Page 39 of 53 FirstFirst ... 29343536373839404142434449 ... LastLast
Results 761 to 780 of 1045
  1. #761  
    Quote Originally Posted by unixpsycho View Post
    1) some things in Palm's kernel would break if debugging functions were disabled. I spend alot of time removing them and alot of time with a hosed kernel. it'll be removed in due time. it can probably be removed now since F102/F104 removed huge pieces of code, but I just havent tried it.
    I see. I know you've been putting a LOT of time into stripping out as much of the debug junk as possible. I guess I hadn't realized they built so much into it (and reliance on it) that things would totally fall apart if it wasn't there.

    Quote Originally Posted by unixpsycho View Post
    2) preempt works but it's terribly br0k3n with pulseaudio. sbromwich was playing with that. I like bleeding kernels but a preempt Pre kernel had bones sticking out and raw flesh exposed.
    Pulseaudio. Yea, I should've guessed that was (one of) the hang up(s). It seems like such a good idea on the surface, but in practice I've not read anything all that great about it. I've just been using ALSA stuff for most of these years and it just kinda works. Maybe once Pulseaudio matures more, I'll sacrifice my machines to it, but until then, I'll just run with ALSA. It just kinda works!

    Where there any *perceived* performance boosts from the preempt tests? Or is sbromwich the better person to take these kinds of questions to?


    M.
  2.    #762  
    Quote Originally Posted by Xanadu73 View Post
    I see. I know you've been putting a LOT of time into stripping out as much of the debug junk as possible. I guess I hadn't realized they built so much into it (and reliance on it) that things would totally fall apart if it wasn't there.



    Pulseaudio. Yea, I should've guessed that was (one of) the hang up(s). It seems like such a good idea on the surface, but in practice I've not read anything all that great about it. I've just been using ALSA stuff for most of these years and it just kinda works. Maybe once Pulseaudio matures more, I'll sacrifice my machines to it, but until then, I'll just run with ALSA. It just kinda works!

    Where there any *perceived* performance boosts from the preempt tests? Or is sbromwich the better person to take these kinds of questions to?


    M.
    preempt runs much smoother. but there was alot of kernel oops's in the background due to pulseaudio.

    I gave up on preempt since Palm is coming out with a newer kernel and at that point I will port BFS to it, and maybe even preempt with that.
    Live free or DIE!
  3. irateb's Avatar
    Posts
    257 Posts
    Global Posts
    269 Global Posts
    #763  
    I wasn't sure if user experiences were still needed or being captured for these kernels, but I figured I'd post and then ask a possibly off topic question. I started using F104 about two weeks ago, and it has been pretty amazing. The major thing for me was that my Pre would lock up back when previous kernels allowed scaling down to 250MHz. With F104, not only have I been screen state scaling from 800MHz to 250MHz successfully, I don't have any issues with Music Player skipping at that speed, like has been documented countless times. Every other app seems to be fine with the screen off and processor down to 250MHz as well. I did have moments of slowness when ramping back up or if I had too many apps open (even though I haven't had a TMC error yet), but I've installed JsTop, and with the automatic garbage collection, those issues seemed to have disappeared as well. Great stuff all around.

    My question is, now that UberKernel has compcache in alpha testing, and I assume after the testing, it will go into the stable build, will there be much difference between F104 and that final version of UK?
  4. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #764  
    Are there any real differences from F104 and F104A?
  5.    #765  
    Quote Originally Posted by irateb View Post
    I wasn't sure if user experiences were still needed or being captured for these kernels, but I figured I'd post and then ask a possibly off topic question. I started using F104 about two weeks ago, and it has been pretty amazing. The major thing for me was that my Pre would lock up back when previous kernels allowed scaling down to 250MHz. With F104, not only have I been screen state scaling from 800MHz to 250MHz successfully, I don't have any issues with Music Player skipping at that speed, like has been documented countless times. Every other app seems to be fine with the screen off and processor down to 250MHz as well. I did have moments of slowness when ramping back up or if I had too many apps open (even though I haven't had a TMC error yet), but I've installed JsTop, and with the automatic garbage collection, those issues seemed to have disappeared as well. Great stuff all around.

