How a yearly release cycle could improve the Android experience

by: Simon HillMay 11, 2015
1.1K
android mwc logo 2015 barcelona 3

Back in 2010, when Andy Rubin was still heading up Android, he told Mercury News, “Our product cycle is now, basically twice a year, and it will probably end up being once a year when things start settling down, because a platform that’s moving — it’s hard for developers to keep up. I want developers to basically leverage the innovation. I don’t want developers to have to predict the innovation.”

If you look at the Android version history, you can hardly fail to notice how erratic the release schedule has been. In the early days of the platform, it was important to rush, because Android was playing catch up, and there was a lot to do. Each new version brought a raft of essential features, but that’s no longer the case.

There are now signs that Google might finally be settling into an annual update schedule, albeit later than expected. Android 5.0 Lollipop landed a year after 4.4 KitKat. Google announced it in June at Google I/O 2014, providing time for a developer preview before the consumer launch in November.

Smaller updates with bug fixes are inevitably going to be released throughout the year, but it looks as though Android M will follow the same pattern.

A lot of benefits

You would assume that having a concrete schedule in mind would be a good thing for the Android team at Google. It’s not a rush to push out new features with the decision on when to ship being made as they go. It should improve the chances of a stable release that’s been properly planned and tested, at least in theory. Nexus owners can attest to the potential impact of bugs in new Android versions.

Predictability and stability are big potential pros for developers and OEMs. If they know when a new version is set to land, then they can plan accordingly. Having to guess isn’t ideal, and it can obviously be pretty frustrating. For OEMs it would provide a nice foundation for their own annual flagship releases. Some manufacturers have stuck to big shows, like MWC, others have chopped and changed every year, but a fixed schedule makes it easier to build hype and expectations.

It should also make it easier for developers and OEMs to plan updates. Previews help developers to ensure that their apps and games work with the latest Android version on day one. Manufacturers can theoretically make the necessary tweaks to their UIs, and push new versions over-the-air (OTA) within a predictable timescale. The current situation is a rush. Often by the time the update actually goes out, Google is announcing a new version of Android.

For consumers, a predictable release schedule for new Android versions would be great. As it stands, the fragmentation situation is very frustrating for device owners keen to get their hands on the latest and greatest features. When Google decides to update Android, there’s a wait to see which manufacturers will push it to their devices, and then another wait for the carriers to make their changes before the update can be pushed out OTA. HTC released an interesting infographic about the update process a while back. You can see how a stable timetable and longer gaps between releases might help. Fragmentation isn’t going to be solved by the annual cycle, but it should certainly make things a bit more transparent.

HTC-Anatomy-of-an-Android-2

Click for full version

A yearly cycle also means one big exciting release with a new feature list, instead of a stream of smaller updates. It brings a bit more clarity to the divide between versions, and it makes it harder for OEMs and carriers to justify not updating. With fewer updates to deal with, perhaps they’ll start to update devices for longer.

And a few drawbacks

The reason that Google hasn’t had a yearly cycle thus far, is to do with the pace of innovation. More frequent releases provide the opportunity to push out new features and cool functionality as it’s developed. With a yearly cycle we’re going to have to wait a bit longer to get the new goodies.

OEMs feed off each other right now, with the competition pushing them to innovate, and a steady stream of new devices hitting the market year-round. A single annual update model could deflate this constant jockeying for position, and slow innovation further.

Security is an issue. If there’s one improvement you don’t want to wait for, it’s the closing of a vulnerability. Bugs are never intended to be part of a release, but unless Google ups its game in terms of defects in new versions of Android, the wait for a fix could grow longer. Even with a yearly cycle, small updates for security and bugs will surely be inevitable.

Is it the right move?

There’s another compelling reason for Google to switch to the annual cycle that we haven’t mentioned yet. The move to deliver new features within Google apps is clearly underway. We don’t always need a platform update to get more out of Android. Regardless of how you feel about this trend and Google’s motivations for it, there’s little doubt that it’s happening.

