I have just read the first 3 parts of a series on switching Corporate desktops to Linux.
However, I don’t agree with the approach or the solution. Now granted I have not been the CIO of any large organisation. But I do have extensive experience in organisations of up to 100 staff over 20 years.
I don’t think this is a good solution politically, it certainly would not have worked anywhere I have been employed (or owned). The risks of a complete switch over are too great and won’t be accepted without first proving that a gentle migration works and that there are successful pilots/prototypes already running.
I don’t think it is a good solution technically either. Actually I probably mean that I don’t think it is likely to be an optimal technical solution. Clearly after nearly 6 years of using Linux on the desktop (and servers) myself I believe that Linux on the desktop is a good thing. But you still have all the problems of desktops to manage and that without the skills your users have already built up in knowing how to keep windows running, how to install applications etc.
Some of the comments have suggested re-using the existing hardware with Linux instead of Windows as a way of saving money. Hardly likely to be a way of endearing everyone to you: "Hey guys meet Dave the one who wants you to keep your old computers for longer". It just means more will be out to prove Linux does not work so that they can have a new computer.
For me the big opportunity is to move away from Windows on servers and desktops but to do it by distributing thin clients to the users. So far as I can see the best solution for this at the moment is Sun Rays. Unfortunately, other than a few key bloggers at Sun like MaryMaryQuiteContrary (aka the Sun Ray Girl), also John at The Clingan Zone, there is also the Thin Guy Life is Just a Tire Swing: Sun Ray and Linux (not actually many Sun Ray posts). Instead see my earlier 42: SunRay Links.
In the US for under $14K you can get a 1U two operton CPU server with 20 thin clients (monitors extra). The benefits of this should be noticed immediately (speed of installation/deployment, flexibility in use) and long term (support costs, upgrades etc). Just like a standard Linux desktop running windows applications should be more straightforward using Citrix (or alternatives) but with simpler management and support due to being server based. Same package in the UK is just under £10K and you could get 20 good 17" LCD monitors for under £4K.
So for a complete set of standard applications (OpenOffice, Firefox, Thunderbird etc) you are looking at a total cost of under £14K for 20 desktops ie £700 each. But the real win will be in the long run as these are going to be so much cheaper to support, upgrade and manage.
It seems a good choice to me, the support costs are going to show a significant saving, the security is going to be better, the flexibility is better (provide Sun Rays at home and people can carry on exactly where they left off at work; go into a meeting, use your smartcard and show the presentation you left ready on the first slide; wander over to a co-worker, swap in your smartcard and get immediate help with your work).
So why is it promoted so little by Sun? Sure we see occasional mentions by Jonathan Schwartz (and he clearly uses one for his own presentations at the major Sun events) but the overall visibility is tiny and some of the pricing is silly as I mentioned yesterday.
For example Sun Ray 1g £250 and Sun Ray 170 £750. Now I think £500 is rather a lot to pay for a 17" LCD monitor (see a Difusion P191 19 TFT Display Silver LCD at £170 or a nice Samsung or Iiyama both under £200). When deploying to 20 desks with the bundle then you are looking at about a £6K difference in cost!
There are other use cases I have in mind beyond the traditional corporate where I think Sun have massively under estimated their own product. I wonder when they will really grok their own products in the wider market place.