Mozilla build ID:20140422024003 DEFECT DESCRIPTION: Power consumption on ConnectA2 is too high REPRODUCING PROCEDURES: 1. A device with ConnectA2 pre-installed 2. Test the standby current with ConnectA2 3. The test result is 40ma. EXPECTED BEHAVIOUR: ASSOCIATE SPECIFICATION: TEST PLAN REFERENCE: TOOLS AND PLATFORMS USED: USER IMPACT: REPRODUCING RATE: For FT PR, Please list reference mobile's behavior:

Telefonica will test the power consumption with ConnectA2. According to TCL's analysis, the app sets an alarm, and when the alarm coming it calls requestwakeLock but does not release it, so the device is unable to go to sleep. Please contact the app developer to fix it. Thanks.

Hi Jason - Somehow I cannot change this one to "Tech Evangelism", could you help to do so?

Hi Nicole, please help to assign to correspond developer, we suspect this issue is due to Twitter didn't release the CPU wakelock

(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #4) > Hi Nicole, please help to assign to correspond developer, we suspect this > issue is due to Twitter didn't release the CPU wakelock I mean connectA2...

Please review and provide potential fix date. Thank you!

I am sorry, but the success of ConnectA2 is based in this feature. If the mobile goes to sleep, network goes to sleep and you won't receive your messages online. We have discussed already this case with experts and evangelists, and the consumption with normal usage of the device give till 48 hours of battery life, and till more than 72 hours with low usage of the device. This is a huge amount if we take in consideration the 12-24 hours battery life of other smartphones.

(In reply to dev from comment #7) > I am sorry, but the success of ConnectA2 is based in this feature. If the > mobile goes to sleep, network goes to sleep and you won't receive your > messages online. > We have discussed already this case with experts and evangelists, and the > consumption with normal usage of the device give till 48 hours of battery > life, and till more than 72 hours with low usage of the device. > This is a huge amount if we take in consideration the 12-24 hours battery > life of other smartphones. Could you please give some examples of the device with FFOS 1.3 that can standby for more than 48/72 hours with ConnectA2 enabled. Thanks

Geeksphone Revolution consumed 5%/hour of battery with ConnectA2 enabled, which means only 20 hours battery life, refer to http://forum.geeksphone.com/index.php?topic=6081.0. Flame with ConnectA2 enabled, a 30mA constant current consumed because device can't go to sleep because ConnectA2 locked the CPU, it's meaning that in theory, the standby time of Flame will changed from 12 day to 3 day. Any application that always hold the device to sleep will be abandoned by user because it's not acceptable on a SMART phone, if your success depends on an unacceptable fact, it will fail eventually.

Ok. we are going to include an option in Settings to disconnect incoming messages when the app is in background. This will mean 0 battery consumption for users who don't want to receive messages online and who don't want the app to run in background. Estimated date: ConnectA2 v15. ConnectA2 v15 will be send for revision ~15 days from today.

(In reply to dev from comment #10) > Ok. we are going to include an option in Settings to disconnect incoming > messages when the app is in background. This will mean 0 battery consumption > for users who don't want to receive messages online and who don't want the > app to run in background. > Estimated date: ConnectA2 v15. > ConnectA2 v15 will be send for revision ~15 days from today. Hi, since our partner is going to run a power consumption test in Telefonica by the end of this week, is it possible to have a engineering build first by this Thursday(May/8)? Really appreciate your help Vance

Sure. We will attach a v15alpha to this Bug(1003639) on Thursday. Please, take in account we won't have time to pass functional test to this v15alpha version. Juanma

Please note that just having an open connection/socket does *not* increase the power consumption. Doing a tcpdump of the phone with just your app running, it seems that every 120 seconds (aprox) a packet is *sent* from the phone (which most of the time doesn't seem to be answered). That, on the other hand, do increase the power consumption (on the best case, the cell radio will be have to be awakened from IDLE every time a packet is sent, on the worst case it will never go into the IDLE status). Do you have any kind of alarm that tries to send (maybe a ping) every 2 minutes? Just getting rid of that (or increasing the connectivity check to 30 minutes or so) should help a lot with battery life without forcing the users to decide between killing their battery time or having a IM that only works when they check it.

Hello Antonio, We appreciate your recommendation. We send this packet because after 3 minutes with the screen off and without network activity, the connection will be closed (at least with our test in FFOS 1.1). It could be a good idea to put the cell radio modem in some type of low power consumption after switching off the screen. At least to change the mode from 3G to 2G, or other mode with lower power consumption but still with Internet connection. But this is out of our scope, as App developers. Juanma

(In reply to dev from comment #12) > Sure. > We will attach a v15alpha to this Bug(1003639) on Thursday. > Please, take in account we won't have time to pass functional test to this > v15alpha version. > > Juanma Hi Juanma - Appreciate your help, it would be better if you can provide the alpha tomorrow(May 7) such that we can have more time(the power consumption test normally need to run for 24 hours to measure) to verify if the power consumption of the device can indeed improve with the alpha version Connecta2 Thanks for your strong support Vance

