Mourta

Members
  • Content count

    523
  • Joined

  • Last visited

Community Reputation

659 Excellent

About Mourta

  • Rank
    Addict
  • Birthday 08/02/70

Profile Information

  • Location
    Stockholm

Previous Fields

  • Your Current Device(s)
    LG Optimus 4x HD

Recent Profile Visitors

1932 profile views
  1. New kernel release. Because of recent events the upgrade to 3.10 has been held back but i'll publish the code on my github before i leave, i'm making a WIP branch, I'm halfway through the upgrade to the new suspend/resume code that we need to implement the rest of 3.10 and fully updated for most parts apart from those dependent on that. Support your troops.
  2. So i got a bit curious on the subject and reviewed the code and what i can say is that the new permission system seems to be a very bad idea, it restricts how you are allowed to store things on your phone and it's not really protecting anything since you can just as easily push to data/media if you want to backdoor into someones phone, there is absolutely no addition of security that i could find. When it comes to the power optimizations it's basically faster art and that part i will be backporting to KK anyway, it's too easy not to do it because it is better code and it uses less memory so that is a given, my KK won't be a standard ROM, it will be a collection of parts i and others like along with some upgraded parts, NothingS suggested the new ART being used half a year ago for KK so that is a given. The other power optimizations are on stock rom Kernels using a a new form of min-max configuration on 8974 which has fixed regulators, i implemented that change two weeks after i released the kernel. To make full use of it i'm shipping the next iteration of the kernel with intellimm. Note that fixed regulators can still be trimmed to scale via EDP though so the testing that is ongoing is whether that is a good idea or not. When upgrading to the new versions of hotplug, platsmp, all timers, smp and EDP i found that the new code makes a hell of a difference but that we are not really using EDP in our kernel and it may be a better idea to leave that for the future. So far i've got 3.10 to boot, display video and things like radio and ... well everything works except for deep sleep, the thing is that i've just resolved that by backporting the vsync code from 3.10 to 3.4, all it really takes is actually using the lvds2dsi and just set our GPIO PIN operations to lvds, our regulators support this as well and so does vddio for both our display and hdmi. That way we can get rid of the age old ssd2825, adapt our panel file to the last known stable for T30 and apply all updates. The thing is 3.10 is very special because at that point android moved from specific to standard Linux, this means we can use any patchset on any kernel after 3.10 but we will have to implement code to use new functions.
  3. Neither are interesting in any way and the constant swapping removes any form of power optimization even with the new memory handling, we're running compression algorithms as soon as we start using the phone. The new "optimizations" are like the "memory optimizations" for KitKat, it's kernel side improvements. Compared to LP it's a step forward but both LP and M have to be hacked together to even work. Of course, if people want to use it, it's up to them. I'll make one version and Nothing will take it from there. However, i won't spend more time on it than that, it's not something i'll be using so there is no reason to develop for it, besides, i have a LOT of code on stand still on the kernel side because of this crap.
  4. New testing kernel has been sent to testing crew. We're now fully implementing sync, it shouldn't take much time to pass this since i've been running it myself for three days without issues. Regarding Android M, i'm not going to put much focus on it because it doesn't work properly and i haven't found a single feature that i would really want with it. I'm going back to KK while helping NothingSerious develop his M version if he wants that, i've spent too much time fixing crap to make it just about as horrible as LP has been. I wouldn't run M on a phone without at least 2GB of memory. Expect a KK release with new kernel that has full system wide sync including the 3.10 host sync.
  5. KK is going to be released soon and M will be released in another thread, look for it. With new code we got the best of both worlds, all stock where it's needed and all performance on everything else. NothingSerious will probably release his own version (since i think it's fairly futile to keep upgrading something that is no more functional than 4.4 and isn't at all suitable for our device) and i will probably release one of M too just to show proof of concept for him and other devs. For me it will probably be Vanir with some additions for KK and VanirAOSP for M since that is what i like at the moment, i will provide exact instructions on how to compile this though so whatever ROM you want should be available, the main thing is new video which won't work with most ROM's yet.
