Confronting their rapidly shrinking budgets, public sector bean counters must imagine that someone somewhere has been casting Chinese curses about living in interesting times. Because when money gets tight, things sure do get interesting. You would think that at times like these open-source deployments would be the obvious …
minor correction - do not publish
"it ought to be a shoe-in."
A shoo-in is what happens when there's an obvious winner. A shoeing is what you get when you make obnoxious pedantic corrections in a rough Edinburgh pub.
is what you get when you let BBC Breakfast presenters loose on a language they have no hope of mastering. Simple concepts like 'singular' or 'plural' seem to consistently fox them. Even by chance they should get it it right 50% of the time ffs! But I digress...
Everyone seems to forget: FOSS still needs to be supported and support is a high cost item.
Sure, I can run Fedora and Mythbuntu at home, for free, but I'm not in need of getting someone to sort out serious problems, these systems aren't exactly mission critical. If I were running servers or desktops in a company and something took them down, I would need a vendor to come in and sort it out, rather than relying on web searches.
That must be ...
That must be why so few use Apache.
And why Microsoft support has such a good reputation.
If you run an army of desktops in a large organisation, all of them with the same installation doing largely the same thing, what exactly are your support requirements? If one of them starts acting up or won't boot, you reinstall it or fix the hardware problem causing it. When was the last time you called Microsoft support for a problem you were having with a normal desktop PC?
Don't forget the distinction between the back office side and the front line desktop systems. They have very different issues.
And if you want support, there are organisations out there that will provide it. RedHat, for one, are proving that there is a lot of money in that business.
FOSS needs support
About 3 years ago the company where I work part-time wanted a new e-mail system but didn't have any funds for it. An old PC, SuSE Linux and the Courier mail package later they had such a system - which had over 2 years continuous up time until we had a power outage that exceeded the capacity of the UPS. It is also compatible with and integrated with MS Active Directory for user authentication. Support has consisted entirely of adding things that weren't in the original request - such as automatic "sig" addition, archived mail recovery and automatic enrolling of new staff in role-specific group aliases.
Smaller, £100m-sized chunks
Is it just me or is a £100m-sized chunk still rather hefty?
In my experience, the biggest potential gains from Open Source are in the local government sector. At the moment there are several hundred local authorities, all given the same instructions from central gov, then each going out to different suppliers to procure software. It's a bonanza for the companies involved, but poor value for government.
"Open Source" isn't the answer
The real answer only appears near the bottom of the article - Open Standards. That's why MS spent so much buying votes for their so called open standard for Office documents.
With open standard, you can have more than one vendors working with the same data - you can choose based on criteria other than "does it work with my data". Once you mandate truly open data standards* then things will change. Buyers really will be able to choose options based on things like vendor reputation & stability, cost, value, etc, etc. That is, of course, why MS hate open standard so much and will go to great lengths to subvert them.
* To be called open, a standard should be available in full to anyone without any restrictions on it's use (no NDAs or restrictions on further dissemination), and it should be complete and implementable in a way that MS's OOXML is not).
As for Open Source - it has a place. Free Software (as in speech, not beer) is arguably more important. Software can be open source without being free, and the free bit is what allows customisation and independent development.
Absolutiely agree about open standards.
Even using the word consultation elevates the status of this survey.
As the official guidance on consultations says: "Clearly, if there is no scope for consultees to influence the policy, a formal consultation exercise should not be launched"
Note also that the survey resiles on the definition of open standards in the Cabinet Office procurement notice by calling the definition a "draft". Note also the use of the phrase "common standards" whatever they are. We deconstruct it here http://bit.ly/ifd7Yl
Take more than a shoe horn
I agree with open standards being the next step forward.
The problem with local government at the sharp end is the shear amount of purpose built solutions such as parking fine software, election systems, bus pass registrations, planning and mapping systems, revenues, benefit and non domestic rate systems and many, many more. The software houses are fairly well all geared to operating in Windows.
Heck, some of them program so badly that they require specific versions of Microsoft Office to be on the servers, so that the libraries that they have used are present; and these are not always compatible with other software so each system ends up requiring its own discrete server.
The other problem, of course, is licensing. When Office comes with the workstation along with the ability to use any version of office back to previous versions, then there is no financial benefit to replacing it with open source software.
Of course, all the integration from the other systems are all hooked in to office also, so if we did replace office with libre, we would lose considerable integration functionality.
So the first stepping block really is open standards. In order to get significant open source in to local government, it will take more than a shoe horn. More like a bazooka.
Might be useful if they dragged out and dusted off the Office of Government Commerce Document 'Open Source Software - Guidance on Implementing UK Government Policy'.
To help out a bit the bean-counters can find it at http://www.ogc.gov.uk/assets/images/OGCOpenSourceSoftwarePolicy.pdf
Amongst the weaknesses is 'perceived lack of support (but they do go on to say it is only perceived).
BTW, it's dated September 2002