(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #15) > (In reply to dev from comment #12) > > Sure. > > We will attach a v15alpha to this Bug(1003639) on Thursday. > > Please, take in account we won't have time to pass functional test to this > > v15alpha version. > > > > Juanma > > Hi Juanma - > > Appreciate your help, it would be better if you can provide the alpha > tomorrow(May 7) such that we can have more time(the power consumption test > normally need to run for 24 hours to measure) to verify if the power > consumption of the device can indeed improve with the alpha version Connecta2 > > Thanks for your strong support > > Vance Hi Vance, Tomorrow in what local time? Midday PDT or other? Juanma

(In reply to dev from comment #16) > (In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #15) > > (In reply to dev from comment #12) > > > Sure. > > > We will attach a v15alpha to this Bug(1003639) on Thursday. > > > Please, take in account we won't have time to pass functional test to this > > > v15alpha version. > > > > > > Juanma > > > > Hi Juanma - > > > > Appreciate your help, it would be better if you can provide the alpha > > tomorrow(May 7) such that we can have more time(the power consumption test > > normally need to run for 24 hours to measure) to verify if the power > > consumption of the device can indeed improve with the alpha version Connecta2 > > > > Thanks for your strong support > > > > Vance > > Hi Vance, > > Tomorrow in what local time? Midday PDT or other? > > Juanma Hum...ya Midday PDT will be fine, BTW, where are you based?

(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #17) > (In reply to dev from comment #16) > > (In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #15) > > > (In reply to dev from comment #12) > > > > Sure. > > > > We will attach a v15alpha to this Bug(1003639) on Thursday. > > > > Please, take in account we won't have time to pass functional test to this > > > > v15alpha version. > > > > > > > > Juanma > > > > > > Hi Juanma - > > > > > > Appreciate your help, it would be better if you can provide the alpha > > > tomorrow(May 7) such that we can have more time(the power consumption test > > > normally need to run for 24 hours to measure) to verify if the power > > > consumption of the device can indeed improve with the alpha version Connecta2 > > > > > > Thanks for your strong support > > > > > > Vance > > > > Hi Vance, > > > > Tomorrow in what local time? Midday PDT or other? > > > > Juanma > > Hum...ya Midday PDT will be fine, BTW, where are you based? Our company is based in the UK (London). The developing is based at the moment in Luxembourg , UK +1. Juanma

ConnectA2 v15 pre-alpha is attached. Please, in the main screen "CHATS", go to menu "▼", then "Config" and disable "Notifications in background". Juanma

Hi Juanma When will you release an official version into Market place, so that end user can update to the version with the option to disable notification in background? Thanks

Hello Hong, It is planned to be ready for the final user in 2nd/3rd of June. Juanma

Juanma, could you please check the suggestion provided in Commnet 13? We do think it is a better option to the issue. >In reply to Antonio Manuel Amaya Calvo (:amac) from comment #13) > Do you have any kind of alarm that tries to send (maybe a ping) every 2 > minutes? Just getting rid of that (or increasing the connectivity check to > 30 minutes or so) should help a lot with battery life without forcing the > users to decide between killing their battery time or having a IM that only > works when they check it.

Hello Beatriz, I have answered in the comment 14. We did some test and the connection goes off if there is not network activity in a 3 minutes period. At least with FFOS 1.1 in ZTE Open. Juanma

(In reply to dev from comment #23) > Hello Beatriz, > > I have answered in the comment 14. We did some test and the connection goes > off if there is not network activity in a 3 minutes period. At least with > FFOS 1.1 in ZTE Open. > > Juanma What do you need the connection for if the user isn't using the app at the moment? What you can do is, if the app is on background, just let the connection lapse and reconnect on a timer (a configurable timer so the user can decide if he wants to get the messages faster or the battery to last more). And yeah, apparently WA closes the connection (it's not FirefoxOS the one closing it) after 3 minutes of inactivity.

Hello Antonio, Many thanks for taking the time to check our improvements. We appreciate it. We are going to do what you told us. We are going to include a functionality in Settings to remove completely the background activity. We are also considering including a time range as you recommend, but we don't know if we will have time to deliver this functionality for v15, because we have to pass many test cases, and these cases take a lot of time due to the 30-minutes periods we have to wait to check it. Finally, related with WA connection. In FFOS 1.0, we only sent PING every 15 minutes. After the update from FFOS 1.0 to 1.1 we received many emails of complaints because they didn't receive the messages in background and we discovered this new network disconnection 3 minutes after switching off the display. Juanma (In reply to Antonio Manuel Amaya Calvo (:amac) from comment #24) > (In reply to dev from comment #23) > > Hello Beatriz, > > > > I have answered in the comment 14. We did some test and the connection goes > > off if there is not network activity in a 3 minutes period. At least with > > FFOS 1.1 in ZTE Open. > > > > Juanma > > What do you need the connection for if the user isn't using the app at the > moment? What you can do is, if the app is on background, just let the > connection lapse and reconnect on a timer (a configurable timer so the user > can decide if he wants to get the messages faster or the battery to last > more). And yeah, apparently WA closes the connection (it's not FirefoxOS the > one closing it) after 3 minutes of inactivity.

link all Fire C (codename: Sora) bugs to a meta one.