  6. Everyone hates Java. I'm going to ask the others what they think but the final decision is mine. Mostly needed is basic fixing of some code, i've fixed most of the power code up to work with 3.10 but there are still issues that i'm noting, i could really use someone well versed in C to catch my slack on these issues and it's novel programming bringing old code up to date, as i figure it we need to move to a new stardard for video to make this work and i've been using LVSD instead of RGB controls for DC with some success, it doesn't work perfectly and needs adjustments in specific code for our phone. So if you are willing to take a shot at improving this code to make it more suitable as it is at this point a general driver that works but doesn't work perfectly with our device then i will give you access to the hidden repos. I'll get back to you in PM.
  7. It's abandoned and no new development will be made. If you are looking for a 4.4 that works then i have one coming, cherrypicked and ready for prime time with an assembly of code from several different projects and a shiny new 3.10 kernel.
  8. I have a lot of things going on, deaths, birthdays and someone in the hospital. I'll fix it when i have the time or someone else will, the code is published (apart from 3.10 because i can't reach my server from here) and there is a lot of new stuff, as soon as i can send it to bitbucket someone will publish it and you'll have an entirely new set of issues to deal with. I'm really badword done right now, i'll be back but at this point i'm going to be with family and friends. This isn't my job, i have a great uptime on my ROM's, better than any other including stock which should count for something but there will be issues. Now i can fix those when i have the time but when i have the time i'm sending 3.10 and a standard patch (which should work on any version up to 3.19 without much editing, i removed all irrelevant code that isn't even used by our phone, that alone took over 150 hours) and a working tree for M upstream for others to handle. Nothing will handle that stuff and i'll do the kernel work with input from Giuseppe and return to 4.4 which is the last known good but it's going to be with a twist, i grew tired of the idiotic fixes so i rewrote stagefright instead and it can be patched up and compiled as is with the latest drivers for Nvidia OpenGL. It's going to be quirky for a while, this is the new s***.
  9. Kernel or ROM, C, C++ or Java? Please say Java.
  10. That is for the ROM and i've reviewed that project, i'll happily support it in anyway i can and i ask everyone else to do the same. It's a good project and a needed one since google has gone the way of including actual spyware in their code as of late.
  11. I've been deployed but i will ask the testing team to release the last version. More to come, i was nearly done upgrading to 3.10 so be patient. It has some really cool stuff, mainly the ability to track corrent.
  12. I... i can't even explain why that works, it's a glitch but what works works. OTG stays, your method will go into the OP. I'll be home on wednesday, expect a new kernel then. It's past testing, that was the last complaint and it's been a problem ever since i implemented the OTG code again. Oh and Juantech, i forgot to send you your kernel for testing, i'll do that too on wednesday, i haven't had the opportunity to test HDMI output but it solved another problem with the Chrome stick so either way, i fixed things. I do think that the new regulator and it's GPIO descriptors will work with HDMI, if it doesn't, i'll have to rewrite the si2k drivers but that won't be a priority. It really should work now though. I realized that it's calling an output that doesn't exist in 3.4 (which is in part why this kernel doesn't play well with other ROMs) but i made room and implemented it.
  13. Yes, when connected to a usb interface you can pull down and click the "connected as media storage" and change it to charging only. This is also a good thing for those with old media connect USB units because i adapted the USB code to work either as an external usb plug or a USB HDD, this works through the media connector trying to connect so if it's not successful, don't pull it out, just let it be plugged in and then try again. I also have a new kernel but there is a question here, do we want OTG or do we want off line charging to work, because at this point i'm going to France and i don't have any time to do much more than a recompilation, or in short, i don't have the time to fix it so we can have both.
  14. Send the log and i'll decide that.

MoDaCo is part of the MoDaCo.network, © Paul O'Brien 2002-2016. MoDaCo uses IntelliTxt technology.