A