[Q] loop-img vs. discrete ext2-part - Touch Pro, Fuze Android Development

Can someone summarize the pros and cons of looped system/data.img partitions, like prefered in XDAndroid FX0x vs. discrete ext2 patitions like prefered in NEOPeek-builds in terms of speed, battery-drain, stability, etc.

rer1 said:
Can someone summarize the pros and cons of looped system/data.img partitions, like prefered in XDAndroid FX0x vs. discrete ext2 patitions like prefered in NEOPeek-builds in terms of speed, battery-drain, stability, etc.
Click to expand...
Click to collapse
Loop pros:
Easy to install
No need for partitions
You can have builds as much as your space
Loop cons:
Editing is a bit harder
Slower
Making them bigger/shorter is harder and could make problems
Partition pros:
Faster
Editing is easier
Resizing is easier
USB Mass storage
Partition cons:
You can have as much builds as you want,but you'd need to hack init
Partitions are needed so first setup is harder

I seriously question the faster/slower arguments.
I say try 'em all, see what works best for you. That's what I do .
IMHO the biggest bonus from having dedicated partitions is USB mass storage is possible. Other than that, I didn't notice any significant benefits.

Thanks for your contributions!
Maybe I use the wrong rootfs, but with looped-parts I always realized fs-corruptions (e2fsck) after a restart, that I don't have with dedicated partitions. So in terms of stability and data-loss prevention, dedicated fs seem to be the better solution, I also realized a slightly lower battery-drainage (15%) with dedicated partitions, but as long as the panel-collapse-problem exists thats a minor issue. In desktop-linux systems loop-devs tend to be slower than dedicated parts, don't know if thats true for mobile-devices

Related

do we really need so many competing roms?