It’s also worth remembering that Google is starting to push Android for Work and it wants the platform to challenge for the enterprise. Businesses, IT departments, and enterprise app developers want stability, and expect a stable update schedule. For planning purposes it may be seen as a prerequisite that’s currently harming Android’s credibility.

The Android platform is mature. There’s always room to innovate and improve, but we aren’t seeing vital new features in every update. As Google moves towards the refinement process, it seems to make sense to slow down and reduce the disruption of a faster release cycle. It may be difficult to determine whether it will have any impact on the pace of innovation, when it feels as though innovation is already slowing, but the potential benefits for developers and OEMs will hopefully be felt by end users as well.

  • abazigal

    I guess the biggest issue is – how do you quickly and efficiently push out an Android software update to all your customers? As it is, barring stock, updates are still subject to OEM and carrier intervention. Lollipop is still at what…10%? You put all that time and effort into overhauling the look of the OS, and it’s reaching only a minority of your user base. What’s the point of constantly pushing out updates if you know that they never reach the majority of the intended audience?

    I think it is also worth nothing that even iOS is having problems sticking to its annual upgrade cycle. iOS 7 and 8 were huge updates with major redesigns and tons of new features, and which also brought with it its fair share of bugs of problems. It took up till iOS 8.3 to iron out the majority of them, which is a full 6 months after the release of iOS 8, and we are currently looking at iOS 8.4 prior to the announcement of the iOS 9 beta in June. This was tempered only by the fact that Apple could easily push .X software patches to its customers.

    Can you imagine companies like Samsung, HTC, Verizon and ATnT being expected to do the same?

    • Guest123

      I know. IOS 7 was terrible and it was hanging there for 6 months or so with so bad performance issues. Even IOS 8 didn’t really solve it until 8.2 or something. With Android there’s like few level of bugs solving to go thru, Google -> OEM -> Telco and on top of that resources have to be spread out over various devices. Apple only have 1 level . Sometimes i wish 1 Company would just take the Apple path and release only 1 or 2 devices with stock. I’ll probably look at them more seriously.

      • Diamen91

        probably, your best option is Motorola, then..

    • Lazy Sapper

      Buy new phone with new OS and quit whining.

  • OhStopItYou!

    I feel like Google should work on some sort of deal with OEMs to use standardized hardware (ofc if they do like, they can go ahead and use more than the essential stuff). This would ensure that you could flash Google’s stock image if necessary (you’d loose all the extra fluff but you get dem updates)

    OR

    Google should work with Qualcomm on better driver support

  • philosopher_Mk

    Google should stop treating OEMs like developers . Lets say Android M is going to be unveiled this month on io and preview version is gonna be available later in June. OEMs should have preview version right know(while I am typing this ) and working on updates (so they can push preview version on their software later this summer ) if Google wants less fragmentation. But no,OEMs must to wait for aosp code to be released and then start working on updates, that ok for OEMs that don’t use google services,but for others like Samsung,Htc,Lg,Sony its not ok.

    With this practice we will never se share higher than 40-50 % and fragmentation is going to be worse year after year.

    • abazigal

      I presume that Google is going to continue working on debugging Android up to the day it is released for download.

      What happens if, in your scenario, Google issues the developer preview to all OEMs, then discovers a bug and fixes it or finiteness a feature like a few days before it is supposed to be released, and after all developers have already modified the software with their own skins and service?

      Granted, it’s really no better than the current scenario we are seeing now, where Google is already releasing 5.1 before many OEMs have even released their own version of 5.0. If I were the company in question, I would wait until the barrage of software updates stabilise before sitting down to work on the latest software version. Why bother jumping in with the earliest version if I know that it will just be obsoleted by a subsequent software patch? Just imagine if I were working at Samsung. I spend months skinning Lollipop 5.0 with Touchwiz, then Google announces 5.1. Now i have to do the same thing all over again?

      • philosopher_Mk

        Well your comment is what is happening right now with lollipop with one difference , bugs are not fixed. OEMs are waiting for M and Google to fix bugs in lollipop.

