Linux not the Cloud will Save Schools?

PM Gordon Brown has finally used the word 'cuts' last week with regard to public sector spending. He is determined, in his words, to cut unnecessary spending. M'thinks this must include school ICT unless of course enough money is saved by sacking senior teachers.

We know Gordon's Education Secretary Mr Balls has a long-term Orwellian fantasy whereby ICT is delivered to all schools from a centralised, Cloud-based app-store run by the Cabinet Office.

Such desire for total control probably exceeds even my jaded imagination but with luck he will run out of money and time to realise such a nightmarish vision. Even if you have no qualms about 'Big-State' control you can bet that there would be only a very few chosen vendors allowed to provide state ICT (Capita, Serco, RM, Microsoft would be about right). Down with the Cloud, I say.

So, leaving LaLa land, let's go back into the real world and look at the real savings and freedoms that can come by making use of Free, Open Source Software.

Free, Open Source Software (FOSS) offers schools colleges and the public sector in general enormous savings. These savings are wide ranging. They encompass software purchase and licencing costs, support overheads and energy consumption. If you want to cut spending then you need FOSS.

But before we get carried away we have to realistic as to what can be made FOSS and what cannot.

Too many false dawns and over-ambitious projects have bedevilled FOSS's earlier promise so what follows is a hard-nosed assessment of FOSS's real and important role in education.

We will start with what it cannot do.

What you can't do

Having come successfully to the closing stages of a major open source project in a large school, certain software realities were inescapable.

These realities concern 'untouchable' software. Some software packages have become indispensable to schools. I do not mean this in the sense that they cannot be replaced because of a dastardly labyrinthine plot by their vendors, though these undoubtedly exist, I mean software that cannot be replaced for user-driven reasons.

These are:

SIMS and SAGE

The management information software Capita-Sims.net, has an install base of something like 80% of schools and in those schools it is effectively indispensable. Sims.net runs on Windows 2003 Server and its client software requires a Windows desktop..end of story

Okay, there is a browser option for the desktop, but this is one heck of a lock-in...oh, oh they are not planning to put this in the Cloud app store..are they?

Usually, at this point, some bright spark says, 'Open Source software will never make it in schools until someone writes a decent MIS to replace SIMS'. Not only do they fail to say who this 'someone' is, it ignores something called 'people'.

From an HR perspective the embedded experience and skill-base required to use SIMS successfully means that the FOSS MIS version would have to be exactly the same as SIMS or else the retraining overhead (not to mention the resignations) would scupper the entire idea.

Ditto, the accounting package SAGE. SAGE is the defacto SME accounts package, it cannot be replaced with GNU Cash or whatever simply because no-one will know how to use it (see SIMS for the rest)
Legacy Edu-Ware

By this I mean software say that is used for example by Design and Technology or a Science department that has platform-specific drivers. It will have been used in teaching programs for many years and the skills to use it will be embedded in the teaching staff.

The original authors of the software will have been a small company or even an individual. There are simply not the resources/will to port it to all platforms or open source the code.

For example if a department has 5 years of work based on Corel Draw you will have a devil of a job to get them to use the similar vector drawing package Inkscape. Don't under-estimate what happens if you try to take away some good ICT-based work and proffer your 'better idea'.

This all means that for the foreseeable future the above are going to hang around. Therefore schools will need Windows Servers and Windows desktops.

What you can do

Open Office and Firefox

Open Office, is a blessing to the cause of FOSS in schools and Public Sector in general. It's free and so similar to MS Office 2003 that minimal training is required and certainly less than would be needed to adopt MS Office 2007. You would be barking mad not to consider using it when faced with the cost of MS Office.

Ditto Firefox, cross-platform, secure, free etc.

However the really large gains can be made behind the desktop.

Systems Integration

Systems integration sounds like the kind of phrase trotted out by every consultant you've met. In the context of schools however it has some real meaning.

Traditionally schools were Windows only, or 'Mac Schools', very occasionally and usually temporarily 'Linux Schools. Vendors like RM and Viglen pushed the Windows only total solutions to its limits. Other (arty) schools demanded Macs at least for their art, movies and music.

It is blindingly obvious in an educational context that computers are tools to deliver certain skills. As such, some tools are appropriate for one job and others for different jobs.

An all Mac school is an expensive luxury. An all Windows school is depressingly mediocre but a bit cheaper. An all Linux school is incredibly low cost, especially if you go with thin-client but very limited in the software it can run. None are totally satisfactory.

The obvious answer is a mixed site. One snag, the dominant unifying technology to allow users to share resources on an network is Microsoft's Active Directory (AD). AD unsurprisingly doesn't much like Macs or Linux workstations.

However, by using OpenLDAP and SAMBA on Linux all three platforms can be integrated seamlessly.

A school saves in three ways:

1. by only using expensive computers and software where the task requires (you don't need a G5 to Twitter from the staff room)
2. by deploying the lowest energy consumption and free software desktops where that's all is needed like in the Library or general ICT room
3. minimising the number of legacy computers required to run locked-in software

This is not pie in the sky. As I mentioned earlier, we just finished such a project and the school increased its desktop numbers by 20% but actually reduced its spending on electricity and software by a similar amount.
Virtualisation

We have all heard of this exciting technology. Using virtual servers instead of physical servers dramatically cuts energy consumption and increases reliability by allowing for more fail-over servers.

School ICT technicians are regularly sent on MS Hyper-V seminars but as I write I know of none in UK schools who have dared to deploy it on their own. Even the great RM and Viglen are pretty coy about it. MSCE training simply deprives them of the confidence to unpick their precarious networks.

In contrast, a recent press report shows that virtualisation adoption is globally being lead by Linux Sys-Admins, why would that be?

Try then using free Xen or KVM on Linux, we do, and it works fine for SIMS on Windows 2003 virtuals amongst a stack of Linux virtuals.

And yes, 18 KVM on Linux virtuals are running in above said school, saving about £1200 in electricity per year in the server room alone.. yet now with a fail-over for every service!

Summary

In the real world, as opposed to the Government's cloud-based dream, real and significant savings can be made using Free, Open Source software. The technologies described above work on all platforms, they are unifying technologies.

I would love to see schools free themselves from the closed-shop model of ICT in education whether it be a mini-Microsoft network behemoth or a Stalinesque government app store delivered by the Web.

But then I believe in freedom and civil liberty, you know, those unfashionable things... and of course the tooth fairy.