CleverBlogName

I’ve had this phone for a few months now, an upgrade from my S-E X10 Mini Pro; it’s a phenominal improvement on the previous version... so much so I’m a little lost as to where to begin with it. Perhaps I should start with it’s magical and wonderful design...

I’ve been very happy with a number of aspects of my Sony-Ericsson X10 mini pro. In many regards it’s a fantastic piece of kit; the hardware is almost perfect (if it had a front-facing camera and a 3" screen with a slightly higher resolution it would be absolutely perfect). But the software...

The X10 mini pro was released last year with Android 1.6. Some time after Android 2.2 was released Sony eventually rolled out Android 2.1 for the X10 series of phones. Then they announced, less than a year after first releasing the X10 range that there would be no more upgrades, ever.

After a few months of being flamed everywhere, S-E revised this; the X10 would get a neutered upgrade that would ship without any real support for the sexy camera that was supposed to be a main selling point. The X10 minis and the X8, released only a few months ago, would be forever stranded at 2.1.

This made me rather unhappy, to put it mildly. Android 2.2 is a huge step up from 2.1—it contains the JIT-enabled Dalvik JVM, massively improving performance by 3 - 5 times. It also makes calendar sync viable for third-party tools like Funambol. These are big features. I’m not especially interested in doing without them. So this was pretty much my breaking point for investigating third-party upgrades again; last time I’d looked, late last year, they’d been pretty horribly, with such minor features as WiFi not working. I was delighted to discover things have moved along quite considerably; as well as a third-party Android 2.2 upgrade, there’s a 2.3 one as well. I decided to play it a little conservative and aim at the 2.2 upgrade, since it’s a bit more mature.

Now, while I wrangle technology for a living, I was a bit nervous about going down this route. I have a bit of a case of “machanic’s car” with home technology, and I didn’t want to irrevocably brick the phone. In the end, though, the Xcyanogen and XDA community have done a fantastic job of making the process not so much “idiot-proof” as “knows-enough-to-be-dangerous-proof”, a far more significant accomplishment. A look around the Xperia mini Cyanogen blog provided most of the information and pointers I needed, and the combination of the excellent Flashtool and the Xcyanogenpro Froyo image made it simplicity itself: plug phone in, use Flashtool to add Xrecovery, copy the Cyanogen image over, factory reset the phone, and then load the Cyanogen image.

Obviously you’ll want to have backed up everything you care about, since it will be gone after this.

(Standard disclaimer: hey, it worked for me. No crying if it blows your phone away.)

The Xcyanogen standard distribution is a huge, huge step up from the vanilla S-E Android build in most respects; one of the claims for not being able to ship newer versions of Android more quickly (or at all) is that S-E have had to customise things for a better experience, especially with the small screen layout of the Mini and Mini Pro. After running Xcyanogen for a few days, I can confidently say, “bullshit.” Or, more accurately, that a bunch of unpaid hackers have assembled something that is, for the most part, streets ahead, through a combination of their own code and utilising what’s available on the market.

Take the screen layout: the “innovation” S-E provided was to allow one widget per horizontal workspace, and four corner buttons you could use to launch things from. The latter were fixed per workspace, and you then had a popup menu for all your other icons. The Xcyanogen team provide you with the ADW.launcher and the free edition of Launcher Pro; I’ve been using Launcher Pro, which has much better-sized icons. I have now got a hot bar with three groups of 5 icons, and the ability to arrange a 4 x 4 grid of small, attractive icons in each of a user-defined number of workspaces. I have infintely more flexibility, and a far more intuitive, responsive interface to work with.

The S-E camera is replaced the by MIUI camera (although a bug in the current edition of the firmware means the Android camera comes up when I click the shutter button, which is annoying), and an MIUI player displaces the terrible S-E music player easily. The standard Google apps are all there, as is Market acces. The Gallery and Gallery 3D apps are about a billion times better than the thing that ships with the S-E Android built. Rotating screens work better—the standard firmware only supports two the four possible orientations. The lock screen actually has the media buttons on it when you’re playing music.

And everything is much, much, much faster.

I don’t have the vast volume of crapware S-E ship as standard, either.

There are a few things missing: I don’t have a built-in QR code reader. Nor is there a standard video player, and the S-E SMS messaging add-on is actually a bit better than the standard Android one.

