Wednesday, July 25, 2012

Is Windows 8 launch the time to move desktops to open source?

How long is it since you read a ‘Year of the Linux Desktop’ article? Having been a regular feature since around the year 2000 they seem to have tailed off. The promise of a free and stable alternative to Windows was often cited but never really seemed to take off. So what happened? Is open source on the desktop doomed to be a non-starter, or was it just an idea ahead of its time? The launch of Windows 8 is an ideal opportunity to revisit those questions and consider if that time has finally come.

The reasons why operating system choice is becoming more relevant are most apparent for mobile workers. The emergence of cloud computing, smartphones and tablets have ensured that applications are being written for the web with users starting to see beyond the world of Windows and the traditional desktop PC. That many of these devices are running on a Linux system is in itself an interesting shift.

The browser and access to ‘apps’ is what really matters now, which is why the so-called browser wars are more important to Google and Microsoft than they ever were. Google have taken the distinction between operating system and browser furthest with their release of Chromium OS and Chromebooks. They may be based on a Linux kernel, but as far as the user is concerned the browser is the system. Apps are installed to the browser, printing is controlled from the browser and files are accessed from the browser. Even offline applications are accessed by the browser thanks to HTML5.

Windows 8 is Microsoft’s acknowledgement of these changes, designed as it is to work on non desktop devices and connect to cloud services. This is probably exactly what Microsoft need to do given the circumstances, but their move to a post PC world is also an acknowledgment that they can’t dominate the operating system market in the same way they once did.

This trend has already started. Netbooks are being sold with Android installed, Chromebooks are being heavily pushed by Google and the similarities between iOS and MacOS are clear to see. These are all systems with Unix type distributions at their core. We may not be quite ready to celebrate the year of the Linux desktop, but it’s probably safe to say that we’re already there in terms of its use everywhere else.

Using Linux based systems on a device is one thing, but can it really substitute a managed desktop operating system like Windows? What about authentication, file storage, centralised updates and security? Moreover, what about the legacy windows applications most notably MS Office that we’ve come to know and love?

MacOS has been a great example of how these issues can be overcome. With the right backing it has been shown to be perfectly possible to provide a Unix type desktop system that works for users and can be successfully integrated into the enterprise.

Enterprise requirements like print queues, file storage and centralised management are all areas that Unix type systems shines anyway, but the move to web based provision of these services is making the question even less relevant. For some, having desktops that behave more like devices means much less effort on all these fronts.

For the core office productivity applications, Microsoft and Google are already locked in battle to provide them from the cloud to enable remote working. Microsoft are offering richer functionality for users of Windows and Office, but nobody denies that these applications will become increasingly cloud based.

There has always been a strong cost benefit for Open Source, but the need to run applications on the platform for which they were mostly developed, usually Windows, always won the day. Now that applications are developed for the web rather than Windows the argument for continuing to pay for a Windows license is not so strong. It may be called ‘Android’ or ‘Chromium’ rather than Linux but the key difference will be the cost compared to Windows 8. Governments are keen to keep costs down too and are also looking at open source as a quick win.

In a recent BCS discussion on LinkedIn members were asked their thoughts on using Ubuntu instead of Windows. Ubuntu and Red Hat are often touted as the most feasible enterprise alternative to Windows because of their large backers, but Ubuntu is where much of the attention is now focused. They provide a very easy to deploy system and offer commercial support, cloud storage and centralised deployment options.

It was clear that a number of members have used Ubuntu and variants for their personal and family computers but there was less evidence of its use (other than for servers) in the work environment. A growing relevance to government was cited but at this stage there isn’t too much evidence of it actually happening.

For many the idea of alternatives to Windows on the desktop remains as far away as it ever was. For all the rapid adoption of consumer devices and cloud services, the applications used in most companies remain firmly tied to Windows. MS Office remains a favourite with users and has no Linux edition. Even for the users who are moving to Office 365 Microsoft have cleverly engineered a dependence on windows based software for full functionality. Accounting software, HR packages and custom databases are also genuine reasons why a move to Linux would not make sense.

One thing is certain. If 2013 is the year, it won’t be for the reasons we thought back in 2000. It won’t be because Linux has suddenly become more ‘ready’. It won’t be because Windows has got any worse. It won’t necessarily be about money and it won’t be about security. It will be for the simple reason that the operating system is becoming increasingly irrelevant and unnoticed by users who are spending increasing amounts of their time on platform agnostic web apps.

[A version of this article has been published on the BCS website]

Monday, July 9, 2012

How to make 'going Google' easier.

Over the past 12 months I've been working with a number of companies who have 'gone Google' and implemented Google Apps for their business email, calendar and document collaboration. So, what lessons have I learned so far?

