Reliable Write: The reason /system reverts back to stock - HTC EVO 3D

This is a quote from the Sprint forums. I happened to subscribe to the thread and got this through e-mail:
I have seen first hand the current level of protection HTC has built into the device, and it is impressive. Should some rogue software use an exploit to obtain root, and make changes to the /system partition, the system detects this, revokes the root priveledge from the process that obtained it, then reverses all changes to /system. The technology behind part of this is a feature of eMMC called reliable write. A write-up on a commercial adaptation of this is http://blog.datalight.com/doing-in-place-os-updates-for-embedded-devices here. This sounds ideal for ensuring that a failed OTA update won't break the phone, and protection against malware and the like, but it also makes it impossible to flash roms or make other customisations to the core of the phone.
Click to expand...
Click to collapse
-Posted by some guy named MS072467
Here is the link to the original discussion on Sprint.com:
http://community.sprint.com/baw/message/313016#313016

Wow is all i can say thatrs really involved thanks a lot HTC (eyes roll)

Sucks for root but makes the phone much safer from malware.

Seriously, it's a piece of brilliance as far as system security goes, and I hope we can continue to use it to our advantage once we've cracked the phone open, as it's a great idea. The problem is not that the system reverts writes, it's that we can't tell it what writes are good. We're locked out of our own devices, and the built-in security measures are being used against the owners. That's not good.

canteenboy said:
Sucks for root but makes the phone much safer from malware.
Click to expand...
Click to collapse
Malware is really that big of a problem to warrant a complete lock-down of the phone? Been using Android for over a year, downloading any app I felt like, and never once had a "virus" take over my phone.
Mike

mikeyxda said:
Malware is really that big of a problem to warrant a complete lock-down of the phone? Been using Android for over a year, downloading any app I felt like, and never once had a "virus" take over my phone.
Mike
Click to expand...
Click to collapse
Does that mean we ignore system security? Malware might not have been a serious problem before, but you throw something like this in, and it all but guarantees that it will continue to not be a problem for some time to come.
Again, the problem here is not that the phone is locked down, the problem is that the owners didn't get the key when they bought the property.

I honestly dont see this as a big saving grace. After all, most malware is likely to come in the form of rogue legit looking apps like a keyboard that could log our passwords without needing root.
IMO they weren't thinking of malware when they implemented this, they were thinking of xda!
Sent from my PG86100 using XDA Premium App

naw they were more thinking about how much $$$ they would save from being able to effectively push a software update correctly the first time...
if they were really that worried about keeping "xda" out of the phone they wouldnt be unlocking bootloaders. i cant even begin to tell you how much easier my job would be if i didnt have to replace one more device because they did an update and now their phone doesnt work.

sn0b0ard said:
This is a quote from the Sprint forums. I happened to subscribe to the thread and got this through e-mail:
-Posted by some guy named MS072467
Here is the link to the original discussion on Sprint.com:
http://community.sprint.com/baw/message/313016#313016
Click to expand...
Click to collapse
That would be me I started a similar thread on the subject here. It was moved to General from Dev some time ago.
Just to clarify, Reliable Write being the culprit was just an assumption, based on what we all see happening when /system reverts, and that it would be the most logical mechanism for performing the reversions, given that it is a feature of eMMC to begin with.
There was a dev who replied to the original thread that disagreed, and thought it may be a proprietary emulation of similar technology.
I contacted Micron to get permission to view the documentation of the technology but never received a reply.

sn0board - Many thanks for the helpful info.
The Shift uses eMMC and we had perm root after a fairly short time - lotta work (tip of the topper for the many who worked that) - but we got it.
http://forum.xda-developers.com/showthread.php?t=932153
I wonder if they'd implemented that feature for the Shift - visionary or z4root provide workable temp root without the difficulties here.

CyWhitfield said:
That would be me I started a similar thread on the subject here. It was moved to General from Dev some time ago.
Just to clarify, Reliable Write being the culprit was just an assumption, based on what we all see happening when /system reverts, and that it would be the most logical mechanism for performing the reversions, given that it is a feature of eMMC to begin with.
There was a dev who replied to the original thread that disagreed, and thought it may be a proprietary emulation of similar technology.
I contacted Micron to get permission to view the documentation of the technology but never received a reply.
Click to expand...
Click to collapse
Ahh makes sense now. The more information we can get on the protection mechanisms behind eMMC and the Evo 3D, the better.
It's odd that they keep moving these posts that talk about the protection of the eMMC. I mean, if we can't disable this write protection, then how are we going to develop ROMs for it in the first place?

sn0b0ard said:
Ahh makes sense now. The more information we can get on the protection mechanisms behind eMMC and the Evo 3D, the better.
It's odd that they keep moving these posts that talk about the protection of the eMMC. I mean, if we can't disable this write protection, then how are we going to develop ROMs for it in the first place?
Click to expand...
Click to collapse
Given that arguably the largest obstacle to loading the ENG bootloader is this eMMC protection, I don't understand how its considered irrelevant to development either.
In a way, I hope it is irrelevant. I hope its easier to break this thing than what I have read in the articles I posted would suggest.

eMMC JEDEC standard
I'm not sure who makes the eMMC chips in our phones even though I heard someone say Micron, but I would assume the support or use JEDEC standards which attempts to standardize environment, data handling and such. I am not sure what the current standard is even though I think it is 4.41. Here is an interesting pdf on mapped commands and bit setting used to handle the writing of data to eMMC. Standard in this PDF is 4.4.
http://rere.qmqm.pl/~mirq/JESD84-A44.pdf

Jason0071 said:
I'm not sure who makes the eMMC chips in our phones even though I heard someone say Micron, but I would assume the support or use JEDEC standards which attempts to standardize environment, data handling and such. I am not sure what the current standard is even though I think it is 4.41. Here is an interesting pdf on mapped commands and bit setting used to handle the writing of data to eMMC. Standard in this PDF is 4.4.
http://rere.qmqm.pl/~mirq/JESD84-A44.pdf
Click to expand...
Click to collapse
Interesting find

Related

Possible? True Security Protection?

