Page 1 of 2 12 LastLast
Results 1 to 20 of 35
  1.    #1  
    Hands down this is the most important issue with chatter users...there has been so many discussions in multiple threads...theory, hypothesis, conjecture...

    MARC!!!
    is anybody sending you logs?....what are they revealing?...what is this battery drain enigma?...can we have anything conclusive stated?...(eg "it is what it is", "i think this can be fixed", "i don't know what the hell you guys are talking about")

    ADMINSTRATORS!!!
    make a sticky battery drain thread...

    detroitkruk
  2. #2  
    I don't have a battery drain problem.
    Palm V-->Visor Deluxe-->Visor Prism-->Visorphone-->Treo 180-->Treo 600-->Treo 650 on Sprint-->Treo 700p-->Centro-->Diamond-->Pre-->HTC EVO 4g???!
  3. #3  
    I'm not sure which problem you're talking about. The only enigma I'm aware of is that some people (maybe a handful at this point) see big battery drain overnight while Chatter is supposedly not running. I haven't gotten ANY logs for this, AFAIKAFAIKAFAIK.

    Marc
  4. resturg's Avatar
    Posts
    210 Posts
    Global Posts
    217 Global Posts
    #4  
    I do not have the “Battery Drain” problem with my Sprint T650 V1.12 phone. I can get on a full battery get 31 hours of use with out Chatter’s “Efficient Sync” checked. I am using the current Chatter version 1.1RC6.

    Note: The measurements were done on the weekend when my phone usage is the lowest. I will re-test next weekend with “Efficient Sync” checked.
  5. #5  
    No problems on my Orange UK phone... I use IMAP accounts.
  6. #6  
    I used to think there was a battery drain but I've just gotten used to ChatterEmail using that much battery. The fact of the matter is that the only way to please everying is when Chatter uses less than 1% of the battery while it is on all day long.

    Originally I was going to switch away from Chatter because of the battery usage but after using it for two weeks I decided that all I need to do is purchase a charger for my desk at work
  7. DocGo's Avatar
    Posts
    341 Posts
    Global Posts
    593 Global Posts
    #7  
    I use a higher capacity 3rd party battery (Mugen) and I only charge my Treo every night (with battery still around 60% and Chatter running from 6:30 am till 9pm before charging). I have 2 IMAP accounts and they sync all the time in the background(using Matrix SSL).
    The Doctor is IN!
    Gadget: Rogers Treo 680 (unlocked) GSM
    Cases: Still deciding
  8. #8  
    GSM Treo owners don't have issues getting a day's worth of battery in general. For CDMA Treo owners, there are usually ways of getting a day's worth, whether by using options like "Efficient Sync" or getting a larger battery, or a charger, or reducing the number of online boxes, etc. It's a matter of priorities and tradeoffs...

    Marc
  9. #9  
    I get through the day fine with a sprint 650 1.12.. the efficient sync option helps alot I usually have around 45 to 50% left when i get home and that's with a couple of calls and a little surfing.
    The only time I ever made a mistake was when i thought I did and was mistaken.
  10. #10  
    What donric said... http://img.photobucket.com/albums/v1...miley/nods.gif

    The only time I had a battery drain problem from Chatter is about a week when fastmail was acting funky. Chatter kept "Connecting" the whole time i was asleep, so when I woke up my battery was at 5%.
    MaxiMunK.com The Forum That Asks, "Are You Not Entertained?"

    Remember: "Anyone that thinks the Treo should just work right out of the box, shouldn't own a Treo..."
  11.    #11  
    uh...okay...

    where's all the people who've been posting to the other battery drain threads?...

    MARC: have any of those people sent the logs you mentioned to them? what have they shown?

    i have no problem being alone in my insanity...but, uh...seems like there's been tons of stuff on this recently...granted, this may be cdma specific...

    nevermind?...c'mon...

    detroitkruk
  12. #12  
    Quote Originally Posted by detroitkruk
    uh...okay...

    where's all the people who've been posting to the other battery drain threads?...

    MARC: have any of those people sent the logs you mentioned to them? what have they shown?

    i have no problem being alone in my insanity...but, uh...seems like there's been tons of stuff on this recently...granted, this may be cdma specific...

    nevermind?...c'mon...

    detroitkruk
    Are you talking about just the typical battery usage? I thought you were having some wild depletion issues...

    Well, Marc and others have already mentioned things to improve this, Effecient Sync, etc...
    MaxiMunK.com The Forum That Asks, "Are You Not Entertained?"

    Remember: "Anyone that thinks the Treo should just work right out of the box, shouldn't own a Treo..."
  13. JoeSam's Avatar
    Posts
    88 Posts
    Global Posts
    89 Global Posts
    #13  
    Detro - I'm with you. I can't figure out what exactly is causing it, but when Chatter is running, my battery drains at at least twice the rate as when it's not running. And this is when I receive no messages or very few on my IMAP online account (the only one installed).

