783050 - tweak a few more histogram definitions

Status: RESOLVED FIXED

(Toolkit :: Telemetry, defect)

Description

[Nathan Froyd [:froydnj]]

STARTUP_MEASUREMENT_ERRORS should be a HISTOGRAM_ENUMERATED_VALUES.

The cache histograms should use 10000 in their definition rather than 10 * 1000.

The first one is a consistency fix.  The second one makes it just slightly simpler to parse things for automated conversion of TelemetryHistograms.h to JSON.

Comment 1

[Nathan Froyd [:froydnj]]

Attachment: patch

Comment 2

[(dormant account)]

Comment on attachment 652170 [details] [diff] [review]
patch

i think the current conversion script already handles calculated values

Comment 3

[Nathan Froyd [:froydnj]]

(In reply to Taras Glek (:taras) from comment #2)
> i think the current conversion script already handles calculated values

That script makes my eyes bleed.  My current conversion script doesn't require everything to be using the HISTOGRAM macro; things are much happier if HISTOGRAM_ENUMERATED_VALUES and similar are used, so that the JSON file reflects intent and isn't just a conversion of the postprocessed data.

Doing things that way will require some work to handle earlier versions for Metrics's consumption, but I don't think it will be that difficult.

Comment 4

[(dormant account)]

(In reply to Nathan Froyd (:froydnj) from comment #3)
> (In reply to Taras Glek (:taras) from comment #2)
> > i think the current conversion script already handles calculated values
> 
> That script makes my eyes bleed.  My current conversion script doesn't
> require everything to be using the HISTOGRAM macro; things are much happier
> if HISTOGRAM_ENUMERATED_VALUES and similar are used, so that the JSON file
> reflects intent and isn't just a conversion of the postprocessed data.
> 
> Doing things that way will require some work to handle earlier versions for
> Metrics's consumption, but I don't think it will be that difficult.

Ok, can you post the WIP script on github or a user-repo, or attach to some bug? Do we have JSON to give to Harsha after these changes?

Comment 5

[Nathan Froyd [:froydnj]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/c90be658d59d

Decided to just let the computed histogram (10 * 1000) lie, since there were a lot like it and it didn't seem worth the churn to fix them all.

(In reply to Taras Glek (:taras) from comment #4)
> Ok, can you post the WIP script on github or a user-repo, or attach to some
> bug? Do we have JSON to give to Harsha after these changes?

Yeah, I'll stick it in bug 781531 by the end of the week.

Comment 6

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/c90be658d59d