1090026 - [Messages] It cannot display failed icon in conversation list.

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: Gaia::SMS, defect, P2)

Description

[sync-1]

Mozilla Build ID: 20140916000205
 Created an attachment (id=990897)
 issue_pic
 
 DEFECT DESCRIPTION:
 It cannot display failed icon in conversation list.
 
  REPRODUCING PROCEDURES:
 1.Send a message to a invalid recipient（eg.14）；
 2.return to conversation list after tapping back icon;
 3.It cannot display failed icon in conversation list-->KO
 
  EXPECTED BEHAVIOUR:
 It can display failed icon in conversation list.
 
  ASSOCIATE SPECIFICATION:
 
  TEST PLAN REFERENCE:
 
  TOOLS AND PLATFORMS USED:
 
  USER IMPACT:medium
 
  REPRODUCING RATE:5/5
 
  For FT PR, Please list reference mobile's behavior:

Comment 1

[sync-1]

Attachment: issue_pic

Comment 2

[Julien Wajsberg [:julienw]]

Yep, I really want this, but we need gecko support to have this information in the "thread" object [1].

[1] http://mxr.mozilla.org/mozilla-central/source/dom/mobilemessage/interfaces/nsIDOMMozMobileMessageThread.idl

NI Bevis to fire a Gecko bug.

Once we have this it's trivial to do in Gaia.

Comment 3

[Bevis Tseng[:bevis][:btseng](Exited)]

(In reply to Julien Wajsberg [:julienw] from comment #2)
> Yep, I really want this, but we need gecko support to have this information
> in the "thread" object [1].
> 
> [1]
> http://mxr.mozilla.org/mozilla-central/source/dom/mobilemessage/interfaces/
> nsIDOMMozMobileMessageThread.idl
> 
> NI Bevis to fire a Gecko bug.
> 
> Once we have this it's trivial to do in Gaia.

Yes, gecko change is needed to support this feature.

I'd like to set NI to get some input from UX to see what detail information we have to provide in the thread object for this feature, for example,
1. Similar to the |unreadCount| in [1], Is it enough to have |failedCount| for gaia to display this indicator when |failedCount| > 0.
2. Shall the indication be displayed only when the last message of this thread is a outgoing message and was failed to sent.
3. In advance, what about MMS retrieval?
   Do we need another indication to inform user that the incoming MMS message was not able to be retrieved after auto-download or manual-download was requested?

We can check how other reference phone behaves in these scenario to figure a better solution. :)

Comment 4

[田旻]

Dears,
Can this bug been fixed on v2.0?
BRs,
TIAN Min

Comment 5

[Julien Wajsberg [:julienw]]

No sorry, this needs gecko work, and this is always risky.

Comment 6

[Jenny Lee]

Hi Bevis,

Sorry for the late reply! Here's the quick answer to your question. We can find some time to sync up the details you need to implement this feature ;)

> 1. Similar to the |unreadCount| in [1], Is it enough to have |failedCount|
> for gaia to display this indicator when |failedCount| > 0.
On inbox UI, number of fail sent message is not important, as long as there's 1 message failed to send out, show indicator.

> 2. Shall the indication be displayed only when the last message of this
> thread is a outgoing message and was failed to sent.
As long as there's message failed to send out, show indicator, order is irrelevant. So unless user delete or resend the message, the indicator will live on inbox page.  

> 3. In advance, what about MMS retrieval?
>    Do we need another indication to inform user that the incoming MMS
> message was not able to be retrieved after auto-download or manual-download
> was requested?
I don't think we need such indication because user will see the retrieve fail state upon entering thread view, and before entering thread view, we will have an unread indicator on inbox view, so this is covered.
 
> We can check how other reference phone behaves in these scenario to figure a
> better solution. :)

Thanks!

Comment 7

[Julien Wajsberg [:julienw]]

Hey Jenny, I quite disagree with 2. I understand what you mean, but in my personal case, it happens that I have old failed messages that I don't care about. I'd like a way to not have the thread in error in some cases.

Maybe visiting the thread (marking it as read) could also remove the error mark?

Comment 8

[Bevis Tseng[:bevis][:btseng](Exited)]

I am okay for the approach of case 1) and case 2).
However, for case 3), it not true for manual-download case, IMO.

For manual-download, user has to enter the thread view and clicked the |download| button of the mms notification.
However, the time to finish the download depends on the network conditions and the size of the message.
There is no guarantee that user will say in the thread view until the request is finished.
Hence, user always has to go to the same thread view again to know the result of that retrieval.

