Related
Looking for feedback on CM10.1 vs. sticking with the stock ROM. I've used CM10.1 on my GS3 , but wondering what the advantages are of using it on the N4.
Being that you have used both stock build and cm10 you should be able to judge which works better for you.
Sent from my Optimus G using XDA premium
pros: added features, fixes bugs from the stock ROM
cons: none
Sent from my Nexus 4 using xda premium
The pros are the same as your galaxy. It's the same rom
Sent from my Transformer TF101 using Tapatalk HD
Does the Google voice integration work the same (ie when it takes over the phone and messaging functions completely)
Sorry to jump on this, but I'm considering the same largely for battery and some tweaking options.
Sent from my Nexus 4 using xda app-developers app
emailrob said:
Does the Google voice integration work the same (ie when it takes over the phone and messaging functions completely)
Sorry to jump on this, but I'm considering the same largely for battery and some tweaking options.
Sent from my Nexus 4 using xda app-developers app
Click to expand...
Click to collapse
yes it works the same. you can set it to do that if you choose to, if not u can disable it.
The only con is a slightly slower Android upgrade path. Having said that, 2-3 weeks max (for a Nexus device) isn't that long to wait. Beats all the manufacturers.
Ajfink said:
The only con is a slightly slower Android upgrade path. Having said that, 2-3 weeks max (for a Nexus device) isn't that long to wait. Beats all the manufacturers.
Click to expand...
Click to collapse
Not really. When CM does get updated to the next version of Android, Nexus devices are the first to get it on CM whereas other devices will take longer (and things may not be working day one for other phones whereas Nexus it will be working from day one, just not all the CM goodies will be available right away as it'll take time to port them over)
Alternatively, people may flash non-CM roms and just a AOSP version of the new Android to try out new features etc etc
My biggest consideration is battery , I'm not impressed with my battery hardly making it through a 8 hours of the day.
ksujace said:
My biggest consideration is battery , I'm not impressed with my battery hardly making it through a 8 hours of the day.
Click to expand...
Click to collapse
CM10 will most definitely help you out with that. Some people have been getting weird Android OS or Mediaserver troubles that eat away at their battery with stock 4.2.1. How many hours of screen on time are you getting?
ksujace said:
My biggest consideration is battery , I'm not impressed with my battery hardly making it through a 8 hours of the day.
Click to expand...
Click to collapse
Kernel plays a bigger role in battery conservation more than a ROM. People always gets mixed between the two.
To fix battery problems,
1) Ensure Google Latitude is off (it's in Google Maps). Disable automatic reporting of location and all the other checks in that menu.
2) Flash a ROM that meets your requirements (there is no BEST ROM).
3) Flash a Kernel that meets your requirements based upon feedback and past experiences (there is no BEST KERNEL).
4) Install BetterBatteryStats @ http://forum.xda-developers.com/showthread.php?t=1179809
5) Use BetterBatteryStats to ensure there are no kernel and partial wakelocks. If there are debug them. Ensure that your phone is indeed deep sleeping (use BetterBatteryStats to look at CPUStates or download CPUSpy from Play Store).
I recommend CM10.1 Jellybro + Franco Kernel, but that's based on my own personal opinions. Look around the dev forums and decide what you like the most
Edit: Oh yeah, how could I forget... go into Google Now and disable the cards you don't need.
zephiK said:
Kernel plays a bigger role in battery conservation more than a ROM. People always gets mixed between the two.
To fix battery problems,
1) Ensure Google Latitude is off (it's in Google Maps). Disable automatic reporting of location and all the other checks in that menu.
2) Flash a ROM that meets your requirements (there is no BEST ROM).
3) Flash a Kernel that meets your requirements based upon feedback and past experiences (there is no BEST KERNEL).
4) Install BetterBatteryStats @ http://forum.xda-developers.com/showthread.php?t=1179809
5) Use BetterBatteryStats to ensure there are no kernel and partial wakelocks. If there are debug them. Ensure that your phone is indeed deep sleeping (use BetterBatteryStats to look at CPUStates or download CPUSpy from Play Store).
I recommend CM10.1 Jellybro + Franco Kernel, but that's based on my own personal opinions. Look around the dev forums and decide what you like the most
Edit: Oh yeah, how could I forget... go into Google Now and disable the cards you don't need.
Click to expand...
Click to collapse
+1 for CM10.1 Jellybro + Franco. I've also used MiNCO ROM which is pretty good, but not a ton of features yet so I'm sticking with CM10.1
Installed CM10.1 , seems there is more real estate (5 icons across vs 4?) ...
anyone else had the problem where you can't create app folders on the home screen?
Hey what is CM10.1 JellyBro?
Is it the nightly CM 10,1?
ksujace said:
Installed CM10.1 , seems there is more real estate (5 icons across vs 4?) ...
anyone else had the problem where you can't create app folders on the home screen?
Click to expand...
Click to collapse
I hear that's an issue on the latest nightly. I'm sure they'll address that in the upcoming nightly.
mediumsteak said:
Hey what is CM10.1 JellyBro?
Is it the nightly CM 10,1?
Click to expand...
Click to collapse
It's the same thing as CM10.1 but including cherry picks.
Cherry picks are unmerged features that are currently pending/awaiting approval from CM. They are implemented within Jellybro before official nightlies get them. It is a very stable ROM and uses the same sources as CM. So whatsever in CM10.1 is in Jellybro, just jellybro has extras (very minimal)
http://forum.xda-developers.com/showthread.php?t=2015081 is the thread.
mediumsteak said:
Hey what is CM10.1 JellyBro?
Is it the nightly CM 10,1?
Click to expand...
Click to collapse
Check it out ...
JellyBro link
It's basically CM 10.1 with some cherry picks (additions) that either get merged into CM or they are removed because they become obsolete
dmorand said:
I hear that's an issue on the latest nightly. I'm sure they'll address that in the upcoming nightly.
Click to expand...
Click to collapse
It's fixed in 12/27.
To elaborate on dmorand,
If cherry picks disappear, it's because they were merged, abandoned or conflict with the current source or other cherry picks.
Click to expand...
Click to collapse
not only because they are 'obsolete' (not really obsolete but abandoned)
I heard from my friend that CM isn't as stable as stock but does have the added features. Can anyone confirm about CM's stability as to stock?
mgkrp said:
I heard from my friend that CM isn't as stable as stock but does have the added features. Can anyone confirm about CM's stability as to stock?
Click to expand...
Click to collapse
that's only the case on devices that weren't made to run aosp android, such as HTC (sense), Samsung (touchwiz), motorolla (motoblur). on nexus devices that natively run stock android, cm works id say 99% perfectly. and usually if there is a "bug" its cosmetic and doesn't affect the functionality of the phone one bit, and once reported by a user it gets fixed pretty fast (within the next few nightlies) :beer:
Sent from my Nexus 4 using xda premium
Hey guys,
I`am thinking of ordering a Oppo Find 5 32GB, i have a Nexus 4 atm and i`am looking for something new.
Is it a definate upgrade? The specs on paper seem great but how does the phone do in real life.
Would appreciate some user feedback
Same here. I am thinking of doing the same. I am particularly hoping for a stable PA/CM10.1 ROM soon, and more kernel development if possible.
gee2012 said:
Hey guys,
I`am thinking of ordering a Oppo Find 5 32GB, i have a Nexus 4 atm and i`am looking for something new.
Is it a definate upgrade? The specs on paper seem great but how does the phone do in real life.
Would appreciate some user feedback
Click to expand...
Click to collapse
it's not a goodupgrade..
same CPU, RAM, GPU, same problems with: OTG, no microSD expansion, no removable battery
you get a better/larger display, slightly better(in specs) or worse(in terms of software capabilities so far) camera....
and worse battery life / smoothless/performance in ROMs.....and more expensive
no matter how good the CM10.1 port for the find 5 is, nexus 4 ports will be snappier as they have the backing of google
If ports are based on aosp how can they be snappier .....only a bit of kernel tweaks and some camera enhancements will make this phone awesome
Sent from my Find 5 using Tapatalk 2
munchy_cool said:
If ports are based on aosp how can they be snappier .....only a bit of kernel tweaks and some camera enhancements will make this phone awesome
Sent from my Find 5 using Tapatalk 2
Click to expand...
Click to collapse
i'm talking specifically about the device/kernel tree....
for nexus 4, it comes from google (who has access to all the proprietary documents, binaries)
for oppo find 5, it comes from people like XpLoDWilD, Entropy512, ... . They may have to interpret that info based on how the Oppo Stock Rom works, previous experience, hacks needed to make stuff work...
this portion will contain bugs / inefficiencies in the codebase
and when you move from different android versions, the proprietary binaries differ and may not be compatible. For Nexus devices, google gives you those updated binaries... for other devices (like Oppo), you may have to beg Oppo to get the latest binaries, steal it from a similar device that has the update, or hack your way (which creates inefficiencies / bugs)
see: https://developers.google.com/android/nexus/drivers
this is why you have: stable cm10.1 ports on some devices, but still buggy/slow compared to the nexus line
paperWastage said:
i'm talking specifically about the device/kernel tree....
for nexus 4, it comes from google (who has access to all the proprietary documents, binaries)
for oppo find 5, it comes from people like XpLoDWilD, Entropy512, ... . They may have to interpret that info based on how the Oppo Stock Rom works, previous experience, hacks needed to make stuff work...
this portion will contain bugs / inefficiencies in the codebase
and when you move from different android versions, the proprietary binaries differ and may not be compatible. For Nexus devices, google gives you those updated binaries... for other devices (like Oppo), you may have to beg Oppo to get the latest binaries, steal it from a similar device that has the update, or hack your way (which creates inefficiencies / bugs)
see: https://developers.google.com/android/nexus/drivers
this is why you have: stable cm10.1 ports on some devices, but still buggy/slow compared to the nexus line
Click to expand...
Click to collapse
I agree with you but that's the case with all non-Google devices.
if you want latest android version you gotta stick with Google devices.
munchy_cool said:
I agree with you but that's the case with all non-Google devices.
if you want latest android version you gotta stick with Google devices.
Click to expand...
Click to collapse
i'm talking about the device tree on the current android version.... device tree port by CM developers vs device tree from google
since hardware is mostly the same from nexus 4, but you're moving software-wise from a "optimized" port (with device tree from google) to a Cm10.1 port (with device tree that is still buggy and not optimized by google)..
don't see the reason to switch from nexus 4.
