ago

Fenix loads cloudflare.com 5 seconds slower than Chrome

Categories

(Core :: Performance, defect, P3)

defect

Tracking

()

Performance Impact low
Tracking Status
firefox83 --- affected

People

(Reporter: yoasif, Unassigned)

References

(Depends on 1 open bug)

Details

(Keywords: perf:pageload)

Basic information

Saw a video shared on social media with a test between Brave and Fenix: https://vimeo.com/461247541

Recreated the test with Chrome vs. Fenix on a Moto G XT1031 with Warp enabled.

cloudflare.com

chrome - 3.5 seconds
firefox - 8.6 seconds


More information

Profile URL: https://share.firefox.dev/33S7s7m

Basic systems configuration:

OS version: Android 5.1

GPU model: Adreno 305

Number of cores: Quad-core 1.2 GHz Cortex-A7

Amount of memory (RAM): 1GB

Thanks so much for your help.

Summary: Fenix loads cloudflare.com 4 seconds slower than Chrome → Fenix loads cloudflare.com 3 seconds slower than Chrome
Summary: Fenix loads cloudflare.com 3 seconds slower than Chrome → Fenix loads cloudflare.com 5 seconds slower than Chrome

(In reply to Asif Youssuff from comment #0)

chrome - 3.5 seconds
firefox - 8.6 seconds

How reliable is this performance difference for you? (Is Chrome always this good, and is Firefox always this bad?)

(dthayer and I both tried this locally, and we're seeing 1-3 second load times from both Firefox and Chrome.)

Flags: needinfo?(yoasif)

Daniel, there is some weirdness with this page - sometimes Fenix is better than the worst case, and Chrome is worse than its best case. I also noticed that Chrome seems to be employing tricks - their progress bar disappears, but if I immediately tap the main menu, I see that the refresh button has not yet become a refresh button - it is still a X (stop) symbol.

Perhaps this is a UI trick - but even if so, it makes Fenix look slow even if there is nothing happening on screen. My measurements were done based on the browser progress bar, not based on content appearing or not moving around on screen. Given that mobiles are small screened devices, I really have no idea of what is happening after the first screen, so I am relying on the progress bar to give me status on whether the page has finished loading.

Would you like HAR exports from both browsers, or does the above give you any sort of hint? I also tried a LG Optimus G - faster device, 5Ghz WiFi (N) which shows the same variance (with Fenix being "slower"). My Pixel 2 basically shows no difference between the two.

Flags: needinfo?(yoasif)

(In reply to Asif Youssuff from comment #2)

I also noticed that Chrome seems to be employing tricks - their progress bar disappears, but if I immediately tap the main menu, I see that the refresh button has not yet become a refresh button - it is still a X (stop) symbol.

This is some known shenanigans & is tracked in bug 1454477.

Would you like HAR exports from both browsers, or does the above give you any sort of hint?

Thank you for offering! I think "no" (though maybe someone on the performance team will disagree & ask for that).

Given the variance (Firefox sometimes fast, Chrome sometimes slow), and the fact that the progress-bar comparison isn't entirely apples-to-apples here due to bug 1454477, I'm going to triage this as qf:p3 (our bucket for most bugs), and as a dependency of bug 1454477 for now.

I don't think it makes sense for anyone to dive in here too much until we've decided what we want to do about bug 1454477.

Depends on: 1454477
Whiteboard: [qf:p3:pageload]
Severity: -- → S3
Performance Impact: --- → P3
Keywords: perf:pageload
Whiteboard: [qf:p3:pageload]
Priority: -- → P3
You need to log in before you can comment on or make changes to this bug.