do we really need so many competing roms? it;s hard to choose between the many 2.1 roms for my hero
steelbytes said:
do we really need so many competing roms? it;s hard to choose between the many 2.1 roms for my hero
Click to expand...
Click to collapse
Yeah, we do. The ROM devs are in most cases developing the ROMs to suit their own needs for a ROM, but also so kind that they share them with others, while other devs are doing it to learn about android, development etc. If you have problems choosing which ROM to use, stay with the stock - otherwise flash, experiment and have fun!
As a general rule, the more chefs you have, the better the overall quality of all the ROMs.
Whilst it is "logical" to try to have several chefs work on a single ROM, the reality of the situation is that tight collaboration is very difficult when you are talking about something that is generally only developed in peoples spare time.
Regards,
Dave
It is a bit annoying to see 6 replicas of one rom, just with a few different apps chucked in. That's the nature of Android development these days though. I think it was Cyanogen that suggested that a proficiency with Winzip makes you a 'developer' these days...
Personally, I don't see it as a competition. If I could help out Benhaam or Lox in any way with one of their roms, I would.
To be totally truthful, pick ANY of the current Eris builds, and the chances are there isn't much of a difference between them. It's better to pick a rom with an active dev and following in my opinion.
don't worry i have flashed happily, as I can't live with stock 1.5 (come on htc, please release 2.1 for the hero). just felt like whinging after studying this forum to try and guess which rom would best suit me and finding not much clear info.
am running neo 1.3 (2.1+sense) on my hero, and stock 2.1 on my nexusone. very interesting to compare and contrast the two - unsure if I prefer sense or not, but do like (need) the htc exchange client having calendar support (hey googgle, why doesn't stock android have this?)
as i tested all Eris rom here villain3.3 is most stable but i still revert back to SenseHero 1.6 which more smoother run & it is an ART WORK! Hope the offcial 2.1 for hero will out soon ='(
ahh finally i can vent my frustration. its not the amount of roms that are available that bothers me as the devs on here are quite good but the way they dont seem to work together nor communicate. i may be rom. all the devs on here are busy working with this bugged up Eris dump and everything else has been neglected. All the Eris roms are more or less the same and all more or less offer the same positives and negatives. Vanillian which is a fantastic rom has all been neglected so all we have to flash apart from Ahero 0.52 are all the same thing. it doesnt take long to flash all the 2.1 Eris dumps roms on offer before you find yourself running back to MCR or Sensehero due to that bugged up dump. it would be nice to see some form of variety in this forum and not every chef offering 2.1 Eris roms which are far from complete. each dev adds something different to their roms but alas they do not make that mucvh different to the end product. come on guys not all of us want to use 2.1 Eris dumps, we also like the vanilla options too. soon when Legend dump is available we will have similar 6 roms with Legend roms. there is not much vareity in this forum and it be nice to see more vareity on here. i have flashed them all, yep all the bugged up Eris ports and always head back to 1.5 for various reasons...........however Ahero 0.52 is there to save the day and offer a different alternate solution.
please dont flame me for expressing my opinion. feel free to express yours as i have done mine.
shingers5 said:
its not the amount of roms that are available that bothers me as the devs on here are quite good but the way they dont seem to work together nor communicate.
Click to expand...
Click to collapse
Actually, the devs communicate quite well via Twitter. If I'm not mistaken, Lox and Behnaam likes to communicate via Gtalk (from what I followed).
steelbytes said:
do we really need so many competing roms? it;s hard to choose between the many 2.1 roms for my hero
Click to expand...
Click to collapse
No, we don't. And we shouldn't. Why?
Look at the Linux distribution arena (and not only).
They're a lot, but none is really viable on everyday desktop/laptop use.
The reason is simple: fragmentation.
Fragmentation of (scarce) human resources.
Fragmentation of (even more scarce) economic resources.
Fragmentation of the (ever growing) user base.
The same goes for the graphical desktop environments (KDE, GNOME, Xfce to name a few).
On the opposite side we have the internet browser and the *BSD OSes.
We actually have much fewer actors, with stabler products, better engineering.
What I'd like to see in the Android arena is a very stable and effective software base on top of which a relatively large number (I'd say about 5 and less than 10) of UI tweakers and developers push the user experience to the max.
But, again, a single high quality software base (that is the kernel, the system libraries and so on).
This's my EUR 0.005 contribution.
Linux technically is not the OS Linux it is the kernel, its not fragmented at all, as the developers submit the changes and its decided by a set group, the same as any closed dev team would do but they are at different world locations.
Because of how BSD is closed off it takes 3x longer for new software to be implemented..and driver are a nightmare
If the rom dev sub foum was not so cluttered with things that don't belong in it like this thread that should really be in the Q&A section, it would be easier to keep track of the roms..
my 1p woth..
anarchyuk said:
Linux technically is not the OS Linux it is the kernel
Click to expand...
Click to collapse
Linux is the kernel. And GNU/LInux is the (basic) OS with all the GNU stuff.
So I can say that 99% of Linux kernel based distros are Linux.
anarchyuk said:
its not fragmented at all, as the developers submit the changes and its decided by a set group, the same as any closed dev team would do but they are at different world locations.
Click to expand...
Click to collapse
It is fragmented. Take an application at random, let's say Openoffice or GIMP.
You cannot just download ONE aplication and install it on any distro. That's not because of the packaging medium (RPM, DEB, TGZ etc.) but because of software and library dependecies and other system dependent choices.
If you choose, let's say, Ubuntu 9.10 and the package (version) you need is not available for it, then it's up to you to invest in time in makeing an hand made installation ...
anarchyuk said:
Because of how BSD is closed off it takes 3x longer for new software to be implemented..and driver are a nightmare
Click to expand...
Click to collapse
That's because the human resources are scarse, not because of the "closed" model. And then you have a choice among super-distributions on top of FreeBSD, though.
anarchyuk said:
If the rom dev sub foum was not so cluttered with things that don't belong in it like this thread that should really be in the Q&A section, it would be easier to keep track of the roms..
my 1p woth..
Click to expand...
Click to collapse
I can count not less than half a dozen of 2.1 based ROMs all with very minor differences each other. And the same goes for the 1.5.
I cannot swear on it, but I think that the biggest gap is in less than 5% of the whole system.
EUR 0.005 ~ GBP 0.01
Uqbar said:
I can count not less than half a dozen of 2.1 based ROMs all with very minor differences each other. And the same goes for the 1.5.
I cannot swear on it, but I think that the biggest gap is in less than 5% of the whole system.