Find 5: aosp -> cm10.1 + device tree from CM developers = ROM
Nexus 4: aosp -> cm10.1 + google's device tree = better ROM
replace cm10.1 with anything else (AOKP, PAC, PA)...
paperWastage said:
i'm talking about the device tree on the current android version.... device tree port by CM developers vs device tree from google
since hardware is mostly the same from nexus 4, but you're moving software-wise from a "optimized" port (with device tree from google) to a Cm10.1 port (with device tree that is still buggy and not optimized by google)..
don't see the reason to switch from nexus 4.
Find 5: aosp -> cm10.1 + device tree from CM developers = ROM
Nexus 4: aosp -> cm10.1 + google's device tree = better ROM
Click to expand...
Click to collapse
the reason to switch from n4 would be better & bigger screen, but in that case I would recommend the DNA
I'm making the switch. Mainly for the screen and I didn't feel like paying $650 for the dev HTC One. It'll be interesting to see how Oppo's firmware is as well.
I switched from a Nexus 4 that I've had since November. I was just getting tired of the device and was looking for something with a slightly larger/higher resolution display, but I wasn't looking for a full on phablet (Note II). I have been incredibly satisfied with the unit's build quality and am only slightly thrown off by it's less than stellar stock rom. No worries though, as PA has been rock solid for me. :good::good: Two thumbs up on making the switch!
I'm also thinking of making the switch and also coming from a nexus 4. What I want to see get more development though is on the android ROMs as there are none. On the original android development I do see cm 10.1,pa,ect.. Which is pretty much the stuff I'm running on the nexus 4 already. I wanna see oppo find 5 u.i based roms.if I'm gonna make the switch and use the same base ROMs as the nexus 4 I might as well stick with the nexus 4 instead.
marcviado said:
I'm also thinking of making the switch and also coming from a nexus 4. What I want to see get more development though is on the android ROMs as there are none. On the original android development I do see cm 10.1,pa,ect.. Which is pretty much the stuff I'm running on the nexus 4 already. I wanna see oppo find 5 u.i based roms.if I'm gonna make the switch and use the same base ROMs as the nexus 4 I might as well stick with the nexus 4 instead.
Click to expand...
Click to collapse
I think it is pretty unlikely that you'll see a lot of custom ROM's for Find 5, considering much lower number of users compared to other phones.
I think i`am gonna stick with the N4 for now and mainly for its optimized firmware. The firmware of the Oppo is clearly not as good (or finished) as the N4`s. Pitty, would have liked to try it out. Aah well, maybe at a later point or its successor. Thanks for the reply`s guys :good:
marcviado said:
I'm also thinking of making the switch and also coming from a nexus 4. What I want to see get more development though is on the android ROMs as there are none. On the original android development I do see cm 10.1,pa,ect.. Which is pretty much the stuff I'm running on the nexus 4 already. I wanna see oppo find 5 u.i based roms.if I'm gonna make the switch and use the same base ROMs as the nexus 4 I might as well stick with the nexus 4 instead.
Click to expand...
Click to collapse
MIUI is the closest "alternative" to Find 5's stock UI rom... find one (probably won't be on XDA) that works on a phone you would want, then buy that phone
next alternative would be the iPhone
N4 owner as well. While I have a hard time putting down my Find 5, if I'm to be totally honest, the N4 is hard to beat. Mostly software I guess, its hard to find a flaw. The Find 5 is still being optimized and tweaked on a regular basis... which is great but not needed in the Nexus. Still though, the Find 5 is one nice device with a lot of potential!
sent from my <Nexus 7>
Hey guys,
I'm looking for a good Kernel with this features available for my Nexus 4.
I have a Nexus 7 2013 and I'm using ElementalX with these features and I just can't get enough of it.
Help a brother out!
I personally use Faux' kernel, which has those features. Whether it's good or not for you depends on your interpretation of "good"
Here are the results in zipped screenshots after running Geekbench on my Samsung Galaxy Tab Pro 8.4 (SM-T320).
First I did a test on the stock TouchWiz ROM (Sep 3rd, 2014 update).
Then I wiped the device and installed CM11 (M11, Oct 2014 release) , nothing else was added.
I had tweaked neither TouchWiz nor CM11 (no governors changing, no trickster mods, no nothing)
As I had empirically observed, in the Sales Force Automation app that I am developing (which is resource intensive), the stock TouchWiz ROM is faster. (I actually benchmarked it on both ROMs, see post 16 bellow)
Single-Core Test: 7.24% (TW is faster than CM11)
Multi-Core Test: 16.11% (TW is faster than CM11)
*** UPDATE for CM11 (M12) ***
When CM11(M12) rolled out I also benchmarked it with Antutu and it casts CM11 as faster than TW by a persistent 3.51% (did the test 5 times in each ROM, always by ~3,5% CM11 is faster)
Benchmarking CM11 (M11) vs CM (M12) the results are the same, not worth posting extra screenshots.
****************************************************************************
Here are some of my thoughts after using both the stock TW ROM and CM11
I found TW better than CM11 on:
-- The Samsung Email app is far superior than the CM11 Email app
I have configured a corporate Exchange account, CM11's Email app downloads attachments as 0 bytes, I have to use 3rd party email such as "Nine - Exchange ActiveSync" or "CloudMagic" (which is cloud based) if I want to download attachments.
-- The Samsung Contacts app is far superior than the CM11 contacts app (which defaults in portrait mode and needs 3rd party app/mod to change it)
-- I cannot transfer files between my PC's Bluetooth dongle and CM11 (but I can and do transfer files via Bluetooth between the PC and the device on the TW ROM, btw files transfers between my phone and the device work both on TW and CM11 ROMs)
-- Battery drains faster on CM11 (though improved), no noticeable difference on M11/M12
-- Samsung Camera is superior than the CM11's camera (though latest Google Camera builds go on par with Samsung's one, so no problem there)
-- I use both WiFi (at home and at work) and Bluetooth (in my car and out of WiFi reach) connection to the Internet. On TW ROM the WiFi connection disables the Bluetooth connection when connectable WiFi is in range while it is the other way around on CM11.
-- Miracast works fine both on TW and CM11, transmitting the device's screen on TV. But additionally when video is playing on the device, TW transmits only the video on TV while leaving the device's screen operable for other tasks while CM11 transmits both video and the screen, as Miracast default behavior.
-- TW has built in DLNA server, running as a service, with the ability to transmit subtitles. In order to achieve that on CM11 I need to download the Pixel Media Server app (the best I found with subtitle support) which is killed by the OS when swiped off the Recent Tasks pane.
I found CM11 better than TW on:
-- No bloatware, debloatted by default (TW debloatting is always incomplete and trouble prone)
-- Uses less RAM, more active applications on the background (though not that many)
-- Much, much smoother that TW (although using Google Now Launcher on TW improved the experience)
-- Far superior Settings on CM11, the Settings are awful on TW ROM, always a pain to find what I am looking for.
-- Far superior notifications, status bar and quick access tiles on CM11
I really, really like CM11. That's the software that our device should have been shipped with, but I found Samsung's TW more rounded, more polished, more h/w compatible, more worked through than CM11.
You are welcomed to share your thoughts.
.
thdervenis said:
Here are the results in zipped screenshots after running Geekbench on my Samsung Galaxy Tab Pro 8.4 (SM-T320).
First I did a test on the stock TouchWiz ROM (Sep 3rd, 2014 update).
Then I wiped the device and installed CM11 (M11, Oct 2014 release) , nothing else was added.
I hadn't tweaked neither TouchWiz nor CM11 (no governors changing, no trickster mods, no nothing)
As I had empirically observed, in the Sales Force Automation app that I am developing (which is resource intensive), the stock TouchWiz ROM is faster.
You are welcomed to share your thoughts.
Click to expand...
Click to collapse
I think it depends on what apps you are running, some are probably better with TW. However, over all I think that cm11 has better overall feeling.
CM is slower ?
CM is worse on this tablet, I have TW because CM and other AOSP builds are having many problems even on official.
I bet a debloated tw would just crush it.
Please if it wasn't for CM I would have gotten rid of both my 8.4 and 10.1 tabs. They work fine with CM. Useless benchmarks...
Sent from my Nexus 5 using Tapatalk
thdervenis, which exact ROM is your TouchWiz ROM anyways? Is it NG1?
Anyways regardless of a 7 to 15% benchmark performance difference... I am still thinking of dumping TW (NG1) and trying CM11.
I'm just not a fan of any of these Samsung "features" or bloat (I don't care for multi-window, S Finder, etc) and I just want the default Android experience I get with a Nexus device.
At least with CM11 I should be able to have more free RAM because none of the Samsung bloat I don't use will be running in the background (right?)
locutusx said:
thdervenis, which exact ROM is your TouchWiz ROM anyways? Is it NG1?
Anyways regardless of a 7 to 15% benchmark performance difference... I am still thinking of dumping TW (NG1) and trying CM11.
I'm just not a fan of any of these Samsung "features" or bloat (I don't care for multi-window, S Finder, etc) and I just want the default Android experience I get with a Nexus device.
At least with CM11 I should be able to have more free RAM because none of the Samsung bloat I don't use will be running in the background (right?)
Click to expand...
Click to collapse
You should try CM11, I definitely recommend it. But be wary of dumping TW for good.
I develop an Android app which is resource and database intensive. It downloads data from remote servers and stores them locally in an Sqlite database. TW db performance is significantly faster than CM11, over 20% (measured repeatedly with my watch, I haven't found a db benchmark app yet)
I like CM11 better than TW, it is smoother, it uses less RAM (TW boots with 1.35GB in use, CM11 boots with 980MB in use), it's transitions are superior to that of TW and I hate the TW's bloatware.
But no bloatware, does not make CM11 lighter on the battery (it drains faster, yet at least since M11 it improved), neither does it make it faster on opening, switching and running apps. I have managed to keep about 6-7 (heavy) active apps active on the background on TW and about 7-8 (heavy) active apps on CM11 due to more RAM available.
I am very fond of CM11, it is the software that our device should have been shipped with in the first place. Android and CM is one of the major reasons why I wouldn't go iOS (I love tweaking, experimenting and I am an avid ROM flasher)
PS: My TW ROM is KOT49H.T320XXU1BNI1 (build Wed Sep 13 :34:57 KST 2014)
I had Slimkat installed on my 8.4 for a while but i'm back to TW, too.