Well I was just reading a thread about someone buying a Vibrant from someone who "found" it and this person was looking for a way to bypass WaveSecure.
We all know that with a little know how that it is possible with Recovery Mode.
The question I have is there a way to prevent even a Recovery Mode reflash? To absolutely stop someone from touching the ROM at all?
I know the Security Apps out right now can track you from GPS, wipe the phone remotely, etc... But can it stop someone from reflashing a ROM?
If there is a app out there like that please let me know, but if not, what would it take to create such a app.
What are YOUR thoughts??
What if this happens and then you brick for some reason need to reflash and it's locked. I would just bank on the fact that most people think that it's a "Droid" phone and don't know ****.
I was hoping for a question like that.
Either there is a security measure which at some point of using Recovery that it asks for a password or pin. Something that will allow you to access it securely and nobody else.
Yes, it is a droid, very true, but how many droids are out there now, are going to be out there, and with the new laws that allow you to unlock your device and pretty much do anything with it, more and more people are going to start playing around. Not only that, there is always somebody who knows someone, you know.
Personally myself, I would feel secure with having an implementation like this, everything else is pointless.
It's sort of like having a anti virus on your computer but not scanning for rootkits, only viruses.
The idea of that app sounds nice and all that but I seriously doubt that the average Android user would know about flashing ROMs and all that. But if it does get into the hands of somebody that does know how to do it then it can be a problem.
jzero88 said:
Yes, it is a droid, very true, but how many droids...
Click to expand...
Click to collapse
First of all these are android devices / android phones. I was mocking the people who call these phones "droid" phones.
Now on topic: All it takes to break this security is for one person to say, "I forgot my password on for the ==sUPERlOCKER== what do I do to get access?" Then all your worry is for nothing again.
What has been done can always be undone.
Sure, unlike me, I never forget my passwords. Especially for something this serious.
Second, of course something can be undo, but to what extent, after hearing your lack of concern makes me think you don't even have a lock on your phone
Again, would you rather have a password like "1234" that is easily guessed, or would you rather have something like "00LowJK54889$3%#". It's really a matter of personal security.
You sound like one of those people who would have Security Cameras, but never has the DVR on to record anything.
I'm saying your idea is bad. I have illustrated why. You have no counterpoint other than that I am 'relaxed' about my phone security.
How about this, keep your phone in your pocket or hand? 100% security.
This should be in general and not development
Sent from my Vibrant using xda app
This has been discussed a few times, you could compile your own recovery image and program in a password while at it, or you can accept that 90% of theives(or people who would find your phone) cannot get to recovery. If I found a phone then yeah I would go straight to recovery but I'm not your average user.
Sent from my T-Mobile myTouch 3G Slide using XDA App
I'm saying your idea is bad. I have illustrated why. You have no counterpoint other than that I am 'relaxed' about my phone security.
How about this, keep your phone in your pocket or hand? 100% security.
Click to expand...
Click to collapse
First, my idea is not bad, give it time, you will see.
Second, I do not have a counterpoint because my main point is stated in the first post. Read again.
Third, I don't care if you are relaxed about your security or not. This post obviously is not for you, another negative person who stunts development if they do not see a logical use for themselves.
I wish you the best and hope that you do not need to ever use such a tool or measure. Take it easy.
This has been discussed a few times, you could compile your own recovery image and program in a password while at it, or you can accept that 90% of theives(or people who would find your phone) cannot get to recovery. If I found a phone then yeah I would go straight to recovery but I'm not your average user.
Sent from my T-Mobile myTouch 3G Slide using XDA App
Click to expand...
Click to collapse
On the Vibrant forums? Haven't seen anything yet.
Also, I am not betting on a thief or someone who found the phone to be able to get to recovery, I'm worried about who these people might know. It's surprising to see how many people out here think that they are the only person in a 20 mile radius who knows how to do such mods... Maybe it's just the people I know but I know quite a few people who can easily google and find a way, easily.
I can bet that 90% of people here do not know anything except following directions, no pun intended to those who do. I definitely do not know half of what I should know, but again, is it really that hard?
Your own logic defeats what you are saying here. Don't you understand OP?
If there is a security measure, there will be a work around it? So why have more than ONE thing for the uneducated masses and stop there?
If the person who steals your phone knows someone who could get around WaveSecure, or any other security application. Then that same person can get around ANY AND ALL other types and forms of theft deterrent. If not, they will know someone, ask on forums, etc. UNTIL they gain access.
zaduma
Then why have any security on anything at all?
You my friend make no sense, good day!
jzero88 said:
Then why have any security on anything at all?
Click to expand...
Click to collapse
Ok, I will lay it out as simply as I can man. I do not want to argue, but you are missing why this is impossible to accomplish.
The existing security layers can be compromised by lets say... 10% of the population, seeing as most people who are thieves do not talk about it, most people dislike thieves.
So effectively 90% of people will be stopped dead in their tracks by having WaveSecure, etc.
The 10% who are not stopped however, can not be stopped by any means. None. They are the people who read these forums, have technical ability, etc.
Therefore having one layer of security means 90% of people are stopped from using your device. But it has ridiculously diminishing returns. With two layers, say stopping access to recovery, 10% are now stopped. Just boot into download mode and flash with odin. Stop download mode? First of all how? Second of all, there has to be a workaround for people who forget their passwords and stuff. And guess what, those 10% will know about that as well.
So please, address these issues and resolve them somehow, and your idea has merit. Without doing so you are wasting your time.
Also, much to your liking I will assume, I will no longer be posting in this thread due to your constant elevation of flaming.
Any security pro will tell you, if you have physical access to a computer, you can make it usable for you. The only real security you can hope for its to prevent access to your data by the thief. That's what full disk encryption and such is about. For our phones, we could achieve this much with a custom kernel perhaps, but how would you enter the password? No keyboard at that level.
The cellular providers can prevent the stolen phone from getting on their networks, and some do, but that's about as far as it goes.
Its like having a lock on your front door.. Its only going to keep out the honest people... Thats what they are made for, honest people, because dishonest people will just kick the door in.. And the good thieves can pick a dead bolt...
Sent from my SGH-T959 using XDA App
I'm starting to think this request/question is for the wrong crowd, truly it is...
If you build it they will hack it... Hands down... Look at the droid x, the unhackable phone, it took 5 weeks..
Sent from my SGH-T959 using XDA App
I agree, never did I not. This thread wasn't to debate whether a security measure could be hacked or not, the thread was created to see what we could do to implement such a measure.
I am totally aware of that. I know that if there is a will there is a way.
PERSONALLY, that is something I wouldn't mind having. Though some of you disagree and have a right to your own opinion, that is beyond the point. I am trying to see if a) is it possible. and b) what it would take to do so, and possibly c) if anyone was interested in trying or helping out.
So feel free to express your opinion. Mine is that you can never have enough protection cuz I would never bring a knife to a gun fight. But that's just me...
BTW, those who hacked the unhackable phone I would consider being part of the .01%.
jzero88 said:
I'm starting to think this request/question is for the wrong crowd, truly it is...
Click to expand...
Click to collapse
If you mean people that know how things work, I suppose. It's the same problem as drm. When you understand why that's not possible, you will understand this. Read up on jtag as well, you can't protect against that. 90% is about as good as it gets.