Lesson 1: Don't run a pilot study!


Pilot studies are very much suited to products in active development as they allow your users to influence the next release and have it customised to suit. With all the will in the world this is not going to happen with Google Apps, any more than it would happen with Exchange or MS Office (rarely piloted products).

Like any standard software product, the judgement of its suitability is unlikely to come from a pilot study which will just inform you about what users prefer (and most people prefer to stick with what they know). A better approach is to conduct a full change impact analysis and consider the needs of the business over individual preferences.

Lesson 2: Change the way you work.  


The whole point of Google Apps is that you can work on email, calendar and docs from any location on any device. This is often a bigger challenge for the IT department and senior management than it is  end-users! If you don't like the idea of your users being able to easily collaborate and use their own devices, then stick with Microsoft!

Lesson 3: You can do this in less than 4 weeks. 


Once you've decided to use Google Apps, you can bring about the implementation (including the migration of calendar, contacts and recent emails from Exchange) in less than four weeks.  Agree a realistic amount of emails to migrate. One month is usually enough if users can still access the pst archives from Outlook when required (you can always migrate archives after go-live). Remember to plan for shared / public calendars and start your communication plan as early as possible.

Lesson 4: You can continue to use Outlook (if you must). 


A significant number of users will have an Outlook dependency, for example if their CRM product is MS Office integrated. This is easily worked around using Google Apps Sync for Microsoft Outlook. It's still important to train your users how to use gmail though, particularly if they would like to work on other devices from other locations.

Lesson 5: Stop panicking, and just do it. 


If you've planned the migration and kept users informed, the transition will be much less of a nightmare than you ever thought it would be. If you are brave and decide to move the whole company over on the same day, you will probably make life easier for yourself. Users will adapt much quicker than you think and it's always better if you don't have to make your legacy and new system co-exist.  Finally, and of course I would say this, but if you can afford to get a specialist with experience, it will always help!

Ray Allen is a Google Apps Certified Deployment Specialist at Third Way IT.

Tuesday, July 3, 2012

Moving to Ubuntu from Windows 7

I've decided it's time for me to use an alternative to Windows 7 on my main computer. As my dependency on running client side software diminishes (now working mostly from the browser) so too does my need to run Windows as an operating system and its ever increasing startup time.

My motivation isn't just for the sake of reducing startup times. My day job involves helping businesses move away from Microsoft (replacing Exchange, Sharepoint and MS Office with Google Apps), so the obvious next step is to look at a more cost-effective and simple to manage desktop operating system.

Chromebooks are a neat idea, but in practice I don't think I'm ready to replace my main computer with a browser-only device like this. What I'm looking for is a good compromise between a zero-maintenance device like a Chromebook and a high maintenance installation of Windows 7 (a third way if you like).

I've decided to go with Ubuntu which I know to be stable and open source, but flexible enough to run the popular client applications if I need to (Skype, Firefox, LibreOffice and the like). It also has a new user interface called Unity which I'm keen to find out about. 

Installation was straightforward. I remember installing Linux in the early days and it was no fun at all. You had to understand disk partitions and boot managers, but this is now all handled for you, and you can still boot into Windows if you need to. There is even a windows installer option available if you want to install it from within Windows rather than using a boot disk. You can't get easier than that.

Once installed my immediate impression was how similar it was to a MacOS interface. The app launcher is very similar to the mac 'dock', the only difference being it is on the left of the screen rather than the bottom. This makes perfect sense given most screen dimensions, particularly netbooks. You can easily configure this to hide using the settings icon. 

If like me you start looking for the 'menu' so you can add more apps, the trick is to right click the Ubuntu 'Dash' icon at the top of the launcher. This provides you with search and find options for applications, files, pictures and videos. I've now got Chrome, Firefox, LibreOffice, Skype, GIMP and gThumb viewer which covers all my needs. It connected to my USB printer without issue.

The boot up time for me is now 33 seconds, about half what it was taking my relatively clean installation of Windows 7 (Lenovo Ideapad S205). In my experience with Linux this remains consistent, unlike Windows which seems to get progressively slower with each update.

My only disappointment was Skype. I had trouble configuring this on Windows with my internal mic, and it looks like I've got the same problem on Ubuntu. Otherwise, everything appears to work out of the box.

The real insight for me is how straightforward this would be to manage in an enterprise environment. The control panel is limited to what a user would like to do and everything else requires administrative rights. Upgrades are simple, and being open source there will be no cost associated with doing so. Combining open source systems and cloud products is definitely a good approach for companies looking to reduce costs but keep up to date. 

Ubuntu with LibreOffice