I'm using Apex launcher,greenify, xposed and removed most unneeded system apps.
On Custom Rom, there were to many apps crashing randomly.
I can customize the stock rom via xposed exactly the way i want it to be.
Many xposed customizations don't work on slimkat and aren't available in settings, too.
The hole TW UI is much more optimized on the screen size and resolution.
Of course these are just my opinions, but i really like the tw rom.
greetings
Good to know you are on NI1. I am on NG1 but too many things do not work (i.e. Chromecast)
I think I will wait another 2 weeks for a North American NI1 firmware, if not I'll just go with the British one.
I guess it makes sense that official Samsung firmwares are better performing because they can ask their in-house developers to optimize it for the tablet, whereas CM11 might be missing some optimizations?
Hey OP, how about re-doing the benches using M12 (since you did them using M11) ... M12 could be faster, no?
locutusx said:
Hey OP, how about re-doing the benches using M12 (since you did them using M11) ... M12 could be faster, no?
Click to expand...
Click to collapse
I planned to do it on the weekend (or tomorrow if I find an hour), not because I expect any difference but because I want to try another benchmark tool also (I follow MondrianWiFi's CM11 changelog, I haven't seen any changes that are going to improve performance).
I plan to use and post results from both Antutu and Geekbench on a vanilla setup of both ROMs. I am not going to try the battery duration as I do not plan to use CM11. Any other benchmark tool that I should try?
PS: It's not so much the performance penalty that keeps me from using CM11 (I have found it to be smoother and more tweakable) as the shortcomings that I mentioned above in my post (especially the email, the bluetooth and to a lesser extent the battery)
locutusx said:
Anyways regardless of a 7 to 15% benchmark performance difference... I am still thinking of dumping TW (NG1) and trying CM11.
I'm just not a fan of any of these Samsung "features" or bloat (I don't care for multi-window, S Finder, etc) and I just want the default Android experience I get with a Nexus device.
At least with CM11 I should be able to have more free RAM because none of the Samsung bloat I don't use will be running in the background (right?)
Click to expand...
Click to collapse
I would absolutely recommend to try CM11 if that is how you feel, or alternately SlimKat.
I went to SlimKat and never looked back. Its just a personal preference, and also depends on how you use your device. But I also hate the Samsung bloat apps, did not find anything to add any value (for me) and hated how TW handles the basic UI like Settings, Contacts, etc. SlimKat is so much smoother. I don't think the things being benched by the OP apply to how I use my tablet (I haven't witnessed any performance loss in real life, only smoother interface); nor have I noticed battery life to be significantly affected. Again, a lot is probably dependent on what apps you like to use.
Also remember that manufacturers love to cheat on benchmarks. And Samsung has been shown to cheat on GeekBench specifically by 11 to 16% (pretty close to the deltas found by the OP's benchmark comparison - coincidence or no?). Samsung has supposedly gone legit on benchmarks as of earlier this year. But I don't know if I buy that. Probably all the manufacturers cheat the benches in one way or another (HTC was caught, too). As long as there is a means to compare (increasingly similar) devices; manufacturers will try to leverage that benchmark to make their product look better.
http://www.extremetech.com/computing/177841-samsungs-latest-android-update-no-longer-cheats-on-benchmarks
Interesting comparison, nonetheless and the OPs effort is appreciated.
OP, in terms of additional benchmarking apps, how about trying one of the "Super PI" apps for Android. I use that for quick benchmarking runs when I overclock my desktop, you could try it for your firmware benchmarking as well
There are a few on the Google Play store.
For the record, I try to avoid threads like this like the plague... But you seem to actually be have good intentions.
Your benchmarks really don't show much other than the fact that Samsung "cheats". If you look at your graphs, you'll see that the CM numbers are equal or almost equal[*] to those of the Nexus 5. Given that we have the same hardware as the nexus 5 that's actually what you'd expect. The "cheating" is that Samsung has a power manager service that knows about benchmarking apps and tweaks things in the system to make benchmarking apps benchmark better. That part I know, what they tweak, I do not know. But obvious things are to overclock (unlikely), lock processors at max frequency (likely), let processors run too hot for their own good (likely) and keep processors always online (likely).
You did mention that your app is definitely much faster on TW than CM. If you can provide a test program/app and instructions for running it that you think is doing better in TW than CM, please send me instructions either via PM or via the CM thread in original development. The only way to make things better is to have tests cases so if you can share, it helps.
[*] We are slower on the graphics tests because we have a larger display.
Sqlite db operations: receive and write 20,000 records in a batch
Triggered by crpalmer's comments I 've compiled the following benchmarks of the application that I participate as a developer. I 've compiled them after hours at work on my personal tablet (SM-T320).
The application I am developing is an ERP & CRM companion app for my company's BackOffice ERP & CRM application systems. It downloads packets of JSON formatted data (usually 2,000 records a batch) and stores them locally in an Sqlite db. The data downloaded can easily exceed 300,000 records in over 150 Sqlite tables.
In order to achieve "optimal" conditions for benchmarking I did the following
The benchmarks were performed after hours, when the company's personnel had left and network resources were unused (I performed them on our Intranet).
