Did my Android tablet just become useful?

Ubuntu for Android has just been announced by Canonical, and if it is done right, it has the potential to revolutionise the mobile computing landscape.

If you have read my previous blog posts, you will know that I have no love for the tablet form factor.  Tablets provide a slightly better way to use your phone apps in a form factor that is significantly less convenient than a phone.

I've also dabbled with webtop, which tried to provide a desktop like experience with a phone, but doesn't quite get there.

But Ubuntu for Android could provide the solution that I (and I imagine many others) have been looking for.

The truth is that the argument between Android, iOS, Windows and whatever "we don't provide a way to read emails" (ok, they do now, but you get my point) rubbish OS that Blackberry puts out almost always comes down to apps.

The ability to touch, swipe and pinch is no longer a selling point. Your basic apps like email, calendar, maps and social networks are all covered adequately (Blackberry not withstanding). And it's not like one manufacturer has the monopoly on high end hardware (the hardware that drives Siri is the exception here).

No, for the end users it's all about apps. And you have to search long and hard to find apps that actually let you do useful work on the current selection of tablets.

This is exactly the problem Ubuntu for Android will solve.

While Linux on the desktop never really took off, you can quite easily configure a Linux desktop with all the apps you need to do real work. A quick web search will reveal numerous lists of high quality productivity, multimedia, web and entertainment apps available for Linux. And as an added bonus, the vast majority are also free.

So with Ubuntu for Android, the disconnect between your desktop and your mobile device disappears. You can use your existing Android apps for what they do best, which is entertainment, communication and the consumption of media, and you can use your Linux apps for everything else. You no longer need to sit at a PC or laptop to do "x", while having to pull out your phone to do "y".

However, widespread adoption of Ubuntu for Android will fail if:
  • You need to root your device to install it. Very few people will upgrade from a perfectly good phone or tablet just to get a device from a specific manufacturer that supports Ubuntu.
  • You can only run Ubuntu apps on a second monitor. That might be the ideal way to use these apps, but if the apps themselves are effectively locked until you get to a desk with a monitor, you might as well just have a desktop pc.
  • It is not released quickly, before the popular open source apps get ported to Android. This is happening right now with LibreOffice, and others will follow suit.
  • The most popular apps absolutely suck with a touch interface. We all know that apps that were built for a mouse won't work well with a touch screen, but if it is infuriatingly frustrating to even view a document in LibreOffice, then you might as well just use a laptop.
I'm hopeful that Ubuntu for Android will be available on existing devices and will provide a great Linux desktop experience when combined with a monitor, mouse and keyboard, and at least a usable Linux desktop when your are on the move.

UPDATE

I just found out through PCWorld that Ubuntu for Android will not be available on existing hardware, and does not provide the Linux desktop unless you are plugged into an external monitor.

Sigh.

Comments

Post a Comment

Popular posts from this blog

MinHash for dummies

Duplicate document detection is becoming increasingly important for businesses that have huge collections of email, web pages and documents with multiple copies that may or may not be kept up to date. MinHash is a fairly simple algorithm that from all my Googling has been explained very poorly in blogs or in the kind of mathematical terms that I forgot long ago. So in this article I will attempt to explain how MinHash works at a practical code level. Before I start, please take a look at  http://infolab.stanford.edu/~ullman/mmds/ch3.pdf . That document goes into a lot of theory, and was ultimately where my understanding on MinHash came from. Unfortunately it approaches the algorithm from a theoretical standpoint, but if I gloss over some aspect of the MinHash algorithm here, you will almost certainly find a fuller explanation in the PDF. I'll also be using pseudo Java in these examples instead of traditional math. This means when I use terms like Set, I am referring to the gr...
Read more

Authenticating via Kerberos with Keycloak and Windows 2008 Active Directory

Image
The following instructions show you how to configure Keycloak with Windows AD in order to use Kerberos authentication. Assumptions The Kerberos realm is VIRTUAL.LOCAL The hostname used to access Keycloak is virtual.local. This just means we are running Keycloak on the domain controller. In production virtual.local will be replaced with something like keycloak.dev.virtual.local or something like that, giving you a SPN of  HTTP/keycloak.dev.virtual.local@VIRTUAL.LOCAL Configuration Create a windows domain account called Keycloak. Run the following command to assign a SPN to the user and generate a keytab file: ktpass -out keycloak.keytab -princ HTTP/virtual.local@VIRTUAL.LOCAL -mapUser Keycloak@VIRTUAL.LOCAL -pass password1! -kvno 0 -ptype KRB5_NT_PRINCIPAL -crypto RC4-HMAC-NT Verify the SPN has been assigned to the user with the command: setspn -l Keycloak Configure the LDAP settings in Keycloak like this. Since we are running Keycloak on the domain controller, we...
Read more

Fixing OpenVPN "Authenticate/Decrypt packet error: cipher final failed"

When connecting to a VPN I was constant getting the error Mar  8 09:29:27 openvpn[1696]: Authenticate/Decrypt packet error: cipher final failed I had imported the supplied ovpn file and had followed all the other configuration steps, so this was quite frustrating. Then I saw this in the logs: Mar  8 09:31:07 openvpn[1790]: WARNING: 'cipher' is used inconsistently, local='cipher BF-CBC', remote='cipher AES-256-CBC' Changing my client to use "cipher AES-256-CBC" instead of the default (which apparently was cipher BF-CBC) fixed the issue.
Read more