Can we have more information about how other reference phones behave in these scenarios for further discussion?

Comment 9

[Jenny Lee]

(In reply to Julien Wajsberg [:julienw] from comment #7)
> Hey Jenny, I quite disagree with 2. I understand what you mean, but in my
> personal case, it happens that I have old failed messages that I don't care
> about. I'd like a way to not have the thread in error in some cases.
> 
> Maybe visiting the thread (marking it as read) could also remove the error
> mark?

Hello Julien,
Failure to send a message is something we want to make sure user is informed, especially if we're running FFOS on a less stable device:p Making fail sent mark behaves the same as unread mark would seem a little risky to me, plus, user can always choose to delete or resend the message. Let me know if this doesn't make sense to you, thanks!

Comment 10

[Julien Wajsberg [:julienw]]

If you insist on this then I don't disagree, you're our UX master after all :)

Just NI again for comment 8, I think you missed it?

Comment 11

[Jenny Lee]

(In reply to Julien Wajsberg [:julienw] from comment #10)
> If you insist on this then I don't disagree, you're our UX master after all
> :)
> 
> Just NI again for comment 8, I think you missed it?

oops I did miss it, thanks Julien :D

Hi Bevis,

I tried a few times on Android device and didn't manage to create the manual download failure case, iOS 8 doesn't seem to support manual-retrieve. Nevertheless I am able to make decision without the info.

If we tap on download and tap on back to go to inbox immediately, upon the attachment is successfully downloaded, a unread indicator will appear. So it only makes sense to show a *red* failure indicator if the attachment has failed to download. Let me know if you need more info on this, thank you for the feedback!

Comment 12

[Julien Wajsberg [:julienw]]

(In reply to Jenny Lee from comment #11)
> (In reply to Julien Wajsberg [:julienw] from comment #10)
> > If you insist on this then I don't disagree, you're our UX master after all
> > :)
> > 
> > Just NI again for comment 8, I think you missed it?
> 
> oops I did miss it, thanks Julien :D
> 
> Hi Bevis,
> 
> I tried a few times on Android device and didn't manage to create the manual
> download failure case, iOS 8 doesn't seem to support manual-retrieve.
> Nevertheless I am able to make decision without the info.
> 
> If we tap on download and tap on back to go to inbox immediately, upon the
> attachment is successfully downloaded, a unread indicator will appear. So it
> only makes sense to show a *red* failure indicator if the attachment has
> failed to download. Let me know if you need more info on this, thank you for
> the feedback!

Hey jenny,
I just want to make sure I understand: is it overriding what you said in comment 6 point 3?

Comment 13

[Jenny Lee]

Hello Julien,

Yes. Sorry for sharing my thinking process and the inconvenience that might cause you (and Bevis!!) :(
 
After discussing with Bevis, we conclude that whenever user is not on the message thread at the moment when file retrieving fails, show red indicator on inbox view. If user press home to leave Message during retrieving, and later comes back to Message (to the same view), user would see either retrieval is successful or failed, at this time, no need to show indicator on inbox view.

Comment 14

[Julien Wajsberg [:julienw]]

Jenny, so it means that for "retrieve MMS" case, the "error" should essentially works like the "unread" mark (means we remove it if the user sees the failed retrieval), but for other cases, we always show the mark?

To be clearer:
* errors in incoming messages => remove the mark in "inbox" if the user enters the thread. (but we still keep the message in error state)
* errors in outgoing messages => never remove the mark in "inbox" if there are errored messages.

I'm not sure how to do all this :/ currently all this state is kept in Gecko.

Comment 15

[Jenny Lee]

Hi Julien,

Yes this is exactly how this should work ;) 
Hmm I've discussed this with Bevis and it looks like the wanted behavior is doable, but let's double check.

ni Bevis for feasibility comment. Thanks Bevis!

Comment 16

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Julien,

Sorry to say that there was some misunderstanding between Jenny & I regarding this topic.
To prevent further confusing,
we'll review our conclusion f2f right before updating the comment in this bug. :)

Regards.

Comment 17

[Julien Wajsberg [:julienw]]

Mass closing of Gaia::SMS bugs. End of an era :(

Comment 18

[Julien Wajsberg [:julienw]]

Mass closing of Gaia::SMS bugs. End of an era :(