EUR 0.005 ~ GBP 0.01
Click to expand...
Click to collapse
5%? How did you come to such a precise conclusion?
Do share your calculations with us.
I guess you didn't take into account some of the AOSP based ROMs.
I do agree; there's A LOT of cloning; a guy taking someone's ROM, repackaging with some "customization", and calling it his own work.
This is one of the reasons I totally left the WinMo devving, but I knew it was a matter of time till the same happened to the Android scene.
That's sad, but true.
I don't even care about some of the guys calling themselves "devs", just because they "released" a few roms, who don't even give credit to the people, thanks to whom they learnt/started.
As long as they get enough attention to their thread, they no longer care about the aformentioned. What for? ... It's like a machine which just turns itself after a while.
The more ROMS we have the better, the thread title is a backwards title.
No calculation. Simply I compare performances and features from 3.3 to 3.4.
3.4 is no more stable or fast than 3.4, to me at least.
Memory footprint is almost the same. Not FC in aither ROM ...
And I bet that if nprussel brought anything bigger than 5% he would have put it into the annoucement.
Instead it shows little more than new application list and a GUI theme.
Uqbar said:
No calculation. Simply I compare performances and features from 3.3 to 3.4.
3.4 is no more stable or fast than 3.4, to me at least.
Memory footprint is almost the same. Not FC in aither ROM ...
And I bet that if nprussel brought anything bigger than 5% he would have put it into the annoucement.
Instead it shows little more than new application list and a GUI theme.
Click to expand...
Click to collapse
My point about the fragmentation was in relation to the kernel not a minor thing like dependency's that really has no relation to fragmentation, The Distribution is called a Linux based distribution not "Linux" in an entirety!
Anyway... I know what the above posters are getting at when it comes to people making a minor change to other peoples dumps and claiming to be dev's when really they are nothing more than cooks.
No to take away any credit for what the produce but if you look at the g1 forums it seems a mass of good work, but very little low level development actually happens on the hero side.
adwinp said:
5%? How did you come to such a precise conclusion?
Do share your calculations with us.
I guess you didn't take into account some of the AOSP based ROMs.
I do agree; there's A LOT of cloning; a guy taking someone's ROM, repackaging with some "customization", and calling it his own work.
This is one of the reasons I totally left the WinMo devving, but I knew it was a matter of time till the same happened to the Android scene.
That's sad, but true.
I don't even care about some of the guys calling themselves "devs", just because they "released" a few roms, who don't even give credit to the people, thanks to whom they learnt/started.
As long as they get enough attention to their thread, they no longer care about the aformentioned. What for? ... It's like a machine which just turns itself after a while.
Click to expand...
Click to collapse
Hey man.. Am I missing something or aren't you cooking anymore?
OT: i totally agree by what you say about releasing a rom that just doesn't need any hardcore coding. Even I without any linux skills can "cook" (read: remove / add apps) This is not creating a rom. This is personalizing one to suit your needs.
Creating a rom imho is coming up with new (speed) hacks or options to make it more functional. Or better looking. From what I "learned" even creating a skin isn't that easy as it was with winmo with all that scripting.
my 2k
Meh, I have a G1, its pretty obvious when roms stand out, for the G1
SuperD (1.6)
FastTest (1.6) based on SuperD, but bleeding edge
CyanogenMod (1.6) Stable, more than anything
OpenEclair (2.1) Collaborative project between ChrisSaywer/WesGarner
and for 2.1 sense roms, I normally run a KingKlick rom...
it seems, for stability, and general acceptance that Fresh is the way to go, but there are other builds that use his stuff for a base that are more bleeding edge but have improvements... its pretty easy to see whats popular just by a view/post count on the thread.
Uqbar said:
No calculation. Simply I compare performances and features from 3.3 to 3.4.
3.4 is no more stable or fast than 3.4, to me at least.
Memory footprint is almost the same. Not FC in aither ROM ...
And I bet that if nprussel brought anything bigger than 5% he would have put it into the annoucement.
Instead it shows little more than new application list and a GUI theme.
Click to expand...
Click to collapse
VillainROM 3.3:
1. Linpack: 2.336 MFLOPS (more is better)
2. BenchmarkPi: 13664ms (less is better)
3. Benchmark: (more is better)
* Graphic: Total graphics score - 123.18746
* CPU: Total CPU score - 157.91457
* Memory: Total memory score - 139.30379
* Filesystem: Total filesystem score - 25.63833
VillainROM 3.4:
1. Linpack: 2.372 MFLOPS (more is better)
2. BenchmarkPi: 13049ms (less is better)
3. Benchmark: (more is better)
* Graphic: Total graphics score - 129.06668
* CPU: Total CPU score - 149.84908
* Memory: Total memory score - 131.71753
* Filesystem: Total filesystem score - 34.176556
I suppose 5% is close. I didn't claim it would be significantly quicker. It's definitely quicker though.
Anyway, that's besides the point.
Benham is leading the way right now in my opinion, but I have just got my hands on a Legend leak, that I have got to boot on the Hero. I'm currently trying to get all the hardware working (Wifi, mic, back speaker, GPS all not working). I have asked a couple of other developers for some help, and we'll probably release it as a community release.
I think the rom developers should start working together and create teams. Maybe some could focus on fixing stability issues, some on improving speed and compatibility, some on implementing features from other rom releases and some on optimizing and implementing themes. There is a optimum for the OS, so why put a lot of time and effort into fixing the same bugs and problems over and over again in different rom releases? The differences lay within the themes and included features, which could be made available as separate (and optional) installer packages. So I say..... organize and collaborate! Independent Android development (independent from the manufacturers) as a community will be much more powerful and professional.
Different branches could emerge from the Hero branch easily, making the developers base grow and 'interaction' between the different branches a lot easier.
Bram77 said:
I think the rom developers should start working together and create teams. Maybe some could focus on fixing stability issues, some on improving speed and compatibility, some on implementing features from other rom releases and some on optimizing and implementing themes. There is a optimum for the OS, so why put a lot of time and effort into fixing the same bugs and problems over and over again in different rom releases? The differences lay within the themes and included features, which could be made available as separate (and optional) installer packages. So I say..... organize and collaborate! Independent Android development (independent from the manufacturers) as a community will be much more powerful and professional.
Different branches could emerge from the Hero branch easily, making the developers base grow and 'interaction' between the different branches a lot easier.
Click to expand...
Click to collapse
So, err..in other words; Workers of the world unite? well, im a keen supporter of that