When custom roms get made

Just wondering. If and when we get custom rom's (fingers crossed for cyanogenmod) how are we going to deal with playing the HD content we've come to love on the archos'? (Hell, its pretty much the only thing it can do without struggling it seems). As the video player is archos code.. hopefully can pull it out of rom without problem when have full root. But if not its a bit of a drag as I haven't found any other player on android which plays as much and as well as the archos one.
Does the source code which was released earlier Contain the code to get hdmi-out/video player stuff? I'm guessing not
Sent from my HTC Desire using XDA App
They can just leave it in the ROM right? I mean I don't think they will be writing them from scratch. Just modifying the existing roms.
given that the performance of the rom is worse than what I was running on the G1 i kinda hope that people WILL be cooking from scratch.
thefunkygibbon said:
given that the performance of the rom is worse than what I was running on the G1 i kinda hope that people WILL be cooking from scratch.
Click to expand...
Click to collapse
i understand where you are coming from, but imo that's going a little far. maybe i am misunderstanding you a bit. the issues are because it is hardware that is not meant to have android running on it.
even if it was feasible for people to "start from scratch" or even from the base android froyo source(not archos) I think it would call for more work than using the source and going backward.
rom devs have gotten good at picking apart and re-arranging, but starting from scratch is very difficult. rarely in any kind of programming are you "starting from scratch"
digibucc said:
i understand where you are coming from, but imo that's going a little far. maybe i am misunderstanding you a bit. the issues are because it is hardware that is not meant to have android running on it.
even if it was feasible for people to "start from scratch" or even from the base android froyo source(not archos) I think it would call for more work than using the source and going backward.
rom devs have gotten good at picking apart and re-arranging, but starting from scratch is very difficult. rarely in any kind of programming are you "starting from scratch"
Click to expand...
Click to collapse
hardware not meant to have android? eh? what aspect are you talking about? its only the chipset that would really matter. hardware rarely is made for software, its usually the other way around. the snapdragon cpu in this device is the same cpu that is in my desire (apparently) .. yes the "motherboard" or whatever the equivalent is when it comes to these sort of devices and memory might be different spec and maybe worse than other devices, they could be contributing towards the poor performance compared to other devices running the same cpu. but i would tend to suggest that the rom optimisation has a very large contribution to the poor performance of this (and many other) devices.
thefunkygibbon said:
hardware not meant to have android? eh? what aspect are you talking about? its only the chipset that would really matter. hardware rarely is made for software, its usually the other way around. the snapdragon cpu in this device is the same cpu that is in my desire (apparently) .. yes the "motherboard" or whatever the equivalent is when it comes to these sort of devices and memory might be different spec and maybe worse than other devices, they could be contributing towards the poor performance compared to other devices running the same cpu. but i would tend to suggest that the rom optimisation has a very large contribution to the poor performance of this (and many other) devices.
Click to expand...
Click to collapse
that was google's own answer.
android is meant to have a mobile network, cell phone network. it's also not meant to be at the size or resolution of the (A70/101) screen. those two things alone cause the majority of the problems people have with non-archos android apps
Google themselves said they do not support Android on tablets, as they are not made for them. i can't be more specific as i only know so much about Android, but if Google said it wasn't made for it, that's what i say too.
the performance of the system on certain hardware is nothing to do with googles comment. they said that in relation to the user experience/app compatibility/screen scaling side of things. covering their own ass with the plethora of tablets coming out and wanting to distance itself from the responsibility that joe public will assume they should have (when, rightly so, its not)
as for the phone network/gps etc they are the arbitary "minimum specs" that google set out to govern what devices would be allowed to use the google marketplace and other built in google apps. again it is just a way of trying to gain some form of quality control with the amount of people creating android based systems on crap hardware.
google dont "support" them anyway. its open source.
ya know what, nevermind.
you are 100% right, never doubt that.
digibucc said:
ya know what, nevermind.
you are 100% right, never doubt that.
Click to expand...
Click to collapse
ummmm ok
When we start to get some optimized kernels I'm sure we'll see a nice performance boost. You don't have to rebuild the entire ROM from scratch to see significant improvements. Hopefully we'll get a custom bootloader. As a side note the Archos 70/101 has an OMAP 3640 which is a good deal faster than the snapdragon, first-gen anyway.
digibucc said:
i understand where you are coming from, but imo that's going a little far. maybe i am misunderstanding you a bit. the issues are because it is hardware that is not meant to have android running on it.
even if it was feasible for people to "start from scratch" or even from the base android froyo source(not archos) I think it would call for more work than using the source and going backward.
rom devs have gotten good at picking apart and re-arranging, but starting from scratch is very difficult. rarely in any kind of programming are you "starting from scratch"
Click to expand...
Click to collapse
digibucc said:
that was google's own answer.
android is meant to have a mobile network, cell phone network. it's also not meant to be at the size or resolution of the (A70/101) screen. those two things alone cause the majority of the problems people have with non-archos android apps
Google themselves said they do not support Android on tablets, as they are not made for them. i can't be more specific as i only know so much about Android, but if Google said it wasn't made for it, that's what i say too.
Click to expand...
Click to collapse
99% of what you've just written is wrong and/or has never been uttered by anyone at Google.
It's like everything you've just posted was gleaned from a game of telephone rather than a primary, secondary, or even tertiary source.
REAVER117 said:
When we start to get some optimized kernels I'm sure we'll see a nice performance boost. You don't have to rebuild the entire ROM from scratch to see significant improvements. Hopefully we'll get a custom bootloader. As a side note the Archos 70/101 has an OMAP 3640 which is a good deal faster than the snapdragon, first-gen anyway.
Click to expand...
Click to collapse
Per /proc/cpuinfo it should actually be a 3630-1000, same as the Droid X (amongst others).
JasonOT said:
99% of what you've just written is wrong and/or has never been uttered by anyone at Google.
It's like everything you've just posted was gleaned from a game of telephone rather than a primary, secondary, or even tertiary source.
Click to expand...
Click to collapse
so i just take your word for it instead?
how often do programmers re-write something that has already been done, from scratch? i'm not saying it doesn't happen - i'm saying far more often code gets recycled. if something is already made, there is no reason to reinvent the wheel.
and thank you for your primary, secondary and tertiary sources - as well as your enlightening explanation of what is true.
i can accept when I am wrong, but not just because some random person on the internet tells me I am, with nothing to prove his point any more than my own.
@digibucc
thefunkygibbon already gave you the answers why you're wrong
chulri said:
@digibucc
thefunkygibbon already gave you the answers why you're wrong
Click to expand...
Click to collapse
sure -
that's why thefunkygibbon explained it away as unimportant , yet Jason argues it was never even said...
well which is it chulri, if you are paying attention so well?
digibucc said:
sure -
that's why thefunkygibbon explained it away as unimportant , yet Jason argues it was never even said...
well which is it chulri, if you are paying attention so well?
Click to expand...
Click to collapse
Are you being obtuse for a reason?
A device made to run android by Archos, and comes booting ONLY Android OS by default... is being said as not being made to run android....
Smells like trolling to me.
Maybe people just like to troll for no reason, blazingwolf.
why do you people think they didn't come with Google Apps! Google doesn't support it! We never would have had to use g4a if they did...
obtuse? trolling? what is wrong with you you guys! I just said what i read, and everyone starts attacking me without anything to back up their side.
http://phandroid.com/2010/09/10/sho...d-not-meant-for-tablets-in-its-current-state/
http://www.dailytech.com/Google+Say...ets+May+Block+App+Market+Use/article19592.htm
http://androidcommunity.com/android...ays-google-may-block-app-market-use-20100910/
http://www.engadget.com/2010/09/10/google-android-2-2-not-designed-for-the-tablet-form-factor/
need more???
http://www.intomobile.com/2010/09/10/google-android-2-2-froyo-not-optimized-for-tablets/
http://www.techradar.com/news/phone...gle-android-not-optimised-for-tablets--715550
http://nexus404.com/Blog/2010/09/10...droid-market-but-the-galaxy-tab-doesnt-count/
grow up please, I thought this place would be better than archosfriends, but not if everyone comes here and acts the same.
someone show me how what i said is not true. i am not purposely arguing with anyone, and i am not a damn troll! I just read something, repeated it, and got slammed for it. yet no one has shown how i was wrong, you just keep saying it. show me and i'll admit it, otherwise just drop it, please.
digibucc said:
why do you people think they didn't come with Google Apps! Google doesn't support it! We never would have had to use g4a if they did...
obtuse? trolling? what is wrong with you you guys! I just said what i read, and everyone starts attacking me without anything to back up their side.
http://phandroid.com/2010/09/10/sho...d-not-meant-for-tablets-in-its-current-state/
http://www.dailytech.com/Google+Say...ets+May+Block+App+Market+Use/article19592.htm
http://androidcommunity.com/android...ays-google-may-block-app-market-use-20100910/
http://www.engadget.com/2010/09/10/google-android-2-2-not-designed-for-the-tablet-form-factor/
need more???
http://www.intomobile.com/2010/09/10/google-android-2-2-froyo-not-optimized-for-tablets/
http://www.techradar.com/news/phone...gle-android-not-optimised-for-tablets--715550
http://nexus404.com/Blog/2010/09/10...droid-market-but-the-galaxy-tab-doesnt-count/
grow up please, I thought this place would be better than archosfriends, but not if everyone comes here and acts the same.
someone show me how what i said is not true. i am not purposely arguing with anyone, and i am not a damn troll! I just read something, repeated it, and got slammed for it. yet no one has shown how i was wrong, you just keep saying it. show me and i'll admit it, otherwise just drop it, please.
Click to expand...
Click to collapse
1. "it is hardware that is not meant to have android running on it."
It's running the same CPU/GPU SOC as the Droid X with half the RAM. It has a capacitive display. The only real difference is the resolution of the display and lack of cellular radio. That's a far cry from what you wrote.
If it were, say, Android running on an old PowerPC CPU, you'd be correct.
2. "android is meant to have a mobile network, cell phone network."
Uhh, no. There's no reason Android needs a mobile network.
3. "it's also not meant to be at the size or resolution of the (A70/101) screen."
While true, it's rather irrelevant. And the statement that followed it...
4. " those two things alone cause the majority of the problems people have with non-archos android apps"
...is completely wrong.
For starters, the majority of people aren't having problems with non-Archos apps. Secondly, the majority of people having problems, are having problems due to Archos' distribution of Froyo. Those that stayed with 2.1 aren't having nearly as many problems.
Yes, a small handful of apps don't work well due to the increased resolution. A small handful of apps will only cover a portion of the display. Note that they are a small handful. The vast vast majority of Android apps work perfectly well on the Archos tabs.
5. "Google themselves said they do not support Android on tablets, as they are not made for them. i can't be more specific as i only know so much about Android, but if Google said it wasn't made for it, that's what i say too."
"Not made for them" is, again, a far cry from the truth. Not designed with tablets in mind, or even not optimized, is more like it. Like thefunkygibbon already told you, Google making those comments was strictly to cover their own asses when people get upset that they can't play games like Guitar Hero -- that represent maybe 5% of all apps -- because they aren't well coded and cannot scale to greater resolutions.
You've made a mountain out of a moehill, and the 7 links you just posted support this notion. So yes, obtuse and/or trolling are both justified descriptions of your Chicken Little act.
It's true its not made for tablets. It will work on them, but it was made with phones in mind. In other words, its a matter of the thought behind the software. What it is catered to. Its mostly the little things. Icon spacing on the default home page. Lots of little things. If they didn't make that statement they would get slammed about all of those little things. Its not really a hardware issue, more of a functionality and UI issue.

[INFO] eMMC and Data Reliance

First off, I want to apologize if this information is either or both regurgitated and irrelevant.
I was looking for information on eMMC, and there really isn't much, and I found an old article that describes how data reliance works with eMMC. At least a cursory look.
One of the features of Reliance (and Reliance Nitro) file system is that it never overwrites live data. It will always use free space on disk or in case there is no space, it will give “disk full” error back to the application. Reliance also has a special transaction mode called “Application-controlled”. In this case, Reliance only conducts a transaction point when asked by the application.
Click to expand...
Click to collapse
Full article here. Information about integration with embedded linux, here.
What struck me was the "Application-controlled" part. It would explain the technology that is undoing changes to /system when the system kills the temp root. I wonder if its possible for temp root to trigger the "commit" function of reliance once some small changes have been made...
Hope this is of some use.
CyWhitfield said:
First off, I want to apologize if this information is either or both regurgitated and irrelevant.
I was looking for information on eMMC, and there really isn't much, and I found an old article that describes how data reliance works with eMMC. At least a cursory look.
Full article here. Information about integration with embedded linux, here.
What struck me was the "Application-controlled" part. It would explain the technology that is undoing changes to /system when the system kills the temp root. I wonder if its possible for temp root to trigger the "commit" function of reliance once some small changes have been made...
Hope this is of some use.
Click to expand...
Click to collapse
Just an FYI, system is an EXT4 FS. This would require not only a custom kernel, but a lot of one offs in the way it's dealing with data. From what I've seen, this isn't what they are using.
But that's a very good find, I am looking into some of the information. Never heard of this before.
Thanks for the info. I would love to find out more about how this memory technology works. More articles are welcome!
Isn't that basically just wear leveling?
Is your name Ben? Or are you perhaps searching on this because of a post that Ben made on HTC? His claim was that even with an unlocked bootloader, that the eMMC could still be locked and prevent us from getting root. This seems far fetched to me.
edufur said:
Is your name Ben? Or are you perhaps searching on this because of a post that Ben made on HTC? His claim was that even with an unlocked bootloader, that the eMMC could still be locked and prevent us from getting root. This seems far fetched to me.
Click to expand...
Click to collapse
In all reality, I'm thinking this is the eventuality. Sprint knows that with root access we can circumvent the WiFi tether that they want to charge you for. They would never be OK with that.
Sent from my PG86100 using Tapatalk
Just an FYI, system is an EXT4 FS. This would require not only a custom kernel, but a lot of one offs in the way it's dealing with data. From what I've seen, this isn't what they are using.
But that's a very good find, I am looking into some of the information. Never heard of this before.
Click to expand...
Click to collapse
Given that you have taken a much closer look at the inner workings than I have, I will defer to your observation with a caveat
According to wiki eMMC supports something called Reliable Write. This suggests that the reversion capability is a part of the eMMC standard. Reliance sounds more and more like a commercial implementation of this function decoupled from a specific media type. After looking it over again, nowhere in the article about Reliance is eMMC mentioned.
Isn't that basically just wear leveling?
Click to expand...
Click to collapse
Wear leveling is a byproduct of what reliable write is doing. The difference is the ability to defer commitment of file system changes, so that a failed system update wont brick the device.
I do not know if changes made to the device are immediate and revertable (i.e., if eMMC is not told to commit a write, the changes just "go away" when its remounted). Nor do I know if reversions can be made on the fly, as we are experiencing when temp root gets deactivation.
There really isn't much information out there about this that is easy to find.
Is your name Ben? Or are you perhaps searching on this because of a post that Ben made on HTC? His claim was that even with an unlocked bootloader, that the eMMC could still be locked and prevent us from getting root. This seems far fetched to me.
Click to expand...
Click to collapse
Neither. eMMC isn't "locked" per se. HTC is using some mechanism that will revert the contents of /system to a prior state when some unknown condition is met. I do not mean to suggest that this is being done through "reliable write" or "Reliance", since it has already been pointed out by someone much more knowledgable on the subject than I that a standard EXT4 file system is being used. I honestly have no idea. I found this information somewhat by accident, and thought that if it could prove useful I should share it here.
Something is dynamically protecting the contents of /system. Once the phone is rooted, I have no doubt that this "something" will be rendered quite impotent. If it were not possible to do so in the first place, OTAs wouldn't work
Sprint knows that with root access we can circumvent the WiFi tether that they want to charge you for. They would never be OK with that.
Click to expand...
Click to collapse
The first part of your statement is true, Sprint knows full well that we can circumvent their attempts to charge us for WiFi tethering with root access. They have known this for years. They also know that in reality there is no way they can completely prevent someone from tethering their phone in one way or another. Even without root access. Ref: PDANet.
In my opinion, this protection of the eMMC contents was designed to reduce support costs from failed OTA updates bricking phones, and perhaps as protection against malware that can attain root, not unlike what Temp Root does.
I am not as paranoid as some here and refuse to accept that this was done specifically to thwart efforts to root the phone. The vast (and i mean VAST) majority of people who buy this phone will never even consider rooting the devices. This same majority has a subset of people that are easily stupid enough to screw up an OTA update or download and install malware.
I will take it a step further and opine that the only reason HTC is unlocking the bootloader is because we are such a minority AND that by tinkering with an unlocked device, we are actually helping HTC improve their product. They would rather have a more appealing facebook page than worry about losing a minuscule fraction of wifi tethering income.m Moreover, take a good look at where Sprint stands in the market, and what they have done recently to improve their position. They are doing a lot of really cool things, and have taken impressive steps to improve customer service and corporate image. That they would allow this bashing of HTC to continue unabated over a handful of tethering dollars is unlikely.
I appreciate your canter, very informative. A thanks will come your way.
Sent from my PG86100 using Tapatalk
Does pdanet allow wireless tether? I didn't think it did.
Sent from my PG86100 using Tapatalk
Nutzy said:
Does pdanet allow wireless tether? I didn't think it did.
Sent from my PG86100 using Tapatalk
Click to expand...
Click to collapse
It doesn't act as a hotspot, no.
Sent from my PG86100 using XDA App
Nutzy said:
I appreciate your canter, very informative. A thanks will come your way.
Sent from my PG86100 using Tapatalk
Click to expand...
Click to collapse
Much appreciated!
Sent from my PG86100 using XDA App
So, I would be interested in hearing more thoughts on this. Is the eMMC independent of the OS? In other words, would a custom ROM have to obey and work with the eMMC? Or could a custom ROM be made to either disable the eMMC or make it do what we want?
edufur said:
So, I would be interested in hearing more thoughts on this. Is the eMMC independent of the OS? In other words, would a custom ROM have to obey and work with the eMMC? Or could a custom ROM be made to either disable the eMMC or make it do what we want?
