Closed Bug 816297 Opened 13 years ago Closed 13 years ago

Gaia always initiates update poll early in startup

Categories

(Firefox OS Graveyard :: Gaia::System, defect, P3)

defect

Tracking

(blocking-basecamp:+)

VERIFIED FIXED
blocking-basecamp +

People

(Reporter: cjones, Assigned: etienne)

Details

(Keywords: perf, regression)

Attachments

(1 file)

(I thought I had filed this, but apparently not.) STR (1) Reboot b2g phone Note that during startup, we force an update check, and then start the check immediately when we go online. We absolutely don't want to do this during early startup. It's a waste of CPU cycles on the critical path to showing UI. There's a reason why we have the first-check delay pref! ;) The unpleasantness of this is also exacerbated by bug 813765. Either marshall or etienne suggested that the FTU code someone caused us to screw up and always force an update check during early startup, even after first run. We need to fix that.
Component: Gaia → Gaia::System
Tested a bit more, gaia does set |'gaia.system.checkForUpdates': true| only when we close the FTU, not on subsequent reboots. Marshall do you know if we do a "oh hey! a connection, let's check for updates" on the platform side?
Flags: needinfo?(marshall)
(In reply to Etienne Segonzac (:etienne) from comment #1) > Marshall do you know if we do a "oh hey! a connection, let's check for > updates" on the platform side? Only if a previous check failed because the network was offline. Something needs to force the initial check =/
Flags: needinfo?(marshall)
I always see this in logcat on early startup I/GeckoDump( 106): XXX FIXME : Got a mozContentEvent: force-update-check Gaia is going to be guilty until proven innocent here, sorry :). Especially since this seemed to have regressed around the time of FTU.
Flags: needinfo?(etienne)
(In reply to Chris Jones [:cjones] [:warhammer] from comment #3) > I always see this in logcat on early startup > > I/GeckoDump( 106): XXX FIXME : Got a mozContentEvent: force-update-check > > Gaia is going to be guilty until proven innocent here, sorry :). Especially > since this seemed to have regressed around the time of FTU. Eheh :) Found it, I was innocently looking for code settings the force-check setting to true, while it was true from the get go... Patch incoming.
Flags: needinfo?(etienne)
Assignee: nobody → etienne
Attachment #686738 - Flags: review?(marshall)
Julien, if you want to steal the review I'm sure Marshall won't mind :)
blocking-basecamp: ? → +
Priority: -- → P3
Attachment #686738 - Flags: review?(marshall) → review?(felash)
Comment on attachment 686738 [details] Pointer to gaia PR 6728 r=me
Attachment #686738 - Flags: review?(felash) → review+
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → FIXED
Reviewed and verified on "Unagi" Build ID:201330103070201 "Check for update" no longer offered on the Start-Up menu
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: