XDA Forum Member Opinions on I/O and Android M

XDA Forum Member Opinions on I/O and Android M

Google’s I/O event this year brought many announcements that affected us as developers. To better understand just how our community felt about the new products and services, we reached out to some of you in the forums. We received many great responses; here are just a few of the things you felt were important.



“There’s going to be a huge fuss over the smaller but more specific changes to Android, and they’re going to grab the media limelight for the most part. For example, I can already picture the myriad of stories covering disappointing results from things like Doze, or a million apps ungracefully crashing from declined permissions, but as usual, most of these will affect the ‘power users’. What’s actually caught my attention lies more in line with Google’s wider plan.

Firstly, Google (aiming for ‘The Next Billion’ I’m assuming) looks to be focusing on continuing to make things easier for new developers. The new features and changes to Android Studio are pretty huge, and elements like the reduced build times and scalable vector images go a long way to ensure the whole developing process looks far less imposing. Material Design in itself provides a lot of answers to the eternal back and forth question for many developers, namely, how to make a clear but functional UI, but adding to that this year with the Android Design Support Library is the icing on the cake. Want to add a new button? Here’s one, use this. Now everyone can get used to what this icon does, regardless of the app the you’re in. That’s really important for making Android look less convoluted to all its users, and hopefully the addition of more support annotations will help with the reliability of each app and the speed at which they can be tested and ultimately brought to market. Again, for the new developer, this is brilliant.

Secondly, and this is slightly more implied, to me it looks as if Google is slowly adding more features to AOSP that previously you would have only seen in OEM-skinned UI’s. Besides having almost admitted as much, this could be a tactical move rather than a redundant competitiveness for Nexus users. If more of these features are already built into ‘Vanilla’ Android, then the companies that want to add their own flavor on top of the operating system don’t have to spend as much time (and money) developing them themselves, and that means faster updates. Timely software updates have been a very vocal complaint with Android almost since it’s inception, and if this feature-rich AOSP direction continues, combined with the separation and rapid update schedule of Google Play Services, Google could really be on to something.

Now, I’m not foretelling the death of third-party Android builds; manufacturers will always want to add more features, their own style, and fundamentally attempt to differentiate themselves from their rivals, and that’s part of what makes Android so great. After all, we may not have ever had an (admittedly basic and potentially buggy) implementation of multi-window functionality without the efforts of Samsung, LG et al, but in my mind, anything that can take some of the feature development onus away from said manufacturers is beneficial to everyone.”



“The truth is I was expecting the event to be bigger and more important to the world of android. I’d love if a new nexus 5 was announced this summer. My favourite announcement was the “doze mode”, even if most people didn’t seem to like it. Smartphones these days have great performance, camera, design etc, but the only thing that is maybe getting even worse is the battery, take a look at the S6 for example. So, if doze is gonna make battery better, I like it.

My only concern for Android M, is that Android is becoming less clean. It is now more like a custom rom, this is NOT necessarily a bad thing. What is bad is that manufacturers are going to take that new and “heavier” stock android and add more and more features that will make it run slower. Some of those features may also be the same as stock Android’s so we’ll basically get double all the features or apps. (Android pay-samsung pay, two browsers and other apps, and many battery saving modes etc).”



“App permissions: As a user this is a welcomed change as I get to choose what I want to share and my data is finally under my control. As a developer I am worried will the new feature keep bugging the user (if he selected “no”) for that one little network connectivity permission so that I can display my ads for revenue in my file manager app, will it cut in on my ads revenue? (turning down the network permission could be taken out)

Photos (purely a user’s POV): unlimited storage up to 16mp? Sure they compress the image, but this is ok as long as there won’t be any perceivable change in quality of the image. The ability to get a link to photos will be extensively used by all, as this makes it incredibly simple to share the selected album.
NDK support in android studio (a dev’s POV): finally! this is long overdue, sure Android depends a lot on java, but games are still done better using the old native languages (c++). Eclipse has been missing a lot of things from android studio (sure plugins help but they can only do so much) it will be really interesting to see how NDK will evolve with the studio.”