[Q] [HELP] Which Lagfix to use?

Anyone please?
My holiday is ending, and I still have to find an answer!
Thanks.
There are a lot of lag fixes, and their differences are beyond 'this one is better'.
So in the end it's down to personal preference and the way you use your phone.
I personally like supercurio's Project Voodoo lagfix the most. He will release a public version (beta2) pretty soon .
If you want a lagfix right now, use One Click Lag Fix. If you can wait, wait for supercurios Voodoo lagfix beta 2.
There is a tiny battle going on between filesystem of choice - EXT2 or EXT4 - for lagfix. This is good for us users for speed and stability of the fix. I would suggest you wait for a couple of days to see if Voodoo Beta 2 fits you. Else you have OCLF which yields a lot of Quadrant scores and keeps a lot of people happy here already.
Personally, I am on JM7 with Voodoo Beta 1.
I have JM7 with the One Click Lag Fix, suits me perfectly, the phone just "flies". Also, the ease of install is the best - just install it from the market.
I have a couple of questions. Do I need to root in order to do the lag fix? I noticed that the One Click Lag Fix app can root my device too.
Also, how will my phone react when Froyo is released? Can I upgrade my phone through Kies without doing anything, or do I need to somehow get rid of root/lag fix?
Tried all the lag fixes EXCLUDING Voodoo. I can say my phone is better off them even when it WORKS. While they do some "cosmetic" speed improvement, I find them actually INCREASING lag in some typical activities like installing, uninstalling apps, multitasking, etc.
Latest I tried OCLF 1.6 and finally revert the lag fix to see my phone working "normally". Hope you could read through the punctuations
Prankey said:
Tried all the lag fixes EXCLUDING Voodoo. I can say my phone is better off them even when it WORKS. While they do some "cosmetic" speed improvement, I find them actually INCREASING lag in some typical activities like installing, uninstalling apps, multitasking, etc.
Latest I tried OCLF 1.6 and finally revert the lag fix to see my phone working "normally". Hope you could read through the punctuations
Click to expand...
Click to collapse
what version rom are u using??
The easiest lagfix (which also greatly improves GPS (in my opinion) as a side benefit) is to just upgrade your firmware to one of the following: XFJM6, HVJH1, or XWJM7. Note the bolded F in the JM6, XWJM6 is not lag-free.
Those three firmwares seem pretty lag-free out of the box, though after a lot of up time, you might get some lag, but that took a couple of days for me, and a reboot fixes.
The next easiest lag-fix is to disable all the fancy animations and window transitions in the GUI. Do this in Settings.
The next easiest lag-fix is to get AutoKiller from the Market and use one of the medium to aggressive settings in the pre-sets. This works well at reducing some lag, and has the benefit of being easy to uninstall if you're not happy with it for some reason.
Now, if those three things don't give you a happy phone experience, personally I am in the Voodoo (by curio/supercurio) camp to be the BEST experience so far. Easy to install, best overall reduction in lag (as in I have ZERO lag even after 90+ hours of uptime and heavy usage), no Market install/download issues, no random freezes. EVERY OTHER lag-fix, from Mimocan to RyanZA's OCLF have those issues still, Voodoo does not, for me.
If I couldn't use Voodoo, I would use the mimocan fix, specifically by using the Samset/mimocan kernel.
I won't get into the ext2 vs ext4 debate that the supercurio and ryanza camps have fallen into.
My advice right now, try the first three things I suggested and WAIT until the end of the week to try supercurio's Voodoo Beta 2 when he goes public with it.
As of right now, I am on XWJM7 firmware with voodoo beta 1, and installing voodoo beta1 is the first time since buying this phone that I have been satisfied with its performance. YMMV, of course.
edit: ps - to be clear, I have tried just about all of them...
distortedloop said:
The easiest lagfix (which also greatly improves GPS (in my opinion) as a side benefit) is to just upgrade your firmware to one of the following: XFJM6, HVJH1, or XWJM7. Note the bolded F in the JM6, XWJM6 is not lag-free.
Those three firmwares seem pretty lag-free out of the box, though after a lot of up time, you might get some lag, but that took a couple of days for me, and a reboot fixes.
The next easiest lag-fix is to disable all the fancy animations and window transitions in the GUI. Do this in Settings.
The next easiest lag-fix is to get AutoKiller from the Market and use one of the medium to aggressive settings in the pre-sets. This works well at reducing some lag, and has the benefit of being easy to uninstall if you're not happy with it for some reason.
Now, if those three things don't give you a happy phone experience, personally I am in the Voodoo (by curio/supercurio) camp to be the BEST experience so far. Easy to install, best overall reduction in lag (as in I have ZERO lag even after 90+ hours of uptime and heavy usage), no Market install/download issues, no random freezes. EVERY OTHER lag-fix, from Mimocan to RyanZA's OCLF have those issues still, Voodoo does not, for me.
If I couldn't use Voodoo, I would use the mimocan fix, specifically by using the Samset/mimocan kernel.
I won't get into the ext2 vs ext4 debate that the supercurio and ryanza camps have fallen into.
My advice right now, try the first three things I suggested and WAIT until the end of the week to try supercurio's Voodoo Beta 2 when he goes public with it.
As of right now, I am on XWJM7 firmware with voodoo beta 1, and installing voodoo beta1 is the first time since buying this phone that I have been satisfied with its performance. YMMV, of course.
edit: ps - to be clear, I have tried just about all of them...
Click to expand...
Click to collapse
I agree completely with you, except about waiting for Voodoo BETA 2 at this point. Please note the BETA - even though it will be pretty heavily tested, please don't jump onto these lagfixes until they are out of beta, especially if you don't know what you're doing. There may be complications, especially if you grab it as soon as it's released.
Please, this type of thing is marked BETA for a reason! When it goes to RELEASE, then you can jump on it!
Of course if you want to test stuff out, then please try the betas, but things can and do go wrong!
As far as the OCLF fix, it has proven so far to be very safe, with thousands of people using it without issue so far. Comments and stats available here, and this only includes market downloads: http://www.appbrain.com/app/com.rc.QuickFixLagFix
The general consensus is that OCLF does make your phone faster, but these types of things are very subjective. It also doesn't use custom kernels or similar, and works on all firmwares, which is a plus for some.
Your point that Voodoo is in beta, make that BETA, is well taken. Supercurio won't release it to public until it's well tested. Some of the other methods released here weren't approached so conservatively in their early days, but anyone in this particular forum should be aware of the risks.
Please note that the original poster said "I want the best/newer one (in terms of speed and stability), I don't mind if its not the easiest to install," which implied to me he's up to some beta stuff.