The company's SQL Servers that provide the data were not used by any other but me.
I configured the app to download just 1 table (not 150+ tables) so as to minimize network calls per table to download the data.
I configured the app to download 20,000 records in a batch (not 10 calls of 2,000 records so as to minimize network calls per table) and insert them in Sqlite.
It took me about 1.5 hours, flashing and reflashing, performing the tests. I did the test for only 20,000 in a batch because if I had done it for the whole dataset it would take me about 30 minutes a try.
As you can see bellow TouchWiz outperforms CM11 consistently by 10 seconds for a 2 min to 3 min operation. Keep in mind that the actual business operation of the app is to do multiple calls over the network downloading and inserting 2,000 to 2,500 records in a batch, a full fetch cycle includes over 120 network/Sqlite batch insert calls. In a full process cycle this may take 30 mins, TouchWiz outperfoms CM11 by about a noticeable 5 - 7 minutes (~17% faster). I 've have stamped up on the delay consistently over the past 4 months that I own the tablet and pin pointed it on Sqlite operations, not on network conditions.
TouchWiz (KOT49H.T320XXU1BNI1) (clean flashed, no accounts, no updates, no apps, only WiFi setted)
1st try: 2 mins and 48 secs
2nd try: 2 mins and 31 secs (no app reopen)
3rd try: 2 mins and 31 secs (no app reopen)
4th try: 2 mins and 30 secs (no app reopen)
5th try: 2 mins and 30 secs (no app reopen)
CM11 (M12) (clean flashed, no accounts, no apps, only WiFi setted)
1st try: 3 minutes and 2 seconds
2nd try: 2 minutes and 42 seconds (no app reopen)
3rd try: 2 minutes and 41 seconds (no app reopen)
4th try: 2 minutes and 41 seconds (no app reopen)
5th try: 2 minutes and 43 seconds (no app reopen)
Bellow are the instructions on how to use it and get a glimpse of what I did.
Sorry crpalmer, I cannot give you a version of the app to perform the test I did, I would have to go to the legal department of my company to get clearance, I don't want to do that for my personal tablet. Also the "demo" user that I am posting bellow (it is the publicly available login that we encourage our customers to use so as to demonstrate our app) downloads a small set of data (not actual full work load data), usually done in 2 to 3 mins, so the difference between the two ROMs may not be noticeable or depended on network conditions.
Here is the link to the app on Google Play http://goo.gl/JTd3SR . Make sure your tablet in in English (or if it's in another language it will default to Greek)
Login using "demo" as both username and password.
Press the first option labeled "Company Selection"
Let it finish until it shows a toast "Work area changed ..."
Press back and press back again.
It will start downloading data. (it 'll take a 2 to 3 mins)
Press the "Refresh" button on the main screen and select the 3rd option labeled "Get All" and press ok. I did this final step five times in order to benchmark it..
I am not passing a judgement on CM11 or TouchWiz, just my personal experience on my device on the General Discussion section, hoping someone will also share their expreriences. I know that full professional benchmarking is not what I did.
crpalmer said:
The "cheating" is that Samsung has a power manager service that knows about benchmarking apps and tweaks things in the system to make benchmarking apps benchmark better. That part I know, what they tweak, I do not know. But obvious things are to overclock (unlikely), lock processors at max frequency (likely), let processors run too hot for their own good (likely) and keep processors always online (likely).
Click to expand...
Click to collapse
From what I've read, tweaking the thermal profile and maxing out the CPU clock during benchmarks is what they were doing when they got busted earlier this year (before supposedly coming clean with 4.4.4).
As much as I appreciate CM, I've always found it inferior to the stock ROM on tablets. Especially when compared to a de-bloatified TouchWiz. (Why oh why does Samsung hamstring their devices in that way?)
I haven't rooted my GTab Pro 8.4 yet (I'm running stock w/Nova Launcher Prime) but when I get some free time I'll debloat it... as fast as it is now, I'd love to see how it runs lean and mean.
TW, CM11 (M12) and conclusion
I updated my first post with benchmarks I did using Antutu, it's *** UPDATE for CM11 (M12) ***
The results appear to be inconclusive:
Geekbench casts TW as noticeably faster (~16%)
Antutu casts CM11 as somewhat faster (~3.5%)
Plus my own, non professional benchmark, that casts TW as faster (on network calls followed by db insertions)
A 3rd, a 4th and a 5th professional benchmark tool might also cast different and inconclusive results.
From a consumer point of view CM11 beats down laggy TouchWiz.
So I guess it is down to if Samsung stock software gives you reasons enough to stay with TW or get rid of it and go with CM11.
My thoughts on the experience is in the 1st post at the beginning of the thread.
My CM works super excellent and much better than TW... That is very strange
Code:
/*
* Your warranty is now void.
*
* I am not responsible for bricked devices, dead SD cards,
* thermonuclear war, or you getting fired because the alarm app failed. Please
* do some research if you have any concerns about features included in this kernel
* before flashing it! YOU are choosing to make these modifications, and if
* you point the finger at me for messing up your device, I will laugh at you.
*
*/
Join our Telegram group for faster updates and news!