“Personally, I feel that they have good intentions with the new features that Android M has to offer. Google Photos was separated from Google+ FINALLY! It’s also very user-friendly and easier to use two accounts. The battery saving technique they have come up with is actually rather good. I saw folks complaining that they need to focus on Screen ON Time instead of Screen OFF Time. To me, better SOFFT allows for better SONT.

Users voiced their concerns on LP and it seems as though the team listened. Did they listen to everything? Probably not, but what M has to offer is a good start so far. People however have a lot of concerns when it comes to the new features they have decided to put into M. For me personally, my only concern is with the Automatic Backup feature! User id and password information is sent to Google in plain text format which can be intercepted and there goes your Google account. So far I haven’t seen any secure way mentioned for the data being backed up. I don’t mind basic settings being saved on a server such as XDA Premium App: Dark Mode, notification settings; just to name a few examples. More sensitive data however such as login info, personal notes(Memo app, note app) etc, should stay on the device. I understand how some may find this feature useful in a day-to-day world where people are too busy to save everything they need on an SD Card or their own personal cloud system. However for the majority of us deeper Android users, this can be a nightmare. Now I know a lot of people are going to say that you can just turn it off. Well you can, but from what we saw from the I/O, it’s on automatically and that is very dangerous especially if you’re using an OEM device that receives an OTA. Upon reboot after M update is installed, is my data going to begin backup without me telling it to? That’s what it appears like.

This reminds me of one of the things I hate the most when it comes to Google Play Store; Automatic Updates when connected to Wifi. I flash custom Roms a lot on my device; A LOT and yes I need help. When setting up a new Rom, I like to bring my configuration of apps from the previous set up because I don’t want to go through the trouble of selecting each app to install again. Yes there is Titanium Backup but that has its slew of issues when switching Roms; not knocking the app, it’s great, but it’s not for me. So after I put in my information and I’m brought to the launcher home screen, Play Store goes to work and installs the apps that I had on my list. It also automatically adds apps needing to be updated on the list. This can be a pain if you’re OCD about BLOAT and happen to be trying out a custom OEM Rom that was not debloated to your liking.”



“Since we started to work smart devices into our lives, the one constant we’ve had is a bad to mediocre battery life. Long gone are the days when a charge could last you a week of moderate to heavy use from our old indestructible Nokias. The two features that most hold my interest are the granular app permissions control and the Doze mode.

The first one we’ve all been waiting for since AppOps was “leaked”, and is also the one thing iOS has been ahead of Android for a while. This way we can keep unruly apps under control (looking at you, Facebook!) without a need for root, which may not affect us here in XDA, but not everyone can get root on every device, so this is great news.

And Doze, which is basically the Stamina mode many OEMs, namely Sony, have in their devices. If this one can manage to wrangle Google Play Services, which I’ve found to be the biggest offenders in the battery department, I think M could be a great OS.”



“The things that really peaked my interest is all the back-end stuff in regards to Android M, especially when it comes to performance and battery life. Doze is a feature I’m really excited about because at times some apps can go rogue due to poor coding both for that particular device or in general. Google finding a way to save battery life even further without totally shutting down major services during standby/sleep is a great achievement by itself.

The ability to manage your permissions is something I think many people have been waiting for including myself. Some people like it for security purposes but my reason again is that it provides users with a way to deal with wakelock problems. Many apps have wakelock issues that can significantly effect battery life and having the power to effectively turn off specific permissions within an app can possibly help rectify any of these problems.

These two features should help users get more out of their devices while maintaining the great user experience that Android M will provide.”



“Project ARA progress was my favourite announcement because I strongly believe that modular computing (not only mobile) will be the future and this is the first big and real step in this direction. I was proposing this to my OEM-friends in Shenzhen for around.the last 5 years whilst in the mean time I can see it’s worked on several fronts from different angles of attack. The Nexpaq would be my favourite example while it even has the potential to prevail over ARA because of it’s wider focus.

Doze seems to be quite a hit as well as multi-window, about time I’d say. This is almost as exciting as the choice between dark and light interface and the fact that external SD cards are finally getting out of the corner is just about good style. I have no idea what made the so big, they just make up additional and unnecessary hassle to their users.