That said, your method's lack of needing custom kernels being a plus is spot on as well, but it also opens the door to the whole ext2 vs ext4 debate, since you're using ext2 and that's why it doesn't need a custom kernel, and I specifically said that I wasn't going to step into that one...
Anyways, I think your method has some plusses, of course, and it turns in some amazing Quadrant scores, for whatever that's worth, but in every day use, lag and GUI freezes still exist occasionally with it, and that is not my experience with voodoo. Voodoo also turns in some great Quadrant scores (about right in the middle of what mimocan and OCLF give me (mimo-1400, Voodoo-1800, OCLF-2200)).
Voodoo works best for me in every day "feel" of the phone. It's my recommendation to anyone if they can get into the beta. It's truly lag-free. And free space on the device is accurately reported.
Mimocan works well and is based on the apps2sd concepts that have been tried and tested over many devices and firmwares and mods like Cyanogen's. It's mostly lag-free.
OCLF is super easy to install with apps from the app store and batch programs for Windows users and has a very dedicated and respected devs behind it. It's mostly lag-free, but I'm not comfortable with the whole ext2 loop mount business it uses and the last time I used it had the disadvantage of inaccurate free space reporting, and the huge file it creates makes clockwork backups very slow and very large. I'm not recommending against OCLF, It's just not my first choice.
RyanZA said:
I agree completely with you, except about waiting for Voodoo BETA 2 at this point. Please note the BETA - even though it will be pretty heavily tested, please don't jump onto these lagfixes until they are out of beta, especially if you don't know what you're doing. There may be complications, especially if you grab it as soon as it's released.
Please, this type of thing is marked BETA for a reason! When it goes to RELEASE, then you can jump on it!
Of course if you want to test stuff out, then please try the betas, but things can and do go wrong!
As far as the OCLF fix, it has proven so far to be very safe, with thousands of people using it without issue so far. Comments and stats available here, and this only includes market downloads: http://www.appbrain.com/app/com.rc.QuickFixLagFix
The general consensus is that OCLF does make your phone faster, but these types of things are very subjective. It also doesn't use custom kernels or similar, and works on all firmwares, which is a plus for some.
Click to expand...
Click to collapse
distortedloop said:
That said, your method's lack of needing custom kernels being a plus is spot on as well, but it also opens the door to the whole ext2 vs ext4 debate, since you're using ext2 and that's why it doesn't need a custom kernel, and I specifically said that I wasn't going to step into that one...
Click to expand...
Click to collapse
Not true at all, EXT4 is very easily added to the running kernel as a module. Tayutama has an EXT4 version that works fine on the stock Samsung kernel. People just don't use it very much because it's slower. *shrug* I haven't had any requests to include EXT4 in my app either, so I haven't done it yet. And you stepped into it, but I needed to clear it up.
distortedloop said:
Anyways, I think your method has some plusses, of course, and it turns in some amazing Quadrant scores, for whatever that's worth, but in every day use, lag and GUI freezes still exist occasionally with it, and that is not my experience with voodoo. Voodoo also turns in some great Quadrant scores (about right in the middle of what mimocan and OCLF give me (mimo-1400, Voodoo-1800, OCLF-2200)).
Voodoo works best for me in every day "feel" of the phone. It's my recommendation to anyone if they can get into the beta. It's truly lag-free. And free space on the device is accurately reported.
Mimocan works well and is based on the apps2sd concepts that have been tried and tested over many devices and firmwares and mods like Cyanogen's. It's mostly lag-free.
OCLF is super easy to install with apps from the app store and batch programs for Windows users and has a very dedicated and respected devs behind it. It's mostly lag-free, but I'm not comfortable with the whole ext2 loop mount business it uses and the last time I used it had the disadvantage of inaccurate free space reporting, and the huge file it creates makes clockwork backups very slow and very large. I'm not recommending against OCLF, It's just not my first choice.
Click to expand...
Click to collapse
Unfortunately, both mimocan and voodoo (BETA1, hopefully fixed in BETA2) have very annoying battery issues, and bad battery life is worse than occasional lags. Your experience with it might have varied though.
I'm currently trying to get a fix similar to voodoo going that won't need a custom kernel, but it unfortunately needs gnu parted to do that. And gnu parted is proving to be horribly difficult to port to Android.
RyanZA said:
Not true at all, EXT4 is very easily added to the running kernel as a module. Tayutama has an EXT4 version that works fine on the stock Samsung kernel. People just don't use it very much because it's slower. *shrug* I haven't had any requests to include EXT4 in my app either, so I haven't done it yet. And you stepped into it, but I needed to clear it up.
Unfortunately, both mimocan and voodoo (BETA1, hopefully fixed in BETA2) have very annoying battery issues, and bad battery life is worse than occasional lags. Your experience with it might have varied though.
I'm currently trying to get a fix similar to voodoo going that won't need a custom kernel, but it unfortunately needs gnu parted to do that. And gnu parted is proving to be horribly difficult to port to Android.
Click to expand...
Click to collapse
No, I still haven't stepped into the ext2 vs ext4 debate, but I stand corrected that you don't need a custom kernel to use it. Is that a recent innovation?
I'm specifically avoiding the speed and potential for data loss argument. It's debated ad nauseum in several other threads.
I suspect the reasons you didn't see requests for ext4 in your own apps are two-fold. 1- you are very clearly against it, and argue strongly that ext2 is faster and that that is part of the reason your fix works and that ext4 would slow things down *shrug* and 2- many, like me, incorrectly assumed you needed a custom kernel for it. Go figya!
I've heard battery complaints about your fix method as well. I do not experience them at any noticeable rate with any of the methods. I've never made it through a full day without using up most of battery with any smartphone since my first iPhone; others claim to make it for 2-3 days. Who knows what we all do different that impacts this. It's a cheap and easy shot to take at any app or rom or mod, and for every one complaint of "drains my battery faster" there's usually a post from someone who says "doesn't affect mine."
Battery complaints are such a sore spot in the Android community. Every new custom ROM for the Nexus One, the first thing you'd see is dozens of complaints of battery life sucking. It's the same with the SGS, and now the various lag-fixes.
I'm not sure which is more obsessed about with SGS owners; lag, GPS, or battery life.
