The network is the computer

Google announced a bunch of stuff last week, finally bringing to the mainstream some bits of tech that have been bubbling away for a few years now.

One is the Chrome operating system, a lightweight OS for netbooks that pretty much hand everything over to the web. So, the OS handles keyboard and mouse inputs and that sort of thing, but basically just boots into a browser and lets you do all your stuff online.

After all, with developments in web technology, who needs software anyway? Google Docs does most of the stuff people who need an office suite use, Picnik is a pretty cool image editor, Gmail is a far better mail client than Outlook is and tools like Huddle and Basecamp provide neat ways of organising your work and collaborating on projects.

Even big, enterprisey software is available through the web now. Salesforce provides a pretty comprehensive CRM offering, Kashflow does the same for accounting, and sites like Netsuite and SAP’s Business by Design provide boring ERP software in the browser.

This is the part of cloud computing known as software-as-a-service. Learning Pool‘s stuff runs on very similar lines: our customers have no software to install, and therefore no patches or upgrades to worry about. Everything can be accessed from anywhere with a browser and a connection to the net.

Anyway, back to Google and Chrome OS. Here’s a video with the skinny:

The idea of the online operating system isn’t new – here’s a review of a previous attempt called YouOS (now sadly dead) that I wrote back in March 2006 – but developments in cloud computing and the almost ubiquitous availability of decent speed broadband (ok, it’s not everywhere yet, especially in rural locations) make it a much more realistic proposition.

What’s interesting about the YouOS example is that it included native applications within the OS itself, rather than just pointing people to existing, external apps. I wrote at the time:

The notion of the online desktop is an interesting one, that conjures the image of computer boxes doing nothing other than handling the keyboard, mouse, display and internet connection; and where you can log in with any machine anywhere in the world and get your own desktop. I suspect, though, that the route that YouOS is taking is the wrong one. What the online OS needs to do is not provide the applications, just the means of accessing the applications, which can be developed by other people on other sites, and the means of storing data to be used and shared between those applications.

It seems like I was probably right about this one (it doesn’t happen often).

Chrome OS won’t be made available for existing netbook owners to download and install – although the fact that it is based on an open source project means that someone else could make it happen. This means that it isn’t possible to have a play with it to see how it works, which is a shame.

One thing that you can have a play with – assuming you have access to Google’s Chrome browser (currently my browser of choice, mainly due to the speed and efficiency of the thing) – is the Chrome Web Store.

A healthy proportion of people are pretty comfortable with the idea of app stores – we’ve used them with our iPhones and iPads, and Android phones and Blackberry users have their own stores – reasonably safe places where applications can be found for the device you are using. Linux users have had an app store like experience for years.

Where these differ with the Chrome store is that Google’s offering is all about web apps, those that work within a browser rather than being native applications that you have to download and install onto your computer, or mobile device.

This is something I struggle with slightly, in terms of understanding what the point is. I mean, when a web app is just an app that runs in a browser, and all you have on your system for accessing apps is a browser, what’s the difference between installing an app and just having a bookmark to it in your browser?!

I guess the answer is around a) making it easy for users to find apps, and providing a space for reviews and that sort of thing; b) enabling a more integrated experience between a web app and the system being used; and c) creating a marketplace where paid-for apps can be, well, paid for.

One neat feature is that by using your Google account, you can sync your Chrome web app setup across machines – so if you log into your account on a different computer, albeit still using Chrome, then your apps come with you, which is cool for portability.

Here’s the video:

The good news about the Chrome store is that folk using the Chrome browser on their usual computer can make use of it. There seems to be a couple of example of Chrome web apps which aren’t available for other browsers – TweetDeck being one.

I’m not quite sure why this is, nor indeed if it is a good thing. There’s the possibility of certain apps only being available to users of certain browsers, which isn’t great.

Still, it’s another step forward for the mainstreaming of cloud computing and software-as-a-service in general.

There’s been quite a bit of talk of a government cloud infastructure as well as an app store for public service use. Indeed, some of these ideas are present in the Knowledge Hub project. The USA government has had an app store for a little while now.

As we pass from the age of the stationary microcomputer and the software industry into a world of commodity computing, understanding the benefits of the approach will be vital – and not just for those working in IT. Indeed, the role of IT departments in organisations will almost certainly need a rethink.

Me ‘talking’ about UKGovCamp and unconferences on Local by Social

I am facilitating a session on (at?) the Local by Social online conference on Monday, 8th November at 13:30 on the topic of the GovCamp movement and unconferences in general.

A couple of levels of sign up are required (it’s hosted on the Communities of Practice) but hopefully we’ll have some good discussion and a few folk will be suitably inspired to run their own events.

Click through to the discussion.

Open source and government

Another post I have been sat on and chewing over for a little while…

Charles Arthur in the Guardian highlighted an interesting area of discussion in the use of open source in government a little while ago. He reports on the views of Liam Maxwell, the councillor responsible for IT policy at the Royal Borough of Windsor & Maidenhead, who’d like to see a move away from proprietary software such as Microsoft Office within local authorities.

Cllr Maxwell would like to see the Cabinet Office mandate the use of the Open Document file format within all levels of government. This would be as opposed to the file formats used by Microsoft’s products, as well as other systems in use in the public sector.

