RFP: provide [more plausible] hardwareConcurrency per OS
Categories
(Core :: DOM: Core & HTML, enhancement)
Tracking
()
People
(Reporter: thorin, Unassigned)
References
(Blocks 1 open bug)
Details
This would be nice (and simple to change the value per OS) to get in time for ESR140
Notes:
- Bug 1360039: original RFP implementation in FF55
- Bug 1630089: macOS reports physical cores instead of logical
- Bug 1728741: capped at 16
- https://data.firefox.com/dashboard/hardware
433.5% (trending up)231.5% (trending down)
- Brave: randomizes
- with I assume a range of values [2, 4, 6, 8, 12, etc] - I can find the code later if required
- I don't think there's any real compat issues with reporting a higher value that what the user actually has (see Brave)
Anecdata
8Apple M1 iMac (system report:4performance4efficiency)- JS also returns
8so IDK re Bug 1630089 (IANAE)
- JS also returns
10Galaxy S24 FE (new)8Galaxy A30 (4+ years old)12my windows 11 desktop machine
currently RFP hardwareConcurrency is spoofed the same for everyone at 2. It would be nice if we reported more plausible results in line with real world, per OS - e.g. macs and phones reporting 2 seems super low and totally not real world (red flags for anti-bot scripts). IANAE on hardware.
Could we get a summary from telemetry (e.g. Bug 1879151) of hardwareConcurrency by OS to make some informed decisions on this
| Reporter | ||
Updated•10 months ago
|
| Reporter | ||
Updated•10 months ago
|
| Reporter | ||
Updated•10 months ago
|
Comment 1•10 months ago
|
||
(In reply to Thorin [:thorin] from comment #0)
Could we get a summary from telemetry (e.g. Bug 1879151) of hardwareConcurrency by OS to make some informed decisions on this
Yes; although we don't have mobile data yet. I will leave the ni for me, and we expect to be doing something with this soonish as well - I think we can align.
| Reporter | ||
Comment 2•5 months ago
|
||
FYI: https://firefoxgraphics.github.io/telemetry/#view=system
- This doesn't break it down by platform, but
2is only 8% and I would expect e.g. for mac, this would not be a benign value at all - also see Bug 1984333
| Reporter | ||
Comment 3•5 months ago
|
||
RFP changes were included in Bug 1984333 (FF143+, pending beta uplift)
| Reporter | ||
Updated•5 months ago
|
Description
•