distortedloop said:
No, I still haven't stepped into the ext2 vs ext4 debate, but I stand corrected that you don't need a custom kernel to use it. Is that a recent innovation?
I'm specifically avoiding the speed and potential for data loss argument. It's debated ad nauseum in several other threads.
I suspect the reasons you didn't see requests for ext4 in your own apps are two-fold. 1- you are very clearly against it, and argue strongly that ext2 is faster and that that is part of the reason your fix works and that ext4 would slow things down *shrug* and 2- many, like me, incorrectly assumed you needed a custom kernel for it. Go figya!
I've heard battery complaints about your fix method as well. I do not experience them at any noticeable rate with any of the methods. I've never made it through a full day without using up most of battery with any smartphone since my first iPhone; others claim to make it for 2-3 days. Who knows what we all do different that impacts this. It's a cheap and easy shot to take at any app or rom or mod, and for every one complaint of "drains my battery faster" there's usually a post from someone who says "doesn't affect mine."
Battery complaints are such a sore spot in the Android community. Every new custom ROM for the Nexus One, the first thing you'd see is dozens of complaints of battery life sucking. It's the same with the SGS, and now the various lag-fixes.
I'm not sure which is more obsessed about with SGS owners; lag, GPS, or battery life.
Click to expand...
Click to collapse
I get tons of emails and messages every day about OCLF, and none have mentioned EXT4 or battery that I can think of. About a million other issues though! Mostly getting to recovery console though.
I haven't heard any battery complaints about OCLF so far, but I may have just missed em. Anyway, my complaints with battery life on custom kernels are entirely that - my complaints. Whenever I use a custom kernel and leave my device overnight, it drains around 30% of the battery or more. With a Samsung kernel, it usually drains at most 10%. This seems to be mirrored by others I talk to as well. I believe it may be wifi-driver related - the GPLed driver may not be as good as the proprietary driver or something equally bizarre.
I agree with the fact that RyanZa's solution is by far the simplest of the methods available. You can easily make out that me is a still a noob....
I don't speak for others, but my best guess is for some the ease of use matters, and for some, regardless of the complexity involved, the other methods work just fine. It usually really is a personal choice and an equally personal experience. What works best for some, does not necessarily work best for others. Some like a mild fight put up by the hardware, some dont (I dont. Yet!! )
Not wanting to get caught in you folks exchange here, but I guess the saying sums it all.
To each, his own....
P.S.: Dont flame for it...
Cheers....
skiddhard said:
Some like a mild fight put up by the hardware, some dont (I dont. Yet!! )
P.S.: Dont flame for it...
Click to expand...
Click to collapse
Gonna flame you! What are you doing here if you don't like messing with your phone/toy?!
But yeah, that's the whole point of computers, phones, and almost everything else. Making it easier for others.
Wow! Thanks for the answers guys!
I think I'm gonna wait 1 or 2 weeks... maybe the 2.2 update comes to change all of this.
Regards!
Paulo
I won't go into saying which method is best, definitely not, nor will I mix myself into ext2 vs ext4.
However, I have seen some people claim both in this thread and in other threads that upgrading firmwares alone is enough. I assure you it isn't. Whether you use mimocan's, mine, Ryan's, curio's, etc fix they all result in having a much faster device. Sure, each of the fixes has their own quirks, con's and pro's, but give it time, one of us will get this exactly right, sooner or later. If not, mayhaps Samsung will.
To be honest, I didn't notice _that_ much difference in speed initially either. And indeed noticed that in some places it was occasionally a tad slower (though with the latest versions of the various fixes I haven't personally experienced this). The only place where I noticed an amazing difference was the email app (5 exchange accounts synced, with 100's of messages a day, on 30-day sync). The difference just blows you away - with setting up the accounts too, not just using them.
But today I had the greatest notice of all. By now some of my coworkers have SGS's as well, but I'm the only one running a lagfix at the moment. It's in the small things: when I put my lagfixed device next to one of their non-lagfixed devices, even with the latest firmwares, it's just amazing how much faster mine is, how much more responsive. Simple taps that resond twice as fast. Granted, theirs aren't usually slow, but mine is just noticably faster, to a level where I'm annoyed by the few millisecond delays when handling their devices.
At the moment, I would say curio's method is the most promising, though certainly not what I would advise just anyone to use. And by curio's method I mean converting the flash itself to a filesystem, instead of using a filesystem image inside an existing filesystem. If his implementation will be the best, remains to be seen
Don't make the "n00b" mistake: I don't see a difference, I'll uninstall. Because probably, your device is fresh and clean. Once it gets cluttered a bit the lagfixed definitely help!
I currently use official JM1 firmware with OCLF apk and find it still lags a bit although not as much. I'm a little worried about the custom kernel fixes not working once froyo comes out as it's mere weeks away.
lokhor said:
I currently use official JM1 firmware with OCLF apk and find it still lags a bit although not as much. I'm a little worried about the custom kernel fixes not working once froyo comes out as it's mere weeks away.
Click to expand...
Click to collapse
Like most, I would say wait till Curio's beta2 is released later this week (hopefully). Ive used almost every variation of every method of seen of the forums, and found the using an ext4 partition on my external sd to store my dalvik-cache and app data is the most smoothest & most reliable 'band-aid'.
Although all the loopback methods work well (EXT2 or EXT4, not getting into it), data corruption is inevitable due to the fact that our loopback partition is never getting properly unmounted on shutdown or reboot. Although the corruption may be mild, I found it eventually starts to lag again as you use the fix for a week or more (more corruption each reboot\power cycle). Even using a fschk, you eventually will get corruption
Ive used all the loopback method several times for around a week or so, and found the smoothest, most reliable method is using an ExternalSD for dalvik-cache and app data until a native EXT file system can be put on /data.
Although Ive throughly tested my conclusion (and I wish it werent true, as I love the speed immediately after installing a loopback lagfix), these are simply my observations and I'm throwing them out there for others.
Thanks to everyone who is working hard to fix the "lag" issues on Galaxy S devices. Its quite an ordeal to fix this, and IMO, not enough "thank yous" get passed around, despite differing opinions.