On so many levels, big and small, the Xcyanogen mod is streets ahead of the stock firmware, though, and to a degree that should humilate whoever’s in charge of S-E’s engineering efforts. Perhaps they should stick to porting the core kernel to their hardware, making sexy kit, and pay the Xcyanogen team, and some of the other projects the latter have leveraged, because they sure as hell do a better job.

So I became acquainted with the “Blue Ring of Death” on my old (and Maire’s current) K850i; the phone doesn’t power on, with the screen staying black, the blue ring on the camera coming hard on, and the keyboard lighting up.

Not a thrilling sight.

A little Googling took my to an initially rather unpromising discussion which explained that:

• The K850 is notoriously unreliable and BRODs under all manner of circumstances and firmware revisions.
• The Sony-Ericsson Updater tool can’t fix the problem.
• Sony themselves can’t, either, and will typically take your phone and fail to fix it.
• There are a number of payware unlocking/reflashing tools that can.

The last was rather irritating. Not because I begrudge the people who developed Davinci or the other such tools, but because this is a perfect example of an anti-feature: like most mobile phones, the K850 is so locked down that when it blackscreens and becomes unusable, not only can I not fix it, neither can Sony, and I need to spend extra money to buy a third-party tool to rescue it.

Fortunately a little more reading about the place provided my salvation: the A2 Uploader, which works with a variety of Sony-Ericsson phones, including the K850i; and TopSony for the images needed to make it work. Download the latest firmware, reflash the filesystem, and then upload the customisation files and lo! The phone works (sans all the on-phone data, but you do make backups, right?).

So, phone restored for free. Hell, A2 Uploader will letme rip the obnoxious Vodaphone branding (useless menu and button items) out of my C903). Win, right?

Well, in one sense, yes. In another, though, not really. This whole thing is a parade of antifeatures:

• SE (and other phone manufacturers) don’t want anyone developing other tools to work with their phones, and go out of their way to make it difficult for third parties to help me, the phone owner, customise the phone I own.
• SE don’t want to allow me to up or downgrade to avoid buggy firmware that makes the phone crap itself. I can only get firmware and filesystem images from third parties of questionable legality.
• I can’t customise my phone to get rid of crapware (Facebooks apps and other rubbish).

Pleasantly surprised to note that the Sony-Ericsson PC suite will out-of-the-box sync with any SyncML capable host—Funambol or eGroupware—thereby syncing your phone to the rest of your world. While I’ve got a (tiny) data plan and I’m generally happy to OTA sync, New Zealand belongs to that sub-Third World chunk of the mobile data universe where data is excruciatingly expensive and syncing by cable can be an attractive option.

Anyway, if this is relevant to your interests, it’s pretty straightforward:

Having had a good time getting Funambol working for OTA syncing is great, but it’s one part of a solution; it’s great for backup and multi-device information sharing, but it doesn’t give much in the way of, say, cross-user calender sharing.

Enter groupware solutions. (Look ma, I’m talking like an architect!)

Having poke about a bit, I’m trying eGroupware. It’s an free solution (albeit with a more featureful commercial variant); it supports PostgreSQL (a must-have for me), and it doesn’t require any kind of full stack replacement (other suites of this sort, as I have lamented in the past seem to require you replace your existing mail server, web server, and random other components). This all looks reasonably positive, as does the fact that it’s regularly producing releases and appears to be taking patches from an active community.

Funambol, formerly Sync4J, is a number of things, but the bit I care about is that it’s a FOSS SyncML server you can download, run on your own server, and appears to be widely supported by all sorts of other FOSS tools (groupware, PIM software, and so on), and offer support for a huge variety of free and non-free SyncML clients. Since I can has data plan the idea of over-the-air syncing to a SyncML host has gained greater appeal, on top of the pre-existing appeal of making it easier to share calendars and the like with Maire.

The out-of-the-box delivery for Funambol is a binary Linux package that contains a JDK, Tomcat6, an RDBMs, and Funambol itself. On the one hand, this makes getting it up and running in a one-clickish fashion kind of easy. On the other hand, if you already have a database server and app server configured (for example) this is a bit of a wasteful duplication of resources; Funambol will let you work with (some) alternatives.