    My question is, now that UberKernel has compcache in alpha testing, and I assume after the testing, it will go into the stable build, will there be much difference between F104 and that final version of UK?
    There are still numerous differences, such as: F104/F102 is built on linux 2.6.24.7, UK is 2.6.24; UK has all debugging symbols and code built in, F102/F104 does not; F102/F104 was built using aggressive compile options along with a different compiler... etc...

    Some features will end up in UK eventually.
    Live free or DIE!
  6.    #766  
    Quote Originally Posted by angiest View Post
    Are there any real differences from F104 and F104A?
    F104A is not confirmed to exist.
    Live free or DIE!
  7. #767  
    Quote Originally Posted by unixpsycho View Post
    For those n00bs who don't read:

    PLEASE UNINSTALL BEFORE WEBOS UPDATE!!!!

    Remember, you have a test kernel on your phone!

    I take no responsibility for h0s3d and pwn3d Pre's for not uninstalling!!



    Experienced users ignore this post.
    i can't read
  8. #768  
    Quote Originally Posted by marxman.13 View Post
    The latest UK build from the testing feed with screenstate 800/500 and compcache enabled is giving me epic batter life now for some reason.

    After 12 hours including an hour of listening to music, 10mins of NOVA, 30mins of phone calls, about 20 texts and probably 20mins of using apps like facebook and salez and my battery is still 41%. That is insane :O

    Combined with the no lag and seemingly infinite cards and ability to play PDK games whenever this is the best thing that has happen to the Pre since launch.
    Didn't you get a sudden 0% when you reached 30% or so?

    I'm having THAT exactly, "excellent" battery life, but it's misleading (at least in my case, since for some reason it jumps from 30% or 20% or so to 0% suddenly).
  9. sratson's Avatar
    Posts
    57 Posts
    Global Posts
    61 Global Posts
    #769  
    Still loving F102 a couple days in.. battery even seems to be lasting very nicely.. question though.. I originally installed it with WOSQI with the file downloaded from page 1.. it was v 1.5.0.2.. I added unixpsycho.com to my feeds in preware and it shows 1.5.0.3 as an available download, but not as an update to 1.5.0.2.. they seem to have different ID's.. can you explain why? and if I want to update, can you remove the old one or should you install the 'stock kernel' to get rid of it, then install the latest version from the feed? Or am I totally off base?
  10. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #770  
    Quote Originally Posted by knobbysideup View Post
    The thing that I don't get about gentoo: The systems that would benefit from any performance gains by compiling everything specifically for that system are systems you'll wait a month or so for everything to build on.... :-)

    I only wish that Windowmaker would stay up to date to be used with compiz fusion.

    Oh crap. I'm off topic again, aren't I?

    Hey, unixpsycho: I really dig the F102 kernel! (there, that's better!)
    Well, use Sabayon and you get the convenience of a binary dist but the flexibility of Gentoo for customization.

    Over the last couple of months I switched my various Gentoo, Ubuntu and other systems over to Sabayon. Not having to wait on KDE to compile is nice, and it is extremely easy to build my own dovecot and postfix to add in USE flags the maintainer doesn't use but that I do.
  11. #771  
    Quote Originally Posted by sratson View Post
    Still loving F102 a couple days in.. battery even seems to be lasting very nicely.. question though.. I originally installed it with WOSQI with the file downloaded from page 1.. it was v 1.5.0.2.. I added unixpsycho.com to my feeds in preware and it shows 1.5.0.3 as an available download, but not as an update to 1.5.0.2.. they seem to have different ID's.. can you explain why? and if I want to update, can you remove the old one or should you install the 'stock kernel' to get rid of it, then install the latest version from the feed? Or am I totally off base?
    Just keep in mind..Anytime you want to go from one kernel to another.Uninstall + Running Stock Kernel or Webos Doctor = Excellent performance. Does this msg clarifies things for you. Good luck.
  12. sratson's Avatar
    Posts
    57 Posts
    Global Posts
    61 Global Posts
    #772  
    Quote Originally Posted by rush View Post
    Just keep in mind..Anytime you want to go from one kernel to another.Uninstall + Running Stock Kernel or Webos Doctor = Excellent performance. Does this msg clarifies things for you. Good luck.
    Technically this is updating the same kernel.. so I didn't know if the same rules apply..