Dinc godmode?

I really should read more about evo's godmode before asking. But the way I understand it they partition part of the harddrive to basically another ram partition. Will the incredible ever see this feature since partitioning seems to have evaded us
I've been wondering the same thing. I thought it was a matter of repartitioning and changing file systems. I thought it could all be handled in side the updater-script file, prior to installing to rom. I played with kingdom and tried to see how it was done, but we need a real dev to bring it to the inc.
Thanks for clearing that up cvb. I was still thinking it was gonna revolutionize the dinc. I think neo just meant someone other than himself. We all know Joelz has skills
well
1. its buggy
2. force to use amon ra
3. not really much of a performance boost more quadrant boost than anything
That reminds me of ReadyBoost for Windows, where you can use your old flash drives or SD cards as RAM... It makes a huge difference in older machines & netbooks but not so much on the newer ones that have ballsacs.
Its an interesting theory though. It would be great for a 3.0 rom if anyone wanted to perfect it.
JoelZ9614 said:
well
1. its buggy
2. force to use amon ra
3. not really much of a performance boost more quadrant boost than anything
Click to expand...
Click to collapse
Here's a question for you, since you seem to have a fairly extensive amount of experience with ROM's and the Dinc. What is, in your personal opinion, going to be my best option of customization, performance and please don't use words like MIUI, lol.
***EDIT*** I should've thought to include this info, but currently these are my stats
CM7 7.1.0-RC
Kernel-Incredikernel 2.6.38.8
Radio-2.15.00series
I've flopped between this and MIUI for about a month. Tried Synergy a while back, did like it a lot. Any ideas? Some I haven't heard of ? Thanks.
Thanks for the input, I appreciate it!
There's a godmode version of synergy that I've been trying to find info on what exactly it does, but I didn't find much. So godmode it just supposed to be a higher performance version of some mod due to an added swap partition and some other speed tweaks?