    I ran an unscientific test overnight on Saturday (my lowest-peak night to attempt). I received no phone calls, made no phone calls and recieved no emails. I had chatter running all night. After 8 hours I had <80% battery. This phone is supposed to have a 300 hour standby time. Even if you cut that in half to 150 hours, it seems strange to lose twenty per cent of standby battery capacity (30 standby hours) in 8 hours. Again, unscientifically, Chatter seems to be draining standby hours at a rate of as much as 5x, depending on battery quality.

    I don't see this behavior when Chatter is not running (Menu-Q to quit). Granted, I'm not getting the advertised battery life of 300 hours standby, but I do see a very slow and gradual battery drain when Chatter is off. I disabled Chatter all day yesterday and used Versamail. I received approximately 25 emails and used the phone for about 1.5 hours of talk time. I had 75% battery at 10PM when the phone went on the charger. On Friday, I had roughly the same usage, but used Chatter w/ efficient sync instead of Versamail, and I got into the red zone by 5PM and had to swap out batteries.

    By the way, I'm using a Sprint T650 w/ the 1.12 software, hardware B.

    Marc says that my battery drain data seems normal to him, but I don't think so. I think there is deffinitely an issue here that should be addressed. If Chatter is a battery hog for CDMA users, they might think twice or invest, as Marc says, in a high capactiy battery prior to using this program.

    I will say this -- despite the battery issue, I do think Chatter is a superior email app compared to everything else on the market. The only thing I wish it did was to somehow sync up w/ my exchange-based calendar without doing a hotsync, like Versamail does.
  14. #14  
    JoeSam - Chatter uses battery when it's running; there is no mystery about this! What you're seeing is 100% expected behavior.

    Marc
  15. ptyork's Avatar
    Posts
    69 Posts
    Global Posts
    70 Global Posts
    #15  
    Okay, I worked through all of the issues with battery drain (and missed calls) through personal research and experimentation, and in a characteristically selfish manner, I didn't share. To correct this, I'll attempt to write it all down here for general consumption. This applies to CDMA/1xRTT phones (Sprint and Verizon). GPRS is a much more efficient (albeit slower) protocol, so this won't apply to GSM users.

    The first thing to remember is that any time you have an active network connection, you are in essence making a phone call. Regardless of whether you transfer 1-byte or 1 megabyte of data, the same call has to be placed from the handset to the tower. Because the 1xRTT (data) channel uses the same radio as the CDMA (voice) channel, you will use the same amount of power making a voice call as you would a data call, call times being equal. Note also that whenever you have and active data connection, you are using the one and only radio channel available to the phone. Unlike GPRS (which allows a voice call to hijack the data connection), CDMA phones will simply send any voice traffic directly to voice mail until all data activity has ceased.

    Another factor in all of this is the way CDMA phones seems to handle data connectivity. It seems that data calls will always stay connected for a predetermined minimum amount of time regardless of the amount of data sent and received. I'm not exactly sure the reason behind this, but the effect of it is that even the simplest and shortest data request from a Sprint phone will result in a data connection that lasts for around a minute. So, no matter what, each instance of data access will result in a minimum of a minute of "talk-time" battery usage, and of course a minute of blocked calls. I'm to understand that for Verizon, this "latency" is worse, and thus battery life and blocked call issues will be worse, as well.

    Chatter's IMAP-push implementation utilizes something called an IDLE command, which instructs the server to alert the client any time there is a change in the status of the mail stored there. For most folks, this is most often due to the arrival of new mail (though moves to folders, changes in read/unread status, etc. will also trigger a response), so it serves as a great mechanism for implementing the push functionality. The downside is that, regardless of whether mail is received or not, the client needs to "ping" the server every so often to let it know that it is still there and interested in receiving updates. For Chatter, this "ping" occurs every ten minutes. So as long as a mailbox is "online," Chatter will initiate a data connection every ten minutes. Given the knowledge we know from the above paragraphs, this will result in a minimum of six full minutes of "talk-time" used every hour. Since this ten-minute timeout resets every time e-mail is received, this six-minute usage can increase dramatically under a heavier load of e-mails.

    Since Chatter runs in the background, this activity will occur without you even being aware of it. You are making a minimum of six calls an hour whether your phone is "on" or "off" unless you take your mailbox(es) offline or you quit Chatter entirely. Given that we should realistically expect around 200 minutes or so of talk-time (I know this is supposed to be higher, but I'm being realistic), you will see a 6/200 or 3% drop in battery life per hour AT A MINIMUM whenever Chatter is running with online mailboxes.

    The reason that efficient sync is effective in lowering battery usage is that checking this option will delay updates made at the handheld (reads, deletes, etc.) from propagating immediately to the server. Instead, Chatter will wait until it already has to make a connection (e.g., when it receives an e-mail or the ten minute "ping" time has elapsed) to send the updates. This can, depending on usage habits, cut your data usage considerably and, thus, significantly improve battery life.