    BTW.. what changes are from .2 - .3 anyway?
  13.    #773  
    Quote Originally Posted by sratson View Post
    Still loving F102 a couple days in.. battery even seems to be lasting very nicely.. question though.. I originally installed it with WOSQI with the file downloaded from page 1.. it was v 1.5.0.2.. I added unixpsycho.com to my feeds in preware and it shows 1.5.0.3 as an available download, but not as an update to 1.5.0.2.. they seem to have different ID's.. can you explain why? and if I want to update, can you remove the old one or should you install the 'stock kernel' to get rid of it, then install the latest version from the feed? Or am I totally off base?
    Why would you do that? Who told you to do that?
    Live free or DIE!
  14.    #774  
    Quote Originally Posted by sratson View Post
    Technically this is updating the same kernel.. so I didn't know if the same rules apply..

    BTW.. what changes are from .2 - .3 anyway?
    It is *not* the same kernel. Please do not spread false information.
    Live free or DIE!
  15. sratson's Avatar
    Posts
    57 Posts
    Global Posts
    61 Global Posts
    #775  
    Quote Originally Posted by unixpsycho View Post
    Why would you do that? Who told you to do that?
    Uhh.. was I bad? seemed like it made sense..
  16. sratson's Avatar
    Posts
    57 Posts
    Global Posts
    61 Global Posts
    #776  
    Quote Originally Posted by unixpsycho View Post
    It is *not* the same kernel. Please do not spread false information.
    Sorry.. that is why I am asking.. It SEEMED like the same kernel.. same name, same version scheme.. It made sense to me at the time..

    * I suddenly sense a disturbance in the force...*
  17. sratson's Avatar
    Posts
    57 Posts
    Global Posts
    61 Global Posts
    #777  
    Did I crush some toes here or what? I didn't mean it.. I was just trying to keep up on the changes to this great kernel I have been reading about and finally decided to try.. very happily I might add.. The questions seemed relevant.. ***? I was just looking for direction, not a lashing..
  18.    #778  
    Quote Originally Posted by sratson View Post
    Did I crush some toes here or what? I didn't mean it.. I was just trying to keep up on the changes to this great kernel I have been reading about and finally decided to try.. very happily I might add.. The questions seemed relevant.. ***? I was just looking for direction, not a lashing..
    Nope. I just didnt want passer-by users seeing something then thinking its gospel.

    No lashing. It was a curious question as you all of a sudden make some startling revelations and a non-existent feed.

    The kernels have different display names but use the same version #. Also updating from one to the other without uninstalling may not work as intended in the future. Preware handles this gracefully, but WOSQI will not currently. They might have the same ID but have different support files (or will). So anyone seeing that may frown at the end result.
    Live free or DIE!
  19. #779  
    Quote Originally Posted by unixpsycho View Post
    Nope. I just didnt want passer-by users seeing something then thinking its gospel.

    No lashing. It was a curious question as you all of a sudden make some startling revelations and a non-existent feed.

    The kernels have different display names but use the same version #. Also updating from one to the other without uninstalling may not work as intended in the future. Preware handles this gracefully, but WOSQI will not currently. They might have the same ID but have different support files (or will). So anyone seeing that may frown at the end result.

    I still have the F104 kernel "stuck" as installed and not able to make it through the complete removal sequence. I am able to reinstall, but still the same thing happens when trying to remove again. will the repair utility assist with correcting this issue?

    I have already tried to remove, then instal the recovery kernel, but still the same issue.

    I have since been able to put it back on, so I can still overclock the phone, but just not able to remove completely

    concerned,
    SJ
  20. #780  
    @Sjsharksfan, listen you are about to seriously do some damage to your phone. Webos utility won't really fix the problem at hand. You need to completely bring back the phone to factory settings by running webos doctor on the device. (note) Please do not install any kernel(s) on the phone until you have done a webos doc on it.

Posting Permissions