[Q] Encrypting the Note 8: Performance Impact?

Greetings!
For various reasons I'm thinking about encrypting my Note 8 (and the removable MicroSD card). I would welcome anyone's experience with encrypting the Note 8 regarding the following:
1) Is the performance impact noticeable (ie, does it feel more more sluggish, and if so, in what contexts)?
2) Was any impact on battery life noticeable?
I am on Android 4.2.2. I am interested in the above with regards to that version, as well as KitKat 4.4.2 (since there may be differences).
I'm not interested in discussion regarding the "why" of encryption (as in, Why am I considering this?)--I'm not looking to think through that in the context of this thread (or at least not at the moment). And I'm aware of the standard things such as, once I encrypt I will have to do a Factory Reset if I change my mind.
From what I've read encryption can/does effect performance and battery life in general, but is very device-dependent as far as whether the difference is significant or noticeable (hence my interest in people's experiences).
Thanks in advance to any who decide to share their experiences!

Is there less development for the Pixel (than Nexus devices)?

I was surprised to notice that there is no official version of Lineage OS for the Pixel. Then looking around it just seems like there aren't that many ROMs for the Pixel. And I'm surprised to see that a year after release there still seem to be difficulties with root and using TWRP.
Am I right in perceiving that the Pixel has gotten less development interest than Nexus devices of the past? If so, why?
I ordered a Pixel 2 (to replace my Nexus 4--which has offical lineage support) and was assuming that it would get all the usual developer love that I've exeperienced in the past. Now I'm a bit worried that it will be difficult just to root it and do nandroid backups.
Thanks for any thoughts and observations.
I think they are having issues with the increased security and dual OS partitions. Will take some time.
scottjal said:
I think they are having issues with the increased security and dual OS partitions. Will take some time.
Click to expand...
Click to collapse
Yeah, I saw that the dual partitions are an issue and the dm-verity check (don't quite understand what that is). But still, the Pixel has been out for a year. Nexus phones usually had tons of development by that point. I imagine if the devs were really into this phone there would be more interest. Especially the lack of official Lineage OS support, the most fundamental of all ROMs. There are unofficial versions of Lineage OS for the Pixel, so it works. I guess that's why I was starting to get the impression that the Pixel is not that popular with the devs (compared to past Nexus devices).
Price doesn't help either, I'm less likely to mess with a phone that costs a grand. Also prices it out of the hands of those that just want to tinker.
I also wonder how much baseline performance plays a role. The Pixel is really quite great out of the box, it's the first phone I've owned in recent times that I haven't had the desire to root, mod, rom, etc. custom roms are generally known to improve the end user experience, can't say that has been my experience on the Pixel and that's a big reason why I've decided stock is best for me.
mlin said:
I also wonder how much baseline performance plays a role. The Pixel is really quite great out of the box, it's the first phone I've owned in recent times that I haven't had the desire to root, mod, rom, etc. custom roms are generally known to improve the end user experience, can't say that has been my experience on the Pixel and that's a big reason why I've decided stock is best for me.
Click to expand...
Click to collapse
For me ROMs have never been about performance, they are about features. They often have more privacy features and remove some of the pernicious behind the scenes stuff. There are more options for quick settings tiles. You can control the notification LED better. I don't want a total makeover; I'm happy with basically stock Android. But there are a lot of little things that are improvements to me. I guess this is why I especially like Lineage OS, since it is not a heavily modified OS. Also you sometimes get updates for security problems sooner (the once a month standard from Google is pretty lame--in Linux security patches are pushed out the day they are ready). As for rooting, it lets you use the iptables firewall in the Linux kernel and programs like adaway, as well as a root file manager (there are many times I find it useful to browse the system files); there is Titanium backup that is so much more powerful than any other backup tool that requires root; and a rooted device can also often get around issues with tethering. And having a custom recovery like TWRP allows nandroid backups, which have saved me from disaster so many times. So I think even on a phone like the Pixel, there's a lot to be gained from ROMs, rooting, minor mods, a custom recovery.
Anyway, I guess maybe it is just about the cost of the Pixel. Or perhaps the complexity of the dual partitions--are the Pixel phones the only ones that do this? I wonder what new phones get the most developer attention these days, if it's not the Pixels.
No, there's significantly less development, regardless of what a few users say (when I mentioned it a week ago after coming from a 6P, people were shocked when I said there's no development lol). I had to point out that 5 ROMs does not even compare to the 30 ROMs of the 6P.
At first, the pixel seemed like a downgrade to me. But after using it a week or so, it is a little smoother I got, and has better battery life, which is awesome. But those are about the only pros. Camera on mine is the same quality as the 6P, and I can't get used to the crappy downfiring speakers (last 2 devices had dual front facing).
All in all, it's an ok phone. It's about 2" taller than it should be, due to Google love affair with gigantic bezels. Haha. But sometimes using a bezeless phone one handed if tough, because your thumb can't reach the navbar.
Sent from my Pixel XL using Tapatalk
cb474 said:
For me ROMs have never been about performance, they are about features. They often have more privacy features and remove some of the pernicious behind the scenes stuff. There are more options for quick settings tiles. You can control the notification LED better. I don't want a total makeover; I'm happy with basically stock Android. But there are a lot of little things that are improvements to me. I guess this is why I especially like Lineage OS, since it is not a heavily modified OS. Also you sometimes get updates for security problems sooner (the once a month standard from Google is pretty lame--in Linux security patches are pushed out the day they are ready). As for rooting, it lets you use the iptables firewall in the Linux kernel and programs like adaway, as well as a root file manager (there are many times I find it useful to browse the system files); there is Titanium backup that is so much more powerful than any other backup tool that requires root; and a rooted device can also often get around issues with tethering. And having a custom recovery like TWRP allows nandroid backups, which have saved me from disaster so many times. So I think even on a phone like the Pixel, there's a lot to be gained from ROMs, rooting, minor mods, a custom recovery.
Anyway, I guess maybe it is just about the cost of the Pixel. Or perhaps the complexity of the dual partitions--are the Pixel phones the only ones that do this? I wonder what new phones get the most developer attention these days, if it's not the Pixels.
Click to expand...
Click to collapse
While custom ROMs may not be about performance to you, I know that it is to others. I think there are a lot of variables at play, and I don't discount the out of the box experience as one of them. Take that custom ROMs are prone to decreased stability, especially with the Pixel and the argument for advantages of custom ROMs gets diminished fairly quickly. Unless of course you must have all the frivolous tweaks at the cost of stability and performance.
mlin said:
I also wonder how much baseline performance plays a role. The Pixel is really quite great out of the box, it's the first phone I've owned in recent times that I haven't had the desire to root, mod, rom, etc. custom roms are generally known to improve the end user experience, can't say that has been my experience on the Pixel and that's a big reason why I've decided stock is best for me.
Click to expand...
Click to collapse
I agree, this phone is pretty hard to improve on. I'm missing all the flashing and tweaking but I'm coming to the conclusion that stock is the ticket. The stock firmware doesn't feel like a compromise anymore, it feels like it works best. With payment via phone and bank accounts on board, security is also an issue I've not worried about before but I guess I need to deal with.
mlin said:
While custom ROMs may not be about performance to you, I know that it is to others. I think there are a lot of variables at play, and I don't discount the out of the box experience as one of them. Take that custom ROMs are prone to decreased stability, especially with the Pixel and the argument for advantages of custom ROMs gets diminished fairly quickly. Unless of course you must have all the frivolous tweaks at the cost of stability and performance.
Click to expand...
Click to collapse
I didn't say no one cares about performance. In fact, I only said that there are a lot of things ROMs do other than performance, so even if that issue is less with a phone like the Pixel, it hardly negates the many and varied reasons that people like ROMs. So I was just suggesting that reducing it all to performance as the fundamental issue, I think, oversimplifies how varied and complicated the custom ROM community of users and devs is.
Anyway, my main question in the OP wasn't really why do people like or dislike custom ROMs, nor was it do they think they are unnecessary for the Pixel. My main question was, do people think my perception is correct that devs are less interested in the Pixel than in past Nexus phones?
Personally, I believe not all developers are created equal. The pixel design and new os introduced complexities beyond the abilities of the cut and paste developer. The truly talented developers are either still hard at work or have concluded, no not worth it.
Who knows?
There are many roms out there. There are just not on xda anymore.
Development is just so much harder on the pixel, so we need to wait till the first running custom rom is out there so the other devs can pick from there. It's already been working hard on it.
Development is not so high cause a dev need the device and not every dev has the money to buy a 1000 $ device every year.
Hopefully some developers fell into some pixels when Google started giving them out as replacements for 6Ps. We'll have to wait and see.
I still suck at G+. I can't get the hang of how to search it for roms, or even how to actually get the rom, when I know a developer has one on their page. I wish Google could have tried out the service before they released it, maybe they would have noticed how difficult it can be - or maybe it's just the developers and how they organize their page...who knows.
Sent from my SM-T330 using Tapatalk

Categories

Resources