Cllr Maxwell states:

If one council goes to a service provider such as Capita and asks for a change to its Revenues and Benefits system so it works with OpenOffice and ODF instead of Microsoft Office, Capita will tell them to go away. But if government mandates it, then Capita or any of these other companies that do this work for councils could get it done in six months. It’s a dysfunctional market because it’s set by standards which are set at the centre.

A bit of background for the non-dorks out there. The Open Document Format (ODF) is a non-proprietary file standard for use in office productivity suites, which include things like word processors, spreadsheets and slideshow presentations.

The flagship software to use ODF is OpenOffice.org, as alluded to by Cllr Maxwell. OpenOffice.org was developed predominantly by Sun Microsystems as an open source office suite, which then fed into their proprietary offering, StarOffice.

Now, I am a fan of free and open source software and I try to use it wherever I can. But there is so much misunderstanding out there about the benefits – especially around cost – that I do worry about whether people’s minds are filled with free-as-in-beer.

Here are some of the issues with this particular proposal. I do want to make clear that none of these are insurmountable, nor am I in the business of spreading fear, uncertainty and doubt. I’m certainly no apologist for Microsoft, their software or their business practices. I want government to make better use of open source, just that it needs to do so with its eyes open.

The idea of the Cabinet Office mandating use of ODF sounds good, but the naivety to think that this would happen for free is remarkable – there’s no way those big boys would do that much work and not make customers pay for it somewhere down the line.

Then there is training – the idea that the majority of council workers could use OpenOffice as well as they use MS Office right away is rather optimistic. In my experience, folks can’t even cope with upgrades between versions of Word, let alone a whole new system! The costs need to added in: training, writing documentation, loss of productivity while people figure out how to do stuff, or what they can’t do anymore that they used to, etc etc.

Next up with OpenOffice is the Oracle issue – they’ve already made significant changes to OpenSolaris since they bought Sun and there is no guarantee they won’t do the same to OpenOffice. Part of the pro-open source argument is sustainability, but if the sponsoring corporation (which owns the IP and drives development) doesn’t want to know then it would be very hard in practice for the community to get things up and running again.

(Actually, we kind of know what is happening here, as a separate organisation appears to have been formed to managed a fork of OpenOffice.org called LibreOffice. Confusion abounds!)

Next, support. Where is the organisation that can provide support to large organisations when it comes to switching over office suites? It would drown a council ICT department and I can’t think off the top of my head of any company providing this sort of service at scale for it to be outsourced to.

Finally – do we even know if ODF is better than the current alternatives? Where’s the benefit for the switch?

Now what I have written sounds like a massive anti-open source rant, but it isn’t. It’s just highlighting some of the issues. I suspect, for example, that the total cost of ownership of an open source ICT solution – certainly on the desktop – would be roughly the same as the Microsoft (or whoever) one, especially when you take into account select agreements etc.

The arguments in favour of open source need to be on the basis that the software is better, more reliable and stable, quicker and feature rich, and that it works for the government context – adapted for the sector in a cost effective, maintainable and supportable manner.

This brings in a number of issues, around business models for suppliers, procurement, understanding of licensing, copyright and IP, having actual coding knowledge within organisations.

Learning Pool is also a good example of taking open source, contextualising it, then implementing, supporting and maintaining it. We were recently asked to come up with a few bullet points outlining our approach and experiences, which I drafted up as:

  • There are cost savings to be made with open source, but only when the vendor can provide a genuinely comprehensive service that includes implementation and support as well as code. Otherwise the total cost of ownership can spiral.
  • The argument for open source must be based on better, not cheaper, software. We benefit from hundreds of people tracking bugs, developing plugins and testing betas which helps give our product the edge over proprietary rivals.
  • The flexibility of cloud based applications saves significant amounts of time and therefore money in providing upgrades and new features to customers – who don’t have the bother of installing patches etc.
  • Building sharing and collaboration between our customers into the business model has achieved far greater cost savings than either the open source foundation of our software, or the cloud based delivery of it. The fact that we don’t just tolerate, but rather encourage, our customers to share and redistribute resources means government is redesigning fewer wheels every day.

Having said that, we use the LAMP stack which is pretty much a won argument on open source in many ways, it’s other technology, especially on the desktop, where the debate needs to be refined and informed.

Discussions around open source use in government have to be based on pragmatism: is the OSS solution as good as the competition? Is it comaptible with other systems? What are the training overheads? What are the support, maintenance and development arrangements?

The truth is that replacing enterprise IT systems with open source alternatives is a lot more complicated than deciding to build a new website in WordPress. I quickly Googled for ‘open source ERP’ (ERP is Enterprise Resource Planning, those big internal systems made by people like SAP and Oracle, that run HR, finance, CRM and everything else) this afternoon, and the top result was something called Openbravo. I tweeted about it, and none of my contacts – even the open source IT analyst folks – had even heard of it.

It’s probably not surprising that people procuring this stuff run into the arms of the traditional vendors and system integrators.

DavePress podcast 3 – Tim Davies

Tim Davies

Finally, the third DavePress podcast! Here I chat to Tim Davies of Practical Participation about open data and his research into the subject.

[audio:http://traffic.libsyn.com/davepress/DavePress_Ep_3.mp3]

If you can’t, or don’t want to, use the flash player, you can download the .mp3 instead or subscribe with iTunes.

If you have any feedback – or want to volunteer to be a participant in a podcast, please do so in the comments below, or emailpodcast@davepress.net.

For those that want to know, here’s how the podcast is produced.