751813 - Stop spewing to the console about Places Maintenance

Status: RESOLVED FIXED

(Toolkit :: Places, defect)

Description

[Phil Ringnalda (:philor)]

According to bug 739354 comment 6, we want to promote the antipattern of polling the console and assuming that nothing but your test will be writing to it, so we need to stop putting anything in there about Places Maintenance, so that tests which do that can fail when other things get written in there instead.

Comment 1

[Marco Bonardo [:mak]]

Yeah, I'm fine with that, as I expressed in https://bugzilla.mozilla.org/show_bug.cgi?id=739354#c5, I actually forgot that bug was still open.

Comment 2

[Marco Bonardo [:mak]]

I suppose it'll take less time if I directly do that.

Comment 3

[Marco Bonardo [:mak]]

Attachment: patch v1.0

Comment 4

[:Paolo Amadini]

Comment on attachment 626813 [details] [diff] [review]
patch v1.0

>+++ b/toolkit/components/places/PlacesCategoriesStarter.js
>@@ -79,17 +79,20 @@ PlacesCategoriesStarter.prototype = {
>-          PlacesDBUtils.maintenanceOnIdle();
>+          PlacesDBUtils.maintenanceOnIdle(function () {
>+            // Empty callback used to to eat the log and thus avoid sending it
>+            // to the console.
>+          });

It looks like this is the only place where maintenanceOnIdle is called, apart from test code:

http://mxr.mozilla.org/mozilla-central/search?string=maintenanceOnIdle

What do you think about removing the "else" part inside maintenanceOnIdle that writes to the console when the callback is null, and just using a callback that writes to the console (or do_log_info or something) in tests that really require it? I doubt that any add-on depends on the behavior of writing to the console when the callback is null.

Comment 5

[:Paolo Amadini]

(In reply to Paolo Amadini [:paolo] from comment #4)
> What do you think about removing the "else" part inside maintenanceOnIdle

I meant _executeTasks:

http://mxr.mozilla.org/comm-central/source/mozilla/toolkit/components/places/PlacesDBUtils.jsm#69

I cannot find any caller of checkAndFixDatabase, also.

Comment 6

[Marco Bonardo [:mak]]

(In reply to Paolo Amadini [:paolo] from comment #4)
> What do you think about removing the "else" part inside maintenanceOnIdle
> that writes to the console when the callback is null, and just using a
> callback that writes to the console

Hm yeah that may also be a possibility, will check add-ons on mxr.

Comment 7

[Marco Bonardo [:mak]]

Attachment: patch v1.1

ok, the change makes sense, no interesting results from add-ons mxr.

Comment 8

[:Paolo Amadini]

Comment on attachment 629788 [details] [diff] [review]
patch v1.1

Looks good!

Comment 9

[Marco Bonardo [:mak]]

looks like I never pushed this, will do.

Comment 10

[Marco Bonardo [:mak]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/7530e96d9af6

Comment 11

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/7530e96d9af6