Jump to content

kabkab

Members
  • Content Count

    20
  • Joined

  • Last visited

Community Reputation

0 Neutral

About kabkab

  • Rank
    Newbie

Profile Information

  • Your Current Device(s)
    i900
  1. try to change following settings in registry: [HKEY_LOCAL_MACHINE\\HARDWARE\DEVICEMAP\TOUCH] "Click"=dword:2000 [HKEY_LOCAL_MACHINE\\HARDWARE\DEVICEMAP\TOUCH] "DownUp"=dword:750
  2. Can you tell me - this is "almost like official ROM should look like" - i mean clean WM6.5 + some samsung apps (and only volume control from omnia lite) or what? I'm asking because you did not publish lisf of apps included...
  3. Ryrzy what is difference between normal and lite versions of classic ROM? wanna try official 6.5 and could not choose... i need MS Office and all samsung apps but did not need any 3rd-party programs, including m2d...
  4. well, this is not secret. :-) you need IDA Pro (disassembler) - to disassemble and understand driver. more info here: http://www.hex-rays.com/idapro/ also you need HEX editor and patcher. i prefer HIEW, which support (from 8.0 version) ARM. more info here: http://www.hiew.ru/ before using this tools, be ready to learn ARM-assembler opcodes and programming basics. also, did not forget to sign driver with Microsoft Priviledged Developer certificates.
  5. sorry, as i said before, i did not have i910 and cannot even test driver to see wassup. ps. most funny, that driver did not have inside ANY code for turn backlight ON. it have code for turn backlight OFF in case of 10% of charge left.
  6. this is i910 driver since i900 users did not encounter any problems...
  7. OS itself did not need to know about battery. Any program (including battery icon on taskbar) can: 1. query power status itself 2. subscribe to power change event. first case we cannot control. second case - i'm not sure. try to search for "RequestPowerNotifications" function in MSDN. this notification is raised by (who?) and then OS ask battery driver for status and send notifications to programs. may be this notification is raised by battery driver itself - i don't know. but it should not. quick search in google give me some "power manager". anyway, more research needed. any info appreciated.
  8. Ok, here new test version for i910 256_5 - 256 values used for sum, decrement goes in 5%. 1024_1 - 1024 values used for sum, decrement goes in 1%. check against power drain! THIS IS i910 DRIVERS zylonite_new.zip Please, post feedback here - should i need to move this files to #1 post for everyone or delete this files, if they did not have any positive effect.
  9. 2 All i'm not worrying and relaxing. :-) Technical details how driver working: 1. when system ask for power status driver ask battery IC for voltage. 2. this battery IC return some ADC value (this is voltage value, but indirect, in some relation measurement units) 3. after than driver try to calculate "compensated" value - it did not have current that device consume, but try to understand wassup (phone on, wifi on, gps on) and compensate this value. 4. after that driver calculate average ADC value (sum last 15 values and divide it by 15). 5. after that driver compare this average ADC against constant values of ADC and return real battery % value. Patch so, i patch step 4: right now driver use 64 or 256 (see post#1) values to calculate average ADC (15 is not enough, you'll see a lot of % changing up and down). also i patch step 5: instead of fixing compare and return fixed 100/80/60/40/20... - i substract by 100 and return in 1% step values. Possible problems with power drain seems, that in some firmware's or some OS builds - system/some programs/i dunno who - perform some recalculations when battery % is changed. because in this patched driver % really changed very often - this can happen massive power drain if case of massive recalculations. Possibe solutions When have some free time i can do following: 1. Increase sum of average values from 64/256 to much more value, like 1024/2048 - up to 65535. 2. Decrease accuracy and use 5% instead of 1%. both steps will cause to slow % changing (and hit accuracy of coz) and _CAN_ (but not sure, that WILL) reduce power drain. sorry, i did not have i910 and cannot test or see wassup. suggestions are welcome
  10. this is very funny, because i can print code in driver before and after patch (or you can compare it yourself). and if you are understand ARM assembler - you can see, that i did not make any change that can affect power drain! even some calculations removed! may be this can affect another way (for example, driver return different values more often than older driver and OS perform some recalculations and so on - i dunno). but this is not problem of driver - this is problem of current OS build. anyway - on my i900 it's working perfect. if you not like it - don't use it! :-)
  11. very strange. and what show original driver? bcoz i have same device and same firmware and no problems... (right now, after 1 day of heavy usage i have 73%, original driver show me 80%).
  12. To all users having some issues with 1% battery driver Sorry, may be you have different firmware, may be you device have some miscompatibility with this patch. I don't know and have no ability to install different firmware to my device and check each one. Just assume, that you cannot use this driver and rollback to original one.
  13. again: on my i900 it's working good. i have wm6.1 and DXID1 firmware. i did not checked against wm6.5 or i910 (i don't have such device). what don't work? did not shown or some bugs? do you install it correctly? checked, that file copied and overwrite original? make soft reset?
  14. For all ROM-makers (and everyone else) - yes, i don't mind to include this driver to you ROMs (btw, thx to you work!!!) or anywhere else!
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.