891788 - The Sms compose tests are slow

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::SMS, defect)

Description

[Julien Wajsberg [:julienw]]

this is due to the fact that input listeners are not cleared at each test run.

Comment 1

[Julien Wajsberg [:julienw]]

Attachment: patch v1

see also PR https://github.com/mozilla-b2g/gaia/pull/10892

Comment 2

[Julien Wajsberg [:julienw]]

This is a change only for tests, but since the change is part of the build and we'd like this to go to the branch, I'm asking leo+ on this. This should have no impact for the normal build because the "init" function is called only once when the app is launched.

Comment 3

[bhavana bajaj [:bajaj]]

(In reply to Julien Wajsberg [:julienw] from comment #2)
> This is a change only for tests, but since the change is part of the build
> and we'd like this to go to the branch, I'm asking leo+ on this. This should
> have no impact for the normal build because the "init" function is called
> only once when the app is launched.

Julien, is the slowness a recent regression ? And to confirm here the user flow of the sms code is never touched at all for a normal user ?

Comment 4

[Corey Frang [:gnarf]]

Comment on attachment 773224 [details] [diff] [review]
patch v1

Review of attachment 773224 [details] [diff] [review]:
-----------------------------------------------------------------

r=me

Comment 5

[Julien Wajsberg [:julienw]]

(In reply to bhavana bajaj [:bajaj] from comment #3)
> (In reply to Julien Wajsberg [:julienw] from comment #2)
> > This is a change only for tests, but since the change is part of the build
> > and we'd like this to go to the branch, I'm asking leo+ on this. This should
> > have no impact for the normal build because the "init" function is called
> > only once when the app is launched.
> 
> Julien, is the slowness a recent regression ?

not really, it's there since the start of the MMS work and the new composer. It's just that I decided to fix it today (I did an early try in bug 888950 that I haven't finished yet).

But I think it's one of the things that makes some tests fail on TBPL (that runs on slow VM).

> And to confirm here the user
> flow of the sms code is never touched at all for a normal user ?

The added code will run only once, and therefore won't have any impact because nothing happened yet. It has an impact if it runs again while the app is running.

Comment 6

[Julien Wajsberg [:julienw]]

master: 4b4688bc6c8e0ddc99453378e00fe53fc14576fb

Comment 7

[bhavana bajaj [:bajaj]]

Leo- this is this a test fix only but may have a user impact given comment #5.