    Please note that this behavior is not the fault of Chatter in the slightest. In fact, if you set Versamail to sync every 10 minutes, you should see similar battery life issues. Also note that Chatter allows you the option to quicksync every X minutes, so you can achieve similar battery life gains by utilizing this option and increasing the time between syncs, as well. Further, the option to use an SMS trigger to initiate mail retrieval may also help significantly in improving battery life, especially for those who receive a relatively small load of e-mail.

    There's a bit more that I could share on this, but I think this should give you enough to understand the basic issues related to battery drain (and blocked calls) when using Chatter. If you have further questions for which I have answers, I'll be glad to share so ask away.

    Paul
  16. #16  
    Paul -

    Many thanks for that; it's a terrific primer on the issues involved in CDMA data transfer. Might I use it as an appendix to the upcoming manual revision for 1.1?

    Marc
  17. aldamon's Avatar
    Posts
    650 Posts
    Global Posts
    684 Global Posts
    #17  
    Quote Originally Posted by ptyork
    I'm not exactly sure the reason behind this, but the effect of it is that even the simplest and shortest data request from a Sprint phone will result in a data connection that lasts for around a minute. So, no matter what, each instance of data access will result in a minimum of a minute of "talk-time" battery usage, and of course a minute of blocked calls. I'm to understand that for Verizon, this "latency" is worse, and thus battery life and blocked call issues will be worse, as well.
    Is this true? It seems like the connection lasts far less than a minute. I'm probably splitting hairs here and a connection of less than a minute is a good thing anyway.

    Quote Originally Posted by ptyork
    The downside is that, regardless of whether mail is received or not, the client needs to "ping" the server every so often to let it know that it is still there and interested in receiving updates. For Chatter, this "ping" occurs every ten minutes.
    OK, but do all servers need to be pinged every 10 minutes to maintain the IDLE connection? Would it possible to make the number of pings configurable so we could experiment with what our own mails server need instead of arbitrarily connecting every 10 minutes? Maybe my server only needs to be pinged every hour? How would I know without experimenting?
  18. ptyork's Avatar
    Posts
    69 Posts
    Global Posts
    70 Global Posts
    #18  
    Quote Originally Posted by aldamon
    Is this true? It seems like the connection lasts far less than a minute. I'm probably splitting hairs here and a connection of less than a minute is a good thing anyway.
    It appears to disconnect relatively quickly, but if you look at the green/grey arrows above the signal strength meter, you'll notice that they are green for quite some time after any real network activity ceases. Also, you can, say, do a quicksync and then try calling yourself repeatedly. You'll see that the line stays "busy" for a good while after all network transfers stop. It may not be a minute exactly, but certainly 30 seconds or more.

    Quote Originally Posted by aldamon
    OK, but do all servers need to be pinged every 10 minutes to maintain the IDLE connection? Would it possible to make the number of pings configurable so we could experiment with what our own mails server need instead of arbitrarily connecting every 10 minutes? Maybe my server only needs to be pinged every hour? How would I know without experimenting?
    Good question. I had the same one for Marc a while ago. See the following thread on his forums.

    http://www.chatteremail.com/index.ph...iewtopic&t=670

    Most servers do in fact have a timeout of far greater than 10 minutes. The only danger is that you could have network blip that causes the IDLE request socket to be reset. In this case, a 10 minute interval insures that even if a network error occurs immediately after the IDLE request, you will at most have a 10 minute delay in receiving new mails (until Chatter resets the IDLE request). Setting it to a higher value will of course result in a potential delay of whatever this new value is. It's a tradeoff.

    Quote Originally Posted by chatter
    Many thanks for that; it's a terrific primer on the issues involved in CDMA data transfer. Might I use it as an appendix to the upcoming manual revision for 1.1?
    Marc, of course you can use what I've written in any way you'd like. No worries.

    Incidentally, did you ever increase the maximum allowable IDLETIME value? If not, then I'll make that a condition of using my post in the 1.1 manual.

    Paul
  19. #19  
    Yes, it's increased to 7200 (120 minutes).

    Marc
  20. ptyork's Avatar
    Posts
    69 Posts
    Global Posts
    70 Global Posts
    #20  
    Quote Originally Posted by ptyork
    Incidentally, did you ever increase the maximum allowable IDLETIME value? If not, then I'll make that a condition of using my post in the 1.1 manual.
    Marc, nevermind, I see you already increased the maximum to 7199, which is certainly sufficient.

    Aldamon, use the console in Chatter and enter IDLETIME (seconds) where "(seconds)" is the number of seconds to wait before issuing the "ping." For example, entering IDLETIME 3600 would set the time between pings to 3600 seconds or 1 hour.

    Paul
Page 1 of 2 12 LastLast

Posting Permissions