Check these kernels Popcorn Kernel or Stratoshphere kernel aswell
Enigma has been and always will be a FREE PROJECT.
Introducing Enigma Kernel for Nokia sdm660 platform. Born as a hobby and because of the intense desire for me to create a kernel with what I deem useful and what I want. I will not claim for you to have god-like battery life or legendary speed, But I can surely tell you'll have a fantastic and stable experience.
Use either Franco Kernel Manager or SmartPack Kernel Manager to tweak. If you find a sweet spot that you'd share with the community, please feel free to do so
Feature requests are welcome and encouraged but please refrain from silly requests such as dev increase pubg performance please!!!! .
For more information and F.A.Q, Please check the second post.
Features -
System
- Removed HMP
- Energy Aware Scheduling(EAS)
- Scheduler modified heavily to support EAS
- Scheduler improved
- Improved Schedutil governor
- Removed Unused CPU Governors
- Load Balancing across cores improved
Hardware
- GPU Tweaks
Drivers
- Qcacld-3.0 WiFi drivers from CAF Android 10 tag(LA.UM.8.2.1.r1-04300-sdm660.0)
- Tone down WiFi driver debugging
Memory Management
- Uses LZ4 algorithm as the default ZRAM compressor
- Improved efficiency
- Misc tweaks
Miscellaneous
- Kernel size reduced
- Compiled with Google's Clang v9.0.3
- CFQ I/O Scheduler improved
- Magisk retain present thanks to AnyKernel3. No need of flashing Magisk after every update.
- Improved system responsiveness
- Improved Fluidity
- Maybe even slightly improved battery life
Installation Instructions
- Download Kramel .zip from Github
- Store it on any storage you have on your phone
- Fastboot TWRP
- Flash via TWRP
- Reboot
- Profit!
-OR-
- Download Kramel .zip from Github
- Store it on any storage you have on your phone
- Open Franco Kernel Manager and navigate to Flasher section
- Select the kernel from storage
- Tap on reboot after flashing(Don't worry, Magisk will be preserved)
- Wait for the flashing process to complete and reboot
- Profit!
FOR FULL CHANGELOGS, PLEASE REFER TO THE RELEASES SECTION ON GITHUB.
Compatible with -
> Nokia 6.1 (PL2)
> Nokia 6.1 Plus/X6 (DRG)
> Nokia 7 Plus (B2N)
> Nokia 7.1 (CTL)
Downloads - Enigma Git Releases
Old Enigma Downloads - Enigma Git Old Releases
Kernel source - Enigma Git
Github Commits Advanced Changelog - Here
Have fun and enjoy Enigma!
Feel free to post suggestions and feedback below.
XDA:DevDB Information
Enigma Kernel for Nokia SDM660, Kernel for the Nokia 6.1, 6.1+ and 7+
Contributors
CarbonGTR
Special Thanks
@paul89rulez - For being my inspiration and encouraging me.
@theimpulson - For the LOS kernel sauce on which the early Enigmas were based on.
@Sahil_Sonar - For sharing his server and encouragement. Enigma wouldn't be here if it weren't for him and Paul.
@Sid127 - Mad tester, Great Friend.
@Nikhilkumar038- Goodest Boi. Encouragement.
All the devs whom I cherry-picked from.
Version Information
Status: Stable
Current Stable Version: Back2Basics
Stable Release Date: 2021-11-20
Created 2020-07-09
Last Updated 2021-11-20
F.A.Q
1. How do I install this kernel? What Do I gain from it?
- Flash Via TWRP or Use FK Kernel Manger's Flasher option. I will also soon provide users with a .json file that you can import into FK Manager to get updates directly!
As I said in the first post, I will not claim extreme speed or hyper Battery life. But you will have a stable, fluid and satisfying experience on your phone.
2. Requirements -
- A unlocked Nokia 6.1, 6.1 Plus or 7 Plus
- A PC with fastboot working
- Stock or custom ROM that boots
- A recovery or FK Kernel Manager
3. Device Heats up alot. Can you do something?
- If your phone heats up, then you should probably consider not playing PUBG: Mobile for 5 hours on a hot day or use LTE continuously.
4. PUBG: Mobile and CODM Lags like crazy. Phone heats and up Enigma kernel dev suck!!!!
- No. I don't suck.
Heavy games utilize every last bit of resource on your phone and your phone has to keep up with all that rendering, sound processing and AI/NPC Controls while keeping a check on your BG Apps. This heavy utilization requires the CPU, Memory Controller and GPU to run at their 100% and that believe it or not produces massive amounts of heat and use a lot of battery power.
I can increase thermal throtlling thresholds but am I gonna? Nope. It is stupid and dangerous.
5. Does GCam Work fine?
- Yes since I replaced the camera drivers from Android 9 sources, Gcam mods will not be affected from shutter button not working or similar. You just have to find a stable Gcam mod for yourself and you're done
6. What about Battery Life, Performance, Fluidity - All the good stuff?
- I put a lot of emphasis on fluidity vs battery or performance because it is just terrible to use a phone that lags, stutters and cannot hold a stable 60FPS throughout the UI. These Nokia bois are perfectly capable of stable 60FPS UI rendering. Next to fluidity, I focus on performance and battery equally.
