How to access the UKM data

[uazo]

Hi all,

I was wondering whether there is public access to the data generated by Chrome's UKM and whether there is a public guide that can be followed.
From what I understand they are available via BigQuery but I don't really know how to do it.

My goal is the verification of all PostMessage_Incoming_*_DifferentBucket_SameOrigin.

Thank you for your excellent work.

[Jeremy Roman]

Chrome's UKM data is very sensitive and even most Chrome engineers don't have access to it (requiring additional approval). As far as I know, there is no general process for external query access. (Someone who knows more can correct me if I'm wrong.)

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to the Google Groups "Chromium-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-dev...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/0100e179-5397-47de-ba17-adb1a3643bfbn%40chromium.org.

[Christian Dullweber]

There is an aggregated, public report based on UKM data: https://developer.chrome.com/docs/crux/

I'm not sure if it answers your specific question though. 

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CACuR13eoRU_4CiLftdJmFjXjafsRS6JicNunOR6pf_VnTvoojQ%40mail.gmail.com.

[uazo]

> Chrome's UKM data is very sensitive and even most Chrome engineers don't have access to it

a pity, but I don't understand the difference with the UseConters of blink, because those are "full" public and it seems to me that they are conveyed through the UKM.

perhaps because the browser side is specifically "less" public than blink, in the sense closer to chrome's features?

what is the process that decides whether a log should be UKM or UseCounter?

because, from https://chromium.googlesource.com/chromium/src/+/lkgr/docs/use_counter_wiki.md 

      > "UseCounter data is essential for web compat decision making and the blink process for breaking changes"

and it seems to me that specifically those UKMs I wrote can impact on how blink works.

is it possible to request a switch from UKM to UseCounter, considering the implementation is blink side?

> There is an aggregated, public report based on UKM data

thank you, I was aware of that, but I couldn't find any information on ukm, and now I understand why.

thanks for the answers

[K. Moon]

I believe the publicly-available UseCounter data is based on Web crawling top sites, and not on UKM metrics reported by end user browsers.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/1ffc2725-b4af-42c4-839c-b0fe9d497ec3n%40chromium.org.

[Alexei Svitkine]

To clarify, the use counter data on chromestatus.com (e.g. https://chromestatus.com/metrics/css/popularity) is derived from data collected from UMA (not UKM).

So it is from real world usage. It is not the full UMA data set however, but a specific subset related to use counters.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CACwGi-74aoiTmwz2QU1Fo0TL%2BsDotgByj4%3DOFumx%3DJbbstCJrw%40mail.gmail.com.

[uazo]

> that they are conveyed through the UKM.

Sorry, I double-checked the source.
the opposite is true, i.e. that a subset of usecounter becomes a ukm.
forgive me, but the chromium base code is really vast and sometimes I struggle to remember exactly what I read, especially if some time passes.