Click to expand...
Click to collapse
I think you're misunderstanding this. The eMMC is the memory inside the device that everything is stored on. It replaced the old NAND chips in older devices.
The OS is stored & runs off of eMMC memory, it's not independent. If you were to 'turn off' the eMMC the device would do nothing. A lot of the security features available on the chip itself probably aren't in use. HTC has been using their own form of write protection since early last year, even on the NAND based Evo 4G. I'd stake a bet they're using the same system here, and we just need to find a way to flash the ENG bootloader like we did last year to get around it.
I agree with you. reliance is setup to ward against "unauthorized" changes to the /system partitions. i believe the developer community takes way too deep a look at each action made by a corporation (htc) and view them as "big brother", when infact most changes are actually approved, reviewed, and committed by someone in accounting with no technical skills whatsoever. these people are forced to look at the bigger scheme of things and make a decision about it (after working for sprint for almost 2 years now...i can tell you how many decisions are literally made by someone who has no idea what the heck he is making decisions on).
instead of looking at them "trying to stop the development community from unlocking wireless tether" look at them as a CEO (who most of the time has no technical knowledge) and a PR rep (who really only cares about how their company is viewed) and using this kind of encryption is only there to "safeguard" their devices against attacks.
one would think the secret to perm rooting the device is triggering the reliance write function so it commits the changes instead of reloading them. if /system doesnt get changed unless theres an OTA of some sorts....theres more than likely a hash table that reliance would check against to verify...so an OTA would need to write to that table first, then make the changes....
more than likely some other noob has already said something along those lines and been flamed for it as well...just throwing it out there....
newkidd said:
.........
one would think the secret to perm rooting the device is triggering the reliance write function so it commits the changes instead of reloading them. if /system doesnt get changed unless theres an OTA of some sorts....theres more than likely a hash table that reliance would check against to verify...so an OTA would need to write to that table first, then make the changes....
........
Click to expand...
Click to collapse
that stuck out in bold to me..... hmmmmmm
I probably was overlooking what eMMC was, however based on the links the user gave, I later learned a little more about its potential. It would appear that HTC is doing something along the lines of the operations expressed in the link. And if they are not fully replicating efforts, it would be a shame. I like the concept of wear leveling and efficient read/writes. It would be my hope that we could integrate all those functions within a custom rom.
I found a page on the Micron site on eMMC. In the tech notes section there are informational downloads for just one chip. Specifically, the Qualcomm QSC6695
You have to register to download them. A process I have already started. Their site claims it takes a half hour to register a new account.
Once I have the PDFs, I will attach them to the OP.
I don't know if this is the chip the evo 3d is using, but if it is these may prove beneficial to have.
EDIT: Nevermind. i'd have to sign an NDA first.
EDIT: Although, this looks interesting.
Geniusdog254 said:
A lot of the security features available on the chip itself probably aren't in use. HTC has been using their own form of write protection since early last year, even on the NAND based Evo 4G. I'd stake a bet they're using the same system here, and we just need to find a way to flash the ENG bootloader like we did last year to get around it.
Click to expand...
Click to collapse
Perhaps, but a hint at the design really tells me that it would only make sense to offload this protection to the eMMC. Posted a link just a minute ago with the eMMC "enablement" model in PDF form. Interesting read...
CyWhitfield said:
I found a page on the Micron site on eMMC. In the tech notes section there are informational downloads for just one chip. Specifically, the Qualcomm QSC6695
You have to register to download them. A process I have already started. Their site claims it takes a half hour to register a new account.
Once I have the PDFs, I will attach them to the OP.
I don't know if this is the chip the evo 3d is using, but if it is these may prove beneficial to have.
EDIT: Nevermind. i'd have to sign an NDA first.
EDIT: Although, this looks interesting.
Click to expand...
Click to collapse
VERY interesting link & read for sure
CyWhitfield said:
The first part of your statement is true, Sprint knows full well that we can circumvent their attempts to charge us for WiFi tethering with root access. They have known this for years. They also know that in reality there is no way they can completely prevent someone from tethering their phone in one way or another. Even without root access. Ref: PDANet.
In my opinion, this protection of the eMMC contents was designed to reduce support costs from failed OTA updates bricking phones, and perhaps as protection against malware that can attain root, not unlike what Temp Root does.
I am not as paranoid as some here and refuse to accept that this was done specifically to thwart efforts to root the phone. The vast (and i mean VAST) majority of people who buy this phone will never even consider rooting the devices. This same majority has a subset of people that are easily stupid enough to screw up an OTA update or download and install malware.
I will take it a step further and opine that the only reason HTC is unlocking the bootloader is because we are such a minority AND that by tinkering with an unlocked device, we are actually helping HTC improve their product. They would rather have a more appealing facebook page than worry about losing a minuscule fraction of wifi tethering income.m Moreover, take a good look at where Sprint stands in the market, and what they have done recently to improve their position. They are doing a lot of really cool things, and have taken impressive steps to improve customer service and corporate image. That they would allow this bashing of HTC to continue unabated over a handful of tethering dollars is unlikely.
Click to expand...
Click to collapse
I completely agree with all of that. Other carriers have taken many steps to try to prevent wireless tethering. They've asked google to filter certain apps from the market from their customers, they've sent out letters to their customers who they suspect of tethering, they've used ECM's to try to stop it.
But Sprint...they've been remarkably silent on that front. Hell they don't even seem to plan on putting any usage caps in place. In my opinion, I suspect that Sprint wants to be different from the other carriers. They can't outright allow tethering because people would go nuts with it and it would saturate their network. Instead they have this approach of telling you that you can't do it without paying extra, but they look the other way when you do.
I don't know if I fully agree on why HTC locks the phone so tight though. I mean they really went out of their way to make sure nobody touches it. There could have been far more simple countermeasures in place to prevent malware yet still be open to somebody who has physical access to the phone.
It can't be that Sprint insisted on it being that way, otherwise Sprint would have insisted that the Nexus S be fully locked, so I don't believe that this is a carrier issue at all, at least not as far as the Evo 3D is concerned.
One of my suspicions is that HTC may make a profit off of having certain apps installed, much in the way that PC OEM's get paid to preload different apps (e.g. norton.) It could be that they want to make sure that you can't remove them. However that profit they make off of these apps may be significantly offset by having a really negative facebook page, hence the decision to unlock.
Hard to say really.