        And lollipop on nexus devices has more bugs than OEMs software, samsung even fixed memory leak on note 4 software.

        But lets go back to my scenario. Imagine is samsung release preview version of touchwiz with M later this summer and developers can test their apps on s6,not just on nexus. In the main time Google can provide to Samsung code for the final version of the software so they can work on final version of touchwiz.and push the update a week after nexuses.

        • Diamen91

          What if Google, instead of unveiling the new version of Android in May with all the bugs and everything, than releasing the preview and finally pushing the official first release in november, causing too much hype, people complaining they want the update, etc..and still having to fix lots of bugs that are already pushed to the final user…what if Google just worked on the new version and unveiled it only in november when it’s almost complete and than started pushing it later on when it’s nearly bug-free? wouldn’t it partly solve the problem? Everyone getting the first release of the new version of Android would have a better experience from the day one, and than receive small bug fix pushes on the following months…

          • abazigal

            I think this might be better. As it stands, Apple normally updates iOS once a year to coincide with the annual refresh of the iPhone. They are barely managing this, with some pretty annoying bugs slipping though and key features having to be left out of the initial version and later re-activated via software updates.

            Google is under no such obligation, since their hardware OEMs release hardware all year round. Sure, there’s the nexus line of phones, but that has such a small user base that I wonder if it is worth sacrificing the forest for the trees.

            Or at least, Google may need to rethink its current model of software distribution first.

          • philosopher_Mk

            In that case only nexuses will run latest software. Having a preview is best for fast updates from OEMs.

  • Vanja

    Windows is better than android and android is just a survailence os to monitor humans all around the world.

    • philosopher_Mk

      huh ?

    • Guest

      lolwut

  • Hans Pedersen

    If they do that with the current software model, this is a disaster in the making. Microsoft couldn’t ask for more help getting a foot back in the race.

  • [email protected]

    I agree with abazigal.
    A yearly cycle is a good idea but once the respective carriers intervene and add their crapware that yearly cycle turns into an 18 or 24 month cycle…..

  • Karly Johnston

    I wish they would do that to YT, the constant changes are getting really annoying.

  • Lindle

    Can someone tell me the problem with fragmentation. The way I see it, as long as you’re on 4.1 and above, you can run 99.9999% of apps baring other restrictions such as country and exclusives. Sure there is not ART but most apps can use both runtimes anyway and that problem will go away with time.
    The UI will be a bit different with different versions but skins bring a bigger change anyway

  • Luka Mlinar

    On paper it looks like Android brings a lot with each update yet when I take away the design, I’m having a hard time seeing any difference from JB to L. Put Google Now on them and it’s the same thing. Even ART never brought change so big that I ever went: whoaa, this is amazing. But I guess that statement is subject to opinion and I’m sure people will fight me on it. In my opinion there shouldn’t be more than one update per year. Google should focus more on core functions than UI design that’s subject to change by the OEM anyway. Also opening up a developer ROM 6 months before the release would most likely keep them out of these situations where they need to patch every major update 2 or 3 times over.

  • dfdf

    android is a joke. 40% after year + ? 10% after 6 months + ? only idiot can’t see that.

  • Jerry Li

    REMOVE THE FULL PAGE ADS FOR THE MOBILE SITE. EACH TIME USERS COMPLAIN YOU GUYS TAKE WEEKS TO REMOVE THEM.

    EACH TIME 3 WEEKS LATER YOU GUYS PULL THE SAME SHIT AGAIN. And again it is us users that make your site profitable. It is not wise to piss off the users. We can easily take our page views to all other blogs and websites.

  • Coven1993

    < col Hiiiiiii Friends….uptil I saw the receipt for $6027 , I be certain …that…my friend woz like truly erning money part time from there labtop. . there sisters roommate has done this for only twenty one months and by now repaid the debts on their mini mansion and purchased a top of the range Citroën DS . More Info > SEE MORE INFO<—–

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~n