Here's what bloatware does to your phone

[A.Nonymous]

This thread originate from a thread complaining about bloatware and what effect it was having on the life of their phones. So I figured I'd run some actual tests and post the results here. I know some people actually use at least some of the programs bundled with their phones, but not everyone does. For the sake of this experiment, I'm classifying the following programs as "bloatware" even though I do use some of them quite freqently:

Amazon MP3
CityID
App Sharing
Facebook
FM Radio
Footprints
Friendstream
Google Search
Latitude
My Verizon Mobile
Peep
Skype Mobile
Talk
Teeter
VZW Navigator

In theory we ought to get a phone with no apps or just the apps that Google bundles with Android (Market, GMail, etc...) and none of the other crap. If I'm on Facebook, I'll download the Facebook app. If I'm on Twitter, it's the same thing. Now, I'm not arguing with anyone who thinks they should be allowed to remove these apps from their phone if they so choose. There are no arguments there at all. All I'm interested in is what effects this "bloatware" is having on the phone's performance.

Here I'm running a rooted DInc with 2.1. None of the software bundled with the phone has been removed. I'm running 2.1 because if these apps were going to impact performance, this is the OS that it would do it on as 2.2 is more efficient (supposedly) at managing memory and resources. I wanted to pick something with the best chance of experiencing any sort of
lag. Besides the bloatware (some of which I found on the market and downloaded since it wasn't on 2.1 originally) I installed Watchdog Light and Drocap. That's it. You can see right after boot that 30% of the CPU is being used by background apps and 162736 mb of memory are available to be used by apps. (Strangely enough, 26.2% of this is the Android media process.) Now, the system process under Eclair is using about 7-8% right after boot. What are these bloatware apps using 10 mins after the phone has been sitting at idle and everything has been given a chance to kick into gear?

10 mins later, here's the stats - 30.0% CPU time (still the media service acting up) and only 3548 kb of RAM are available. So, let's see which of these bloatware apps are doing the damage.....

Amazon MP3 - Idle, 0.0%, 11.7 mb
CityID - Not running
App Sharing - Not installed under eclair
Facebook - Not running
FM Radio - Idle, 0.0%, 10.9 mb
Footprints - Idle, 0.0%, 11.1 mb
Friendstream - Not running
Google Search - Not running
Latitude - Not running
My Verizon Mobile - Not running
Peep - Not running
Skype Mobile - Running as service, 0.0%, 12.4 mb
Talk - Not running
Teeter - Not running
VZW Navigator - Idle, 0.0% 11.5 mb of RAM

So, of all the bloatware apps, none of them are impacting the processor in any way at all while the phone is idle. In fact, the Android System process is what is causing the most CPU usage under Eclair along with the media process.

Now I've flashed my phone with a custom ROM running 2.2 and I've removed all of the aforementioned bloatware. I also did a factory reset on the phone so there are no apps that I've downloaded running. Again, the only apps I put on here from the market are Drocap (to capture screenshots) and Watchdog (to measure performance). This is basically what people are saying they want from VZW and other carriers - A phone with no bundled software. So this is what I did. I removed all apps using Titanium Backup, removed Titanium and rebooted the phone.

You can see what an Android phone with no bloatware and no other apps on it runs right after boot. 5.9% of the CPU is being used by background apps/processes and 69536 kb of memory are free. Here's the top running apps (nothing showing up of course). After 10 mins, the CPU % settles down to .4% and memory goes up to 94924kb at idle.

So, in conclusion, you can see that Froyo is much more efficient than Eclair because the System process and the Media process that seem to chew up processor time on Eclair are fixed under Froyo. The phone running Froyo runs faster, but this is only because it's Froyo, not because it has no bloatware running on it. This accounts for the difference in CPU time and RAM between the two. The bloatware that's running under either OS makes no difference at all.

Bottom line, quit complaining about how bloatware is damaging your battery life and your phone's responsiveness because it just isn't. (The fact that I took all that time to put that together just proves I have no life I think.)

 

[amlothi]

It's hard to take your post seriously when your tests were not done scientifically at all. I'll give you an A for effort, but that is all.

You need to:

1) Have repeatable results. Not just a single data point.

2) Compare apples to apples, not apples to oranges. You need to have the same phone, same OS version, same ROM and only change 1 variable (the "bloatware").

3) You need to track the performance over time, not just at a single timepoint. Do the bloatware apps start up on their own, or sync on their own during the course of the day? What resources to they use then?


Yeah, it might not be easy or feasible to do the tests the right way. I get that. However, your data at this point is basically the same as me saying:

1) Yesterday was the 24th. It rained.
2) Today is the 25th. It didn't rain.

Conclusion: It only rains on even numbered days.

 

[copestag]

it rained here yesterday also..... that theory makes perfect logical sense

 

[suXor]

C for effort, A for caring.

As was said, more time needs to be spent and variables limited. Go with a specific rooted ROM and test with and without the pre-installed applications. Allow the test period to run longer than a few minutes. Also, RAM takes power. Holding the runtime for these apps in memory requires power to the RAM. However, I am not sure if power consumption by RAM changes based on the usage of that RAM (amount in use, R/W to it, or if it is a constant). In the end, RAM power probably is a very small portion of battery consumption, but I say this to illustrate other variables not captured here.

Your battery is going to be consumed mostly by the following, in order of level of consumption:
Display
Radio (Carrier network)
Wi-Fi
CPU

9 times out of 10, the battery is being used because of an application using those features. Even GPS is not that problematic on its own. It's merely a receiver. However, the fact that it is communicating over a data network for further calculations, and being used because an application needs it, and likely is an application being viewed with the screen on, makes everyone point at GPS as the problem.

/rant

 

[grainysand]

It's hard to take your post seriously when your tests were not done scientifically at all. I'll give you an A for effort, but that is all.

You need to:

1) Have repeatable results. Not just a single data point.

2) Compare apples to apples, not apples to oranges. You need to have the same phone, same OS version, same ROM and only change 1 variable (the "bloatware").

3) You need to track the performance over time, not just at a single timepoint. Do the bloatware apps start up on their own, or sync on their own during the course of the day? What resources to they use then?


Yeah, it might not be easy or feasible to do the tests the right way. I get that. However, your data at this point is basically the same as me saying:

1) Yesterday was the 24th. It rained.
2) Today is the 25th. It didn't rain.

Conclusion: It only rains on even numbered days.

I like this post.