Would you hardware mod your NT? With other noobish bootloader conversation.

The thread that Adam and others are using for development level conversations about the unlocking of the bootloader is awesome, but it goes over the head of many of us enthusiastic noobs.
This thread is so that we can try and move our questions and confusion to here so as to un-clog the other thread.
It looks like the only way to bypass the bootloader we have seen so far, is to perform a hardware mod. After the hardware mod, we should be able to boot from SD or reinstall the ROM.
Discuss!
If i first get to have the chip from outside USA i surely will go for it, if its totally worthy.
Really, for me, it depends on just how involved the hardware mod is---whether it requires soldering and such.
Sent from my rooted Nook Tablet using Tapatalk 8)
I wouldn't mind doing it myself, no matter the level of difficulty. Also a send-in service by some 3rd party company (N2A comes to mind). We'll see what the success rate is, and what can be done.
I already opted for $50 more than the KF, then another $40 for a 32gb sd card... i cant see spending even mor eto buy the mod chip. At that point, i shouldve bought a better tablet.
I'll wait on a software bypass, it'll get here eventually.
Plus, i have no interest in linux. I doubt i even need ICS, but if it comes and its free... why not?
From what I have read so far the hardware mod would not be challenging or expensive. Two screws and 4 solder points. According to Adam the chip is only $3 USD, and I would assume that it could be programed with a printer port.
I already opted for $50 more than the KF, then another $40 for a 32gb sd card... i cant see spending even mor eto buy the mod chip. At that point, i shouldve bought a better tablet.
I'll wait on a software bypass, it'll get here eventually.
Plus, i have no interest in linux. I doubt i even need ICS, but if it comes and its free... why not?
Click to expand...
Click to collapse
I think the cost of the SD is irrelevant. You would have to do this with any tablet, and if your comparing to the KF then this one section of your argument is even more so invalid.
The mod chip also looks like it will be very inexpensive.
EDIT: I highly doubt ICS will be released for our tablets by B&N ever... So we will be waiting for someone like the CM team to get it working for us.
The hardware mod interests me even if solely as a way to learn a bit about embedded security. I do find it unsettling that I can't install linux on my linux computer.
I don't want to be dependent on having a boot sdcard installed at all times. A hardware modification is good ONLY if it leads to PERMANENTLY disabling the anti-hack mechanism so that the hardware modification doesn't have to be performed again every time the device is booted up.
I would accept precisely this;
Plug in some device,
Boot on sdcard,
Modify the secure boot process WITHIN DEVICE STORAGE (not within RAM),
Remove device,
Store on shelf until it is needed for another hacking project.
There are multiple options for modification of the secure boot process, in particular, replacing the signing keys stored on the device allowing us to use OUR OWN signing key to satisfy the secure boot process. Another option is to eliminate the signature check. I believe that the former will be simpler since it is just a DATA modification rather than reverse engineering.
I wonder if the hardware has an equivalent of "S-OFF" that HTC phones have? As I recall, on most HTC phones, there is a ONE BYTE EDIT that has to be made within one partition of the eMMC to TOTALLY disable all of the device's bootloader sig checks and hardware write protect. The issue was that the partition in question would be hardware write protected. Presumably, with the hardware modification for NT, a similar state *should* be achievable.
The_Joe said:
Would you hardware mod your NT?
Click to expand...
Click to collapse
Absolutely not!
I also wouldn't because you never will get your in the same condition as it was before you opened it and everybody can see that you did something to the device. You also don't have warranty anymore after that process.
Sent from my SGH-T989 using Tapatalk
Pete1612 said:
I also wouldn't because you never will get your in the same condition as it was before you opened it and everybody can see that you did something to the device. You also don't have warranty anymore after that process.
Click to expand...
Click to collapse
Whaaaaa???
Unlike you, most intelligent people will select jeweler's tools to open something like this rather than a crowbar and a sledgehammer.
Further, most people would actually perform the modification using a temperature controlled and grounded soldering iron, rather than an acetylene torch, thus no damage caused even internally.
As far as warranty goes, it is EASY to restore it to factory condition, unless you use the Neanderthal approach to hardware modification, in which case warranty provision is determined by YOUR HONESTY to only warranty it for something that YOU DID NOT CAUSE (in which case there is no morality issue to be worried over).
I also wouldn't because you never will get your in the same condition as it was before you opened it and everybody can see that you did something to the device. You also don't have warranty anymore after that process.
Click to expand...
Click to collapse
Soldering is not that big a deal once you have developed the touch. I am fairly certain that I could solder the four wires required for this chip and then later un-solder them and no one would be able to tell.
It's a great device regardless of installing a ROM. I just rooted it and installed the gapps and OTA block. I still have all the B&N stuff and everything work fine. I have all the apps loaded I want (that make sense without gps and 3g) and I WOULD hardware modify this puppy in a heartbeat to get back what I have if B&N happens to force an OTA through that took away root.
I have been having a lot of fun trying lots of different apps and schemes. The hardware mod is easy and I have already taken it apart and looked at where the soldering goes and it is back together without a hitch.
Big hand to the devs who are doing their thing so we can do ours.
Bill
Oh yes, I'm sure he had a rock and mallet in mind for the modification. How neanderthal of him to suggest that hardware modification is well out of his, and many other's, reach.
It's easy to unroot. I doubt physically messing around with the internals using any manner of ultra sophisticated museum art thief tools will leave behind changes that are unnoticeable.
Duely blundered from my thunderdolt
I would do it just to do it. I am one of those guys that just like to mod stuff for the hell of it. No, I'm not very good at creating any of the mods, but if it will make my experience better and it's not rocket surgery, I'm game.
As for the warranty issue and opening up the case, it really doesn't look like its that big of a deal. I am guessing that with a set of precision screwdrivers and a guitar pick it can be opened simply and be very nearly undetectable. Besides, if I am going to open it and solder it, I am not worried about the warranty anyway. If someone IS worried about it, then they shouldn't be doing anything that will void it.
JM2C

