492953 - gloda activity manager processes have incorrect / incomplete context for account - shows null

Status: NEW

(Thunderbird :: Search, defect)

Description

[Andrew Sutherland [:asuth] (he/him)]

currently they are like so:
  process.contextType = "account";
  process.contextObj  = aFolder;

We are missing a contextDisplayText, but also, Standard8 proposes we use the incomingServer as the context.  Since aFolder is (potentially confusingly) a string in the above case, the gloda indexer listener interface would want to be augmented to take the nsIMsgFolder in question (or null) instead.

example of using the incomingServer context setup:
http://mxr.mozilla.org/comm-central/source/mail/components/activity/modules/alertHook.js#66

Comment 2

[Magnus Melin [:mkmelin]]

I'm not sure if this is the bug for it, but I miss having both server + folder in the activity manager operations. "Inbox is up to date" etc. really doesn't say anything when you have several accounts.

Comment 4

[Jens Müller (:tessarakt)]

Target Milestone should be resetted, 3.0 is out ...

Comment 5

[:aceman]

In addition to the contextDisplayText missing, the aFolder argument seems to be null as if the gloda service is not sending the folder name. The result is that there is a context title above the progress bar just showing 'null', which is bad. So this is no longer an enhancement but a visible bug.

I could update the listener to take a real nsIMsgFolder, but why is gloda not sending anything? Is there a problem with _indexingFolder at
http://mxr.mozilla.org/comm-central/source/mailnews/db/gloda/modules/indexer.js#707

Comment 6

[Andrew Sutherland [:asuth] (he/him)]

Maybe GlodaMsgIndexer._indexerEnterFolder should be calling GlodaIndexer._notifyListeners() around the area after where it logs what folder it is entering.  If we did this, then _worker_folderIndex will no longer need to call GlodaIndexer._notifyListeners.

Comment 11

[arthur.lutz]

getting this too.

Comment 12

[Helder]

Same "null" header is present on TB 45.2.0.

Comment 13

[kobaz]

Same null header is present in TB 52.0b3

Comment 14

[Sylwester Zarębski]

I confirm the same 'null' in TB 52.0.1.

Comment 15

[stephan]

Same 'null' header is present in TB 60.6.1