Wednesday, January 27, 2010

Progress update for next alpha & beta

I am now one bug fix away from launching these The bug is that certain apps are stopping the correct operation of the 2.0 optimized method I am now using. When screen first sleeps, my process attempts to get started and ready to give you a lag-free, unlocked wakeup. Basically primes itself by taking over the lockscreen. But if browser was open, this fails, & instead it only starts when you wake the screen up. It's like the request gets paused and the OS puts it off till we wake the screen up again. The result is you have to sleep it again and wake up again to get unlocked, because the logic is to do nothing at start, and wait for the next wakeup. When google talk or the home screen launcher (or even the home replacement I use- slidescreen) no conflict happens, it works like it should. But then when you're trying to wake it up to resume where you were in browsing, it just doesn't get started correctly. Totally misses the boat. The point of the 2.0 method is that it is very smooth and fast, and never fails due to CPU load. It never has a flicker. You never see the stock lockscreen since we override it at sleep. So this next release will be really solid and a better experience, if there is any fix for this bug.

You can download a totally functional test build that shows where we are at with this, it just will have a large issue trying to work on the first try when browsing or a few other apps. For the actual alpha 3 it will have no widget lockscreen option as that method won't be finished till we have launcher functions & key customization working within it (for the beta). Other than that it should include a preference whether sliding open should cause the unlock, to help protect against accidental pocket slider bumps, have had some requests for that.


Update: Still trying to resolve some more bugs. Going crazy over the failure to wakeup that seems to happen with a lot of attempts following the power key re-sleep. My current fix seems to resolve it when timeout sleeps happen. the bug doesn't want to happen when i'm plugged in to USB monitoring the program log. something in the timing is the only root cause I can think of.
I have a fix so it very reliably yields to any screen popup and runs very smoothly, plus allows exit via home key and use of notification bar. If i can get this bug killed then I will add the option if users want an Unlock slider and roll out alpha 3 with some cool new options and the custom lockscreen....
Posted by at

2 comments:

  1. JustinFebruary 2, 2010 at 1:03 PM

    just wanted to let you know that i'm still following this and excited for any progress. i'm still using the lite version at this point, even tho it is a little buggy. do you think that's the best version to be using if i just want to not be required to swipe?

    ReplyDelete
  2. myLockFebruary 2, 2010 at 5:20 PM

    Well, I think the version 2c might perform more reliably than lite at this point. The alpha 3 progress will create the same implementation as an option where no wakeup customization is done but it more smoothly and efficiently gets past the lockscreen requirement.

    I have fixed the more annoying bug I was describing in this post and am left with just one, which is a fairly rare failure for the myLock process to receive the power key (causing the result that the phone is awake but screen hasn't come on). If I can get a fix for that so it will always be passed the button presses reliably, and then fix a few loose ends that didn't get priority because of the critical bug fix, then a featured alpha 3 version will be ready to roll.

    ReplyDelete

Subscribe to: Post Comments (Atom)