What this means for end users is that, You have a smooth, lag-free and fluid experience with balanced performance and battery life across the board.
7. HMP vs EAS? Why EAS?
- They are just different ways via which the Kernel chooses which CPU to use for the different tasks you perform on a daily basis. HMP is almost deprecated at this point and all the new phones releasing with Kernel tag 4.9 and higher come with EAS by default.
EAS is smoother and with proper configuration, more performant and efficient as-well.
8. I want to create a kernel based on Enigma's sources. Where should I start?
- Nice! You could head over to Enigma's Git repository and fork it. I'd advice you to base your work over the STAGING branch as that's the cleanest and the most stable base of Enigma.
I'd consider Enigma's sources stable as on 29-09-2021.
For in-line ROM compiling, other devs have run into issues and as such, I WOULD NOT recommend you to build Enigma in-line. I will not support you in those cases. Standalone compilers however, would have my full support.
9. Donations and what not?
- You don't need to spend more money to bring out more out of your phone. This is a hobby, not a business. Let's keep it that way
-------------------------------------------------
For More Information on EAS -
- How EAS makes Google Pixel the smoothest phone
- How EAS makes Google Pixel the smoothest phone: Reddit Discussion - This is quite a technical dive, But is quite interesting and absolutely wonderous to see top devs from XDA and other blogs have arguments.
- Kirisakura Kernel for HTC U11: General Explanation on what EAS is.
- RenderZenith Kernel for OnePlus 3: Comparison between EAS and HMP and battery life and CPU usage is affected.
- RenderZenith Kernel for OnePlus 3: Difference between the different load tracking algorithms EAS uses(PELT vs WALT).
Pro af legendary xD
Will it work with Nokia 7.2
SHLD I try
happy3777 said:
Will it work with Nokia 7.2
SHLD I try
Click to expand...
Click to collapse
Nokia 7.2 is unfortunately not supported. They have a different code base than drg, b2n and pl2.
Just a quick update - Enigma now has an working EAS variant. Expect release soon. EAS is now released. HMP is discontinued for now.
Grt work man...
new update?
Tsuragi said:
new update?
Click to expand...
Click to collapse
Check here - Enigma Releases Git
is it possible to add support for usb wlan ralink 3070?
and.. nokia 6.1 is compatible with usb ethernet adapters? i mean, support for that would be great
Tsuragi said:
is it possible to add support for usb wlan ralink 3070?
and.. nokia 6.1 is compatible with usb ethernet adapters? i mean, support for that would be great
Click to expand...
Click to collapse
Lemme look
Edit - Looked into it, Seems like I will not be able add it via kernel
Hello, do you think you could look into why the minimum brightness in novatek screens, for 7 plus, is so bright?
I tried to mess with it using i2cset but I can't control it because it's being used by the driver
dicamarques said:
Hello, do you think you could look into why the minimum brightness in novatek screens, for 7 plus, is so bright?
I tried to mess with it using i2cset but I can't control it because it's being used by the driver
Click to expand...
Click to collapse
Brightness is usually affected in the vendor side of things. Kernel just hosts the touch drivers to power on the display and recognize touch inputs. I can suggest you to switch back to the stock rom/flash stock vendor if you feel brightness isn't enough.
I do not usually play on the phone but the battery lasts much much longer, in daily use (web browsing and messaging) the performance is perfect it is fluid and energetically very very efficient. your kernel it's great
Tsuragi said:
I do not usually play on the phone but the battery lasts much much longer, in daily use (web browsing and messaging) the performance is perfect it is fluid and energetically very very efficient. your kernel it's great
Click to expand...
Click to collapse
Really appreciate the feedback! Quite glad that you're enjoying Enigma!
If you don't mind, Can you tell me which phone do you own?
CarbonGTR said:
Brightness is usually affected in the vendor side of things. Kernel just hosts the touch drivers to power on the display and recognize touch inputs. I can suggest you to switch back to the stock rom/flash stock vendor if you feel brightness isn't enough.
Click to expand...
Click to collapse
I want to lower it xD but i think its in the kernel side. I link you the file I think needs some changing.
https://github.com/archie9211/android_kernel_nokia_B2N/issues/1
CarbonGTR said:
Really appreciate the feedback! Quite glad that you're enjoying Enigma!
If you don't mind, Can you tell me which phone do you own?
Click to expand...
Click to collapse
i have nokia 6.1 (pl2)
dicamarques said:
I want to lower it xD but i think its in the kernel side. I link you the file I think needs some changing.
https://github.com/archie9211/android_kernel_nokia_B2N/issues/1
Click to expand...
Click to collapse
If it's there, It's probably there for a reason or it's a bug. But seeing the amount of display issues with 7 Plus i don't it's a bug.
Anywho, Lemme see what I can do.
I have some screenshots of the full daily battery usage, I am in the pixel experience with Enigma-WAS 08-02-2020 for nokia 6.1
I hope they serve as a reference for your good work :good:
https://imgur.com/a/1CVKaWw
Tsuragi said:
I have some screenshots of the full daily battery usage, I am in the pixel experience with Enigma-WAS 08-02-2020 for nokia 6.1
I hope they serve as a reference for your good work :good:
https://imgur.com/a/1CVKaWw
Click to expand...
Click to collapse
They are pretty good for this phone!