Please explain why some ROMs and kernels work well on some phones

This is a new concept to me. There are no PCs that "just don't like Windows". With the same hardware, a PC will run the same software at the same speed with the same stability. All SGS II i777's have the same hardware, don't they?
There are of course variations in the quality (=bin) of some components, but that just sets voltage boundaries for a given clockspeed (or in extreme cases, makes a phone defective), but this shouldn't affect how a ROM runs.
What am I missing?
Why do some problems persist on what seems to be a clean install on some phones, and don't manifest on others? Do some ROMs run better because of different recoveries? What's left after a full wipe?
Obviously, installed apps and settings will make a phone perform differently, but that aside, shouldn't all SGS II run the same?
Eckyx said:
This is a new concept to me. There are no PCs that "just don't like Windows". With the same hardware, a PC will run the same software at the same speed with the same stability. All SGS II i777's have the same hardware, don't they?
There are of course variations in the quality (=bin) of some components, but that just sets voltage boundaries for a given clockspeed (or in extreme cases, makes a phone defective), but this shouldn't affect how a ROM runs.
What am I missing?
Why do some problems persist on what seems to be a clean install on some phones, and don't manifest on others? Do some ROMs run better because of different recoveries? What's left after a full wipe?
Obviously, installed apps and settings will make a phone perform differently, but that aside, shouldn't all SGS II run the same?
Click to expand...
Click to collapse
Simple answer is no not all devices are made the same. You see this with mass produced hardware. As for your PC ref. My brother and I have the same laptop while mine loves my set up, my brother has nothing but problems with it on his system. It comes alot down to personal set up and device. This has always been the case with mobile devices and custom roms. Stock roms are set to run on all devices the same. This is not so with custom roms.
Eckyx said:
This is a new concept to me. There are no PCs that "just don't like Windows". With the same hardware, a PC will run the same software at the same speed with the same stability. All SGS II i777's have the same hardware, don't they?
There are of course variations in the quality (=bin) of some components, but that just sets voltage boundaries for a given clockspeed (or in extreme cases, makes a phone defective), but this shouldn't affect how a ROM runs.
What am I missing?
Why do some problems persist on what seems to be a clean install on some phones, and don't manifest on others? Do some ROMs run better because of different recoveries? What's left after a full wipe?
Obviously, installed apps and settings will make a phone perform differently, but that aside, shouldn't all SGS II run the same?
Click to expand...
Click to collapse
Zel pretty neatly answered this, but I will also add that modern consumer electronics interact with the immediate environment far more than our pcs do. Light sensors, 3g radios, barometers, etc. are far less deterministic than our classic closed-loop pcs. Part of this perception of flux is based on this real flux, for example one of the core features people will discuss is call/modem quality, but driver tweaking vs. actual signal strength is a pretty fuzzy battle for anyone but an electrical/firmware engineer. And just like in the pc world, when you're talking under volting and over clocking your mileage will vary.
If you are methodical and read all the materials, your phone will operate tip top. It seems to me a lot (not all) of the variances often do boil down to the users configuration.
Sent from my iPad using Forum Runner
I have fixed a lot of computers and have been flashing custom roms for a year and ill tell you, in my personal opinion, problems are 90% user error. If people would all install properly and wipe everything completely and follow everything they're told to do and read all possible material on what they're flashing they can, a huge portion of the problems would dissapear. But is that gonna happen? I hope so
Heck I make mistakes too. None of us are immune to screwing up right? Good luck all, happy flashing.
Sent from my SAMSUNG-SGH-I777 using Tapatalk 2
Eckyx said:
Obviously, installed apps and settings will make a phone perform differently, but that aside, shouldn't all SGS II run the same?
Click to expand...
Click to collapse
Yes, they should and they do, absolutely with every ROM, stock or custom.
If they're purely, properly installed, they're all the same.
By "purely", I meant completely, virgin-like ROM without any add-on.
If you choose to install somethings else, to customize your needs or set it up the way you want, then it's a whole different ball game.
zelendel said:
Simple answer is no not all devices are made the same. You see this with mass produced hardware. As for your PC ref. My brother and I have the same laptop while mine loves my set up, my brother has nothing but problems with it on his system. It comes alot down to personal set up and device. This has always been the case with mobile devices and custom roms. Stock roms are set to run on all devices the same. This is not so with custom roms.
Click to expand...
Click to collapse
Just to chime in on the Windows thing - we buy a standard build HP computer at work for all developers. But if you put the same Windows disk in two of them and boot and install accepting all the defaults, there will always be slight variations in the way it configures itself between the two. It's probably the hardware detection that does it, but I swear you could get two of the same build lot and you'd STILL get something that didn't set up the same way. Sunspots? Power surge during the process? I dunno, but it does vary
You cant change the disk. Your network adapter has a MAC adress on it... windows will know something has changed. motherboard also has one.
A PC component are not the same at all.
You can buy a good I7 2600k or a bad I7 2600k. There are revisions of the very same model of CPU, memory, everything and its really hard to make 1 equal another.
Another thing is that one smartphone is a lot more delicate piece of hardware and the most important, has limited power to it components.
That makes harder to change anything on it. A small change could lead you to a failure.
just blame it on the ghosts in the machine and be done with it
votinh said:
Yes, they should and they do, absolutely with every ROM, stock or custom.
If they're purely, properly installed, they're all the same.
By "purely", I meant completely, virgin-like ROM without any add-on.
If you choose to install somethings else, to customize your needs or set it up the way you want, then it's a whole different ball game.
Click to expand...
Click to collapse
This is not the case. I have tested with two different EVO 4Gs and two different SGS IIs - of the same hardware revision, even.
I performed the same steps to root and ROM both of the sets of phones, and put the same ROM on them. There were no other apps installed, nor themes/addons. I then used each as my phone for a week, making sure that I installed identical apps and even synced app data.
Both performed differently. My i777 is the faster of the two, but my EVO 4G was the slower and more bug-prone.
Yes, this is anecdotal evidence, but I at least am convinced. Take it as you will.
Also the phones are made with lower quality checks than desktop. ie I can oc my sgs 1 up to 1.6GHz but uv of -25, the phone well restart. But other people can't go more than 1.2 and cab apply a uv of -150 on the same step.
This is a fact. Think if every phone it's done with high quality checks the price of the device will raise pretty high.
Once of the frustrating things is that the people responding on the forums, (ROM devs and regular users) often try to have it both ways.
If you're experiencing something they aren't, it's obviously your fault because it doesn't happen on their one phone so it can't possibly be anything but your own fault. Go wipe 50 times and do other things that will take 3+ hours each time and don't come back until you do.
Oh... you've done all that already and still having the issue? Oh well, all phones are different... tough luck, bro. I'm not going to spend any time on anything my one phone doesn't do.
(Never mind that a dozen other people have reported the exact same issue, and another several dozen are experiencing it but are too scared to post about it because they see how everyone else is being flamed, castrated or even banned for daring to suggest a bug.)
One way or another, it seems a convenient way to blow off users having legit issues.
While I suppose there might be slight deviations in components once in a blue moon, I think the "all phones are different" excuse is more often than not used as a way to easily dismiss people and issues without helping. It's been repeated over and over for so long now, the majority take it as "fact" without really putting much thought into how it could possibly be as widespread and dramatic as they're pretending it is.
I don't buy it.
Even just with modems... everyone's like, "Oh well all phones are different... some modems work better on some phones, or in different areas". What kind of BS is that? How on earth could any phone manufacturer then create a mass-market phone that worked across the country without hacking? I'm sorry, but you can't chastise and criticize the manufacturers for not producing the universal "uber-phone" that works great anywhere while at the same time admitting that "all phones are different" and therefore require hacking for your specific flavor or region. It's hypocritical.
I think it's very, very rare that actual hardware differences between the exact same model phone account for issues people experience, and is more often than not either user or dev (as much as they try to paint themselves as infallible gods) error that they don't want to bother with. "All phones are different" = "F-off, I don't want to deal with this"
sremick said:
Once of the frustrating things is that the people responding on the forums, (ROM devs and regular users) often try to have it both ways.
If you're experiencing something they aren't, it's obviously your fault because it doesn't happen on their one phone so it can't possibly be anything but your own fault. Go wipe 50 times and do other things that will take 3+ hours each time and don't come back until you do.
Oh... you've done all that already and still having the issue? Oh well, all phones are different... tough luck, bro. I'm not going to spend any time on anything my one phone doesn't do.
(Never mind that a dozen other people have reported the exact same issue, and another several dozen are experiencing it but are too scared to post about it because they see how everyone else is being flamed, castrated or even banned for daring to suggest a bug.)
One way or another, it seems a convenient way to blow off users having legit issues.
While I suppose there might be slight deviations in components once in a blue moon, I think the "all phones are different" excuse is more often than not used as a way to easily dismiss people and issues without helping. It's been repeated over and over for so long now, the majority take it as "fact" without really putting much thought into how it could possibly be as widespread and dramatic as they're pretending it is.
I don't buy it.
Even just with modems... everyone's like, "Oh well all phones are different... some modems work better on some phones, or in different areas". What kind of BS is that? How on earth could any phone manufacturer then create a mass-market phone that worked across the country without hacking? I'm sorry, but you can't chastise and criticize the manufacturers for not producing the universal "uber-phone" that works great anywhere while at the same time admitting that "all phones are different" and therefore require hacking for your specific flavor or region. It's hypocritical.
I think it's very, very rare that actual hardware differences between the exact same model phone account for issues people experience, and is more often than not either user or dev (as much as they try to paint themselves as infallible gods) error that they don't want to bother with. "All phones are different" = "F-off, I don't want to deal with this"
Click to expand...
Click to collapse
Great so what you're saying is everyone that says this (especially devs) are a bunch of dicks. Nice.
Sent from my SGH-I777 using Tapatalk 2
Nick281051 said:
Great so what you're saying is everyone that says this (especially devs) are a bunch of dicks. Nice.
Click to expand...
Click to collapse
That's actually not what I said at all, but I know it's a lot easier to summarize my post into one sentence that makes me look like an ass, than actually look at the points that I made.
sremick said:
That's actually not what I said at all, but I know it's a lot easier to summarize my post into one sentence that makes me look like an ass, than actually look at the points that I made.
Click to expand...
Click to collapse
I read the whole thing and that's exactly what it says.
Sent from my SGH-I777 using Tapatalk 2
Nick281051 said:
I read the whole thing and that's exactly what it says.
Click to expand...
Click to collapse
Nope. But let me simplify it for you:
1) There can't be the level of hardware variation that people claim. It's not seen in other electronics, and it's not been seen before the recent generation of smartphones. The assumption of its existence is a recent phenomenon however if it were real we'd see it everywhere, on other electronics other than phones.
2) If it were to really exist, people would lose the ability to legitimately blame the phone manufacturers for 90% of what they currently give them crap about, especially in regards to making a quality stock ROM. What are they expected to do, create thousands of variations of stock ROMs, one for ever county in the USA to compensate for this accepted "all phones are different" theology?
3) Due to it being repeated over and over and simply assumed to be true without any actual evidence to the fact, it's become a convenient way to dismiss user issues... even in the face of mounting evidence to the contrary on a specific issue. This unfortunate trend causes lots of friction between users and devs. Even if it were true, it's now become an over-used dismissal without allowing for the chance that the user might be right.
People can't have it both ways, but right now there's a lot of hypocrisy. If it's true, there's been no evidence actually shown... just anecdotal experiences that could be chalked up to any number of other things. And whether it's true or its not, either way a massive amount of thinking and behavior would then have to then change... but right now, people behave like it's true and not true at the same time, which is nonsensical and frustrating.
The simple answer is, there is no answer. Its the nature of the process.
I've had one click roms fail the 1st attempt only to succeed the 2nd without even closing the Odin just reconnect the phone.
Either you accept that and have fun with it. Or stick to stock and move on.
Sent from my SGH-I777 using xda premium
sremick said:
Nope. But let me simplify it for you:
1) There can't be the level of hardware variation that people claim. It's not seen in other electronics, and it's not been seen before the recent generation of smartphones. The assumption of its existence is a recent phenomenon however if it were real we'd see it everywhere, on other electronics other than phones.
2) If it were to really exist, people would lose the ability to legitimately blame the phone manufacturers for 90% of what they currently give them crap about, especially in regards to making a quality stock ROM. What are they expected to do, create thousands of variations of stock ROMs, one for ever county in the USA to compensate for this accepted "all phones are different" theology?
3) Due to it being repeated over and over and simply assumed to be true without any actual evidence to the fact, it's become a convenient way to dismiss user issues... even in the face of mounting evidence to the contrary on a specific issue. This unfortunate trend causes lots of friction between users and devs. Even if it were true, it's now become an over-used dismissal without allowing for the chance that the user might be right.
People can't have it both ways, but right now there's a lot of hypocrisy. If it's true, there's been no evidence actually shown... just anecdotal experiences that could be chalked up to any number of other things. And whether it's true or its not, either way a massive amount of thinking and behavior would then have to then change... but right now, people behave like it's true and not true at the same time, which is nonsensical and frustrating.
Click to expand...
Click to collapse
No you're actually wrong because sometimes things CAN go wrong with flashing stuff. And modems in fact do work better in some places than others even though they try to make them universal it is very difficult to do that. Also, if a dev doesn't see a problem, how is he supposed to fix it? Riddle me that one. Also, restoring should take at most an hour with something like titanium backup. So 3+ hours is bull.
Sent from my SGH-I777 using Tapatalk 2
Nick281051 said:
if a dev doesn't see a problem, how is he supposed to fix it?
Click to expand...
Click to collapse
He can start by not flaming the several people experiencing it (which scares away others also experiencing it, self-validating him and making him think the problem is less-common than it really is)
There's a difference between a problem being simply difficult for a dev to pin down and solve, and just attacking any user who dares bring up an issue that the dev himself isn't experiencing. Therein lies the hypocrisy, though: if "all phones are different", then a dev has to accept that a problem a user is experiencing might not be the result of stupidity on the user's part, and might actually be something in the ROM that simply didn't surface on the dev's phone, but is still something that needs to be addressed at the ROM level.
I totally accept that an issue the dev can't reproduce on his own phone is harder to resolve. But there are several devs who even make ROMs for phones they don't own anymore, and still manage to work with users to resolve issues. App devs do the same thing to resolve incompatibility bugs/issues with specific phone models the app dev don't own.
Also, restoring should take at most an hour with something like titanium backup. So 3+ hours is bull.
Click to expand...
Click to collapse
Not quite. Just the backup alone takes 45+ mins on my phone. The restore is the same amount, so you've got 1.5 hours right there without taking into account everything else that TB doesn't handle that needs to be reconfigured by hand, or the actual flashing, and tracking down other surprises. And because I apparently am one of the few people who care about their data, I also back up my internal SD... just in-case something goes wrong. I know the wipe shouldn't touch it... but "**** happens" and if something goes wrong during a flash and I didn't have a backup, it'd be my own fault and I'd be subsequently flamed for that. So I play it safe.
Once again, a case of users for whom its quick trying to tell everyone else they're full of BS. I accept that for some users, whether it's the # of apps, amount of data, or whatever, it goes fast for them. Lucky them. Why can't they accept the fact that it's not "10 mins" for everyone like they spread around and flame others for suggesting otherwise?
Ok Here is the thing. I have personally been flashing and making custom roms for phones as far back as the Motorola razor v3. I have multiple phone running the exact same firmware, set up the exact same way and the have run totally differently. If all phone were made equal then there would be no returns due to issues the phone had as they would have all run the same and acted the same as the units that the OEM quality control tested. Lets take the Black Jack 2 as an example. The BJII was a WM phone that would self corrupt the system/media folder. This prevented any ringtones from working and the OS had to be reflashed. All this running on Stock firmware. This didnt happen to all of them, but became a well known glitch to anyone that did cell phone troubleshooting and repair. Did you ever stop to think why OEM and carriers dont use the roms from places like XDA? This is because things here are always under development. There will always be bugs. In the end the developers are making things for their phones and are nice enough to share it for others to use. Some do keep making roms for phones they dont have as this is overly not hard to do. They do this just to be nice. Except for the hardware drivers most of the under lying OS is all the same.
If a dev cant reproduce it then there is no way for them to fix it with the way people tend to report bugs. (The wrong way without the proper info)
Now lets jump to the present. I have 4 phones sitting on my desk. 2 are the HTC Inspire and 2 are the Samsung Captivate. Both running the exact same rom and apps, but guess what. They run very differently. On 1 Inspire and 1 captivate, I can OC to almost double, while the other 2 cant handle more then 1.2 over clock.
One of them also doesnt like the AOSP based software while the others are fine.
As for your backing up and restoring. It can take a long time if you have a TON of apps. Flashing custom roms are not for everyone. If you dont have the time or the want to learn something then you are doing it for the wrong reasons.
sremick said:
He can start by not flaming the several people experiencing it (which scares away others also experiencing it, self-validating him and making him think the problem is less-common than it really is)
There's a difference between a problem being simply difficult for a dev to pin down and solve, and just attacking any user who dares bring up an issue that the dev himself isn't experiencing. Therein lies the hypocrisy, though: if "all phones are different", then a dev has to accept that a problem a user is experiencing might not be the result of stupidity on the user's part, and might actually be something in the ROM that simply didn't surface on the dev's phone, but is still something that needs to be addressed at the ROM level.
I totally accept that an issue the dev can't reproduce on his own phone is harder to resolve. But there are several devs who even make ROMs for phones they don't own anymore, and still manage to work with users to resolve issues. App devs do the same thing to resolve incompatibility bugs/issues with specific phone models the app dev don't own.
Not quite. Just the backup alone takes 45+ mins on my phone. The restore is the same amount, so you've got 1.5 hours right there without taking into account everything else that TB doesn't handle that needs to be reconfigured by hand, or the actual flashing, and tracking down other surprises. And because I apparently am one of the few people who care about their data, I also back up my internal SD... just in-case something goes wrong. I know the wipe shouldn't touch it... but "**** happens" and if something goes wrong during a flash and I didn't have a backup, it'd be my own fault and I'd be subsequently flamed for that. So I play it safe.
Once again, a case of users for whom its quick trying to tell everyone else they're full of BS. I accept that for some users, whether it's the # of apps, amount of data, or whatever, it goes fast for them. Lucky them. Why can't they accept the fact that it's not "10 mins" for everyone like they spread around and flame others for suggesting otherwise?
Click to expand...
Click to collapse
I agree with your point about the dev/user relationship. It can get pretty hostile sometimes which is completely unnecessary. Recently, I've been trying to emphasize the respect rule, regardless of who you are. Patience and cooperation can go a long way. And as zelendel said, sometimes people do report stuff the wrong way, but there's no need to be rude about it. Again, patience goes a long way.
Back to the main thing though, I'm not entirely sure how 2 fresh-out-of-the-box phones would work with the same settings and ROMs, but as others said, environmental and external factors can stress the phone and stuff just goes wrong. My phone, for example, would not operate the same as a fresh-out-of-the-box SGS2, even if you put the exact same stuff on it. Stuff just starts freaking out, and I'm pretty sure my phone is having hardware issues :[

Categories

Resources