Reply to "Putin says Russia's government to migrate to OSS" on

Original article

Putin says Russia's government to migrate to OSS

Published on:-

< [ But] In the UK, Birmingham City Council aborted a planned
migration to Linux after a revised costing showed it was cheaper to
stay with Microsoft than to migrate. >

My Reply

I was involved in this project and the facts reported in the media were very different from what actually happened. This is my reply to the above article, refuting the claim that Linux is more expensive than windows.

Subject: Re: [SB] Putin says Russia's government to migrate to OSS (fwd)

Dear Sirs,
I read with interest your article referred to in the subject, in
particular the part concerning Birmingham City Council.

I was part of the technical team working on this project at BCC and I
would like to take this opportunity to add some detail which you have not

The remit for the money allocated for the project was to write a report
detailing the experiences of deploying a Linux desktop environment, 
not the actual work involved in building such an environment. Of course a
Linux environment was required to write the report, but with the benefit of
hindsight it was not the primary objective. And the the project money
bought some new hardware.

As a result the money allocated was far more than needed to do a
deployment, the larger share of the funds went to paying a London
consultancy to write a report which didn't draw out the benefits of open
source in a Public Body.
The report was written in MS word. That speaks volumes on it's own.

Looking back, choosing a public library rather than an internal department
made it a hard task to start with, or maybe we were being set up to fail by
vested interests. 

Politics aside, this is what we achieved.  We built a tailored image to
deploy to new hardware from a boot dvd, because this was to be used in a
public area by anonymous users, the public login wiped it's profile after
each login and rebuilt a predefined desktop to ensure privacy of user data.
In addition, we had a openLDAP backend so registered users could login in
and save work on a network drive. We built a windows and Linux (and Mac if
we had any to include) printing environment using CUPS which would have
controlled the spiralling printing costs in the library but without the
high price being asked by a commercial windows only solution.
Not that it is an objective measure but each time we visited the Library,
all the computers were being used, so the public clearly didn't have too
much trouble using Open Office and Firefox on Linux. 
Had we deployed this internally to BCC council employees, we would had
little trouble integrating with BCC corporate systems, we successfully
tested Lotus Notes, the Java SAP client, Citrix and RDP access to remote
desktops, Firefox and Konqueror Browser access via the corporate internet
proxy. Filesystem access to Novell NetWare and windows shared directories
worked fine and printing to Toshiba MFP's and HP laserjet printers worked
without problems. I know this because for the entire time I worked within
the BCC corp environment, I used a Linux Desktop, a period of 6 years. From
time to time, there would be formatting problems with MS office documents,
but this was before the move towards OpenDocument formats. 
The only application I never managed to interact with was our change
management application, Assyst which was dreadful on windows anyway. (Note
to CIO's, if anyone suggests using Axios Assyst for change management, sack
them instantly for incompetence. I'm not joking.) Citrix solved this
problem for me.
As we move away from installed desktop applications to browser delivered
applications, the move towards Open Source desktops becomes ever easier and
the past experiences should not discourage future attempts, the biggest
lesson we could have learnt was to beware of hidden agendas.

> In the UK, Birmingham City Council aborted a planned migration to Linux
> after a revised costing showed it was cheaper to stay with Microsoft than
> to migrate.

Absolutely not true! Take out the cost of the report writing and the
result would have been very different.

The best advice I could offer to Mr. Putin is to make strenuous efforts to
keep all the potential users informed of the reasons for the change, the
benefits and progress.  Microsoft has a very good publicity engine with
deep pockets, and we have seen the depths Microsoft will sink to with no
sense of shame.

The Open Source community has built a better OS than Microsoft, it now has
to build a better PR machine.

Andrew Stringer.

Lessons Learnt and how to do it better next time

There are several lessons to be learnt from our experiences and those of others who have tried this.

  • Do it in stages. If you have time left in windows / windows only application licences, you have already paid for that time, don't waste money by not using it to the full. Use the time it buys you to develop your migration strategy.
  • Don't try to convert all users immediately, the last 10% who really need windows because their applications are not available as Linux versions or it is not practical to substitute alternatives, will take by far the most effort. Don't let this sidetrack you! Target the easy conversions.
  • Do involve the end users. Explain the business case to move away from microsoft and windows. Actively promote training for the Open Source applications running on windows before even thinking about moving the OS.
  • Don't be put off by the number of installed windows machines, don't confuse widespread with popular. Many of the doubters will say windows is popular when they really mean used on many machines which the users have not had any choice over. My experience is that many windows users are only windows users either from ignorance of something better or are given no choice. 'Flu is widespread, but most certainly not popular.
  • What are you planning as a directory service? It's essential to think this through. This is enterprise computing, not hobbyist level tinkering. You could plan on using MAD (active directory) for authentication, we used OpenLDAP but having used OpenLDAP in my current job, I have realised how much more use we could have made of it.

Start by introducing Open Source application on windows desktops, start to make OpenOffice formats the defaults to use, and for the Open Office suite to be the default Office application. Most users are not power users and won't have any problems converting with some training.

Start to change browser habits. Install Firefox and Chrome and make them the default browsers. The internet at large is moving away from the flawed Internet Exploiter, so there is no reason for Local Government users to find this move difficult also.

Start to plan for a changeover from exchange / lookout! to something better. There are better alternatives out there which can work with lookout! as a client but can also work with a browser. Intranet delivered applications don't have the headache of maintenance that installed applications have.

If you are using Lotus Notes, it's far from being OpenSource, but from personal experience the Linux client works well so stick with it. If you are an Administrator/Designer, Linux may not be a option initially but “normal” users are catered for well.

Develop a strategy to run windows application which cannot be converted from a pool of windows VM's. Part of this strategy is to inform vendors of windows only applications that at the next software renewal phase, you will be evaluating competitors who are in tune with your migration, either by producing Linux versions or by moving to a browser based solution.

Your strategy MUST address training. Don't sell this as a cost, but a benefit. If your windows licences for the os or office suite are up for renewal, you will probably be forced to upgrade to the next version. The differences between the installed to target microsoft application / os will be great enough to warrant training anyway, so you are substituting one training package for another. Additionally, most windows/office users use few of the inbuilt features because they don't understand them, if you can turn an ordinary user in to a power user, you will get a more productive workforce as a side effect.

Open Source is great and the Gold standard to aim for, but if your application choice runs on Linux but is closed source, don't do a Stallman and dismiss it out of hand. It's part of a bigger picture, don't loose sight of the eventual goal.

Would I do it again?

Yes!!! It was great fun, and we achieved many objectives overcoming lots of barriers on the way. All in all, a satisfying project to have been involved in.

corr/bcclibrarydesktop.txt · Last modified: 10/10/2013 14:21 (external edit)