I’m indeed concerned about letting Google backup my data “automatically” with no idea as to where, for how long and who will have access. Even though I can choose freely, I prefer to transfer app data via OTG USB-Stick over cloud backups, not even talking about the fact that they can take forever (chrome backup haha) I shall certainly stick with offline backups which I later can copy to my computer for safety.”



“Android M: While it didn´t get everyone excited, it got me, sure it won’t bring a lot of new things or new UI elements, however android M seems to be marking a point in the history of android, it finally seems that Android is getting mature, it’s finally a system made with every part in mind, the “core” with all Android’s customizability and power, the UI and how the System interacts with the user (and Vice-versa), with focus on features to make android even more powerful, without being less attractive for new users.

App Permissions: FINALLY, We can’t deny android didn’t really handle permissions well before, sure in a security point of view it was working, however the interaction between how it handles them was non-existent. If an app didn’t have a permission for what it was trying to do, it would just crash, the user itself would not even know. As a developer you would need to read the logcats to see what happened, it was a mess for the user, you either had the app and you let it use all the permissions it wanted, or you didn’t install it. Finally this is over, however I still feel sad at the fact it won’t work with apps targeted for earlier versions than M. It’s still hard to find versions targeting L or even KK (most apps are JB, with features via app compat, and I know the difference between min sdk and target sdk).

Doze and per app hibernating: Useless on phones and to be honest it’s not more than just a hacky fix. The system should have better integration with apps, it’s sad but the more apps you have the more sluggish your system will run. This isn’t something we see on iOS and WP. I’m sorry but once you need to make a new service to make a custom keyboard, something just seems wrong.

Cardboard: Good idea, however since I live in Brazil I will not see anything from it, for me this is just an experiment, I believe in the future google will make another VR project based on cardboard. Which would explain why they have not mass marketed it.

Project ARA: Oh this one, this should have been on the keynote alongside android M, the demo was good and bad at same time. I personally have a lot to say, ARA will fail even counting the fact that google will sell it, most of people who will buy are techies and such some months later it’s going be doomed. The project however, will go on. I don’t know but a lot of things from the small to the bigger just felt wrong on that demo, it seems that it is my paranoia but:

  • The pieces were added in such a particular order.
  • He said “applications processor” instead of SOC (and I bet he knows the difference), which lead me to think that there is a co-processor handling all the communication inside the main frame.
  • How it will handle other SOC (remember, the SOC controls everything, not just CPU/GPU, it goes from display to usb port to the bootloaders).
  • Why didn’t they made the display also detachable? they could have made a frame with just the base chips,a big nand (8gb) JUST for the system and some low level bootloaders (like a bios in a pc), on it you could attach what you wanted, and with 8gb just for the system, they could have made a build of android with assets for all densities, etc.
  • How are the antennas being handled? Another chip inside the main display frame? If so how again, will it handle other processors, camera, etc.
  • Why it was running launcher2? AOSP by default builds launcher 3 since KK, launcher 2 we only see in phones upgraded from JB to KK or L.

However the demo at least showed us that ARA was actually a thing and it was somewhat working, even being a “safe” environment.

I/O overall: It seemed more organized than last years however I didn’t like it that much, it didn’t bring much. The only thing for me that helped when developing was the update to the Material design guidelines and the fact that finally, after almost 2 years Android Studio is fully supporting the NDK. I also liked the fact they gone and made more I/O widespread not just in the main cities, e.g. we had I/O extended to my town, and it was pretty cool.”


These were just a few of the responses we received, but you can already see recurring themes and concerns. Coming up to the final release of Android M this fall we will no doubt see more opinions and thoughts arise. To read the concerns of Senior Recognised Developer Pulser regarding backups head over to here. Or to see our summary of the event head over here!

What do you think to the Google I/O and Android M announcements? Leave a comment below!

About author

Mathew Bloomer
Mathew Bloomer

He fell in love with Android after buying a T-mobile G1 in 2008 and hasn't looked back since. He firmly believes the future of technology lies within bio-hacking and is an NFC implantee.