732310 - A11Y_INSTANTIATED telemetry is rather useless

Status: RESOLVED DUPLICATE of bug 723846

(Core :: Disability Access APIs, defect)

Description

[alexander :surkov (:asurkov)]

Because it shows nothing. Ignoring that boolean can be false this telemetry item makes me think that accessibility is always instantiated, in every case.

Comment 1

[alexander :surkov (:asurkov)]

the same is applicable to A11Y_XXX_USAGE items. 

I'd say we need to collect this data on shutdown whether a11y was instantiated for example but per SHUTDOWN_OK telemetry we don't shutdown properly in 10% of case so results won't be correct. That means we need to the same thing as SHUTDOWN_OK does, i.e. collect data and send them on next startup.

Who's willing to take?

Comment 2

[David Bolter [:davidb] (NeedInfo me for attention)]

I'm not convinced shutdown is the right place. Regarding A11Y_INSTANTIATED, I think we need to see the total number of telemetry pings for the same period, that way we'll know the percentage of users that instantiate a11y (who also enable telemetry).

For the USAGE stuff, I think we should only report whether it was used even once (e.g. bug 721772) or not. Then we can do the same calculation above. Incrementing a counter on each usage foils that calculation.

Ideally we should also find out how to turn # of telemetry pings into # users (estimates).

Taras, does this make sense?

Comment 3

[Nathan Froyd [:froydnj]]

I'd like to see this fixed with bug 723846 instead.

Comment 4

[alexander :surkov (:asurkov)]

(In reply to David Bolter [:davidb] from comment #2)
> I'm not convinced shutdown is the right place. 

I said we need to telemetry on startup not shutdown.

> Regarding A11Y_INSTANTIATED,
> I think we need to see the total number of telemetry pings for the same
> period, that way we'll know the percentage of users that instantiate a11y
> (who also enable telemetry).

this sounds like workaround of the problem

> Ideally we should also find out how to turn # of telemetry pings into #
> users (estimates).

Also doesn't perfect since the user can change his environment. Anyway this bug is about different things.

Comment 5

[alexander :surkov (:asurkov)]

(In reply to Nathan Froyd (:froydnj) from comment #3)
> I'd like to see this fixed with bug 723846 instead.

ok