Posted by: Frank | January 24, 2009

Android Market Storage Issue

On Matthew Miller’s blog post about Android Cupcake in which Matt wrote that Google needs to improve storage management (Matt calls it memory managment, but this is really about storage), I replied about how Android Market is consuming more and more storage on my G1. It is currently at 23 MB on my device. I don’t know why, but I had not done a search to see whether anyone else is experiencing the Market storage issue, and I have since learned they have. Apparently this is a known issue to Google that they claim to have fixed in their source code, but obviously the patch has not been released.

In the mean time, the only apparent direct way to fix the problem is a hard reset. However, Android is supposed to automatically decrease this storage when the device gets into a low storage situation. I have seen this several times, each time of which I have manually gone in to the app manager and either cleared the browser cache or deleted applications. According to some, the trick may be that when you get the lower storage warning to turn off the G1 and then turn it back on. It seems as though if Android detected a low storage condition when powered off, it will do some clean up at boot such as with the Market.

So the good news is that Market should not continue to consume more and more storage because at some point Android should clean it up. Hopefully this issue will be fixed in the next release of Android, that by all reports should be coming in the next 2 to 3 weeks. In the mean time I’ll wait for the next time I get a low storage warning and try power cycling my G1 and see what happens.

All of this highlights two things. First, that Matt’s premise in his blog post is absolutely correctly. Android’s storage management must be improved. Ideally, it should incorporate storage cards as part of the total device storage so that users have a way to increase the storage on their device. Minimally Android needs to provide more ways for users to easily free up storage, and it must address the browser cache issue that causes users to lose favorites and cookies whenever they delete the browser data.

In my mind this issue also reinforces to me that Android on the G1 is a beta, and in my mind that means Google ought to focus on bug fixes over adding features. Bugs, including anything that causes a negative user experience, will cause more damage to a platform than not having a built-in video application. What is needed with Google is active release management in which every 6 to 10 weeks a new release is distributed that fixes bugs. I say this because I think one of the threats with Android is fixation on features versus getting the core product working perfectly. In my opinion this is the problem Microsoft has with Windows Mobile that allows bugs like the alarm notification bug to linger through new releases of the product while we get new features. When developing a new platform, the platform vendor (Google) needs to focus on making the operating system perfect and let the developer community focus on providing new features. I think this has been one of the fundamental differences betwee how Apple has developed the Mac operating systems compared with how Microsoft has developed Windows.

Advertisements

Responses

  1. “Bugs, including anything that causes a negative user experience, will cause more damage to a platform than not having a built-in video application.”

    Based on other comments in other venues, many people consider “not having a built-in video application” to be a bug because it “causes a negative user experience”.

    “What is needed with Google is active release management in which every 6 to 10 weeks a new release is distributed that fixes bugs.”

    Unlike, say, a PC operating system, there are more distribution layers between you and a “new release”. The core Android team could issue a “new release” every week — but unless HTC adopts one for the G1 and T-Mobile elects to distribute it, those releases matter not a whit.

    In many respects, the Android open source project is a bit like working on the Linux kernel, and buying a G1 is a bit like buying a TiVo. New Linux kernel releases are made frequently. They get bundled into distros at varying frequencies, depending on the distro. They get pushed out to consumer devices when deemed necessary. Having the kernel developers issue new releases more frequently than they do today will have negligible impact on how frequently your TiVo gets updated.

    I’m not disputing the premise that there needs to be more frequent public release management, but I don’t think it’s going to help your particular concerns very much.

    • You are right that one man’s “feature” is another man’s “bug.” For example, the inability of a user to shut down an application is probably considered a “feature” while many, including myself, may consider it a bug. My point though is that with a scarce amount of resources to do work, there has to be prioritization and I think Google’s priority with Android ought to be on perfection of the core operating system and not so much on features. The power of Android ought to be the developer community, so let the developer community focus on adding features. However, Google has to make support of the developer community a priority. (BTW, I have no reason to believe Google is not doing this.)

      I was under the impression that Google had a more direct path to getting releases of Android to consumers than the lag that exists with Windows Mobile, which also has hardware OEMs and carriers to work with. I would hope that you agree that a consistent release of updates to Android will help the platform grow, though we can reasonably debate the amount of time between releases. I personally think that time cannot be longer than 6 months. My point is that it is important with starting a platform that there be consistent improvements that allow users to tolerate problems because they know that within x period of time there will be a new release likely to fix the problem. Google cannot do what Microsoft does with Windows Mobile and have new releases every year or so, instead I advocate a model like Apple has with OS X where the are regular and frequent updates to the operating system. Eventually as the platform matures the frequency of releases can start decreasing, but at the beginning its important for there to be consistent growth.

  2. “My point though is that with a scarce amount of resources to do work, there has to be prioritization and I think Google’s priority with Android ought to be on perfection of the core operating system and not so much on features.”

    There is no Google with respect to this subject. There is the core Android team. Admittedly, many are Googlers, but not all are. And it is extremely unclear where Google’s power ends and the Open Handset Alliance’s power begins vis a vis the work priorities of the core Android team.

    “The power of Android ought to be the developer community, so let the developer community focus on adding features.”

    If I were a member of the Open Handset Alliance, I would disagree with you vehemently. You’re basically telling the Open Handset Alliance to abandon Android in droves, since they would, by definition, have no control over Android’s feature set.

    Personally, I think there is some middle ground here.

    “I was under the impression that Google had a more direct path to getting releases of Android to consumers than the lag that exists with Windows Mobile, which also has hardware OEMs and carriers to work with.”

    The core Android team has a *less* direct path. Android is licensed under Apache License 2.0 — users of Android (e.g., HTC) can do what they damn well please.

    “I personally think that time cannot be longer than 6 months. My point is that it is important with starting a platform that there be consistent improvements that allow users to tolerate problems because they know that within x period of time there will be a new release likely to fix the problem.”

    No quibbles there.

    “Google cannot do what Microsoft does with Windows Mobile and have new releases every year or so, instead I advocate a model like Apple has with OS X where the are regular and frequent updates to the operating system.”

    I suspect the results will be in the middle.

  3. Agree with your points although I think the next release is gonna be slightly different. I think they need to include the virtual keyboard in there ASAP as many manufacturers are gonna be releasing android phones without keyboards, presumably because Google have promised them that the virtual keyboard will be there for Q1.

    I would suspect that once most of the big features are in, then they can start attacking the bugs bigtime.

  4. […] Storage Finally Passes I have chronicled the issue that I have been having where the Android Market was consuming an alarming amount, 24 MB […]

  5. […] Managing Browser Storage on Android You should be able to tell from the number of posts that I have written on this topic, that storage space management on the T-Mobile G1 is a problem. The problem is caused by the limited amount of usable storage space (256 MB) with the G1 and the fact that you cannot install applications on the SD card or configure the web browser to use the SD card for the browser cache. (Not to mention how Market chews up storage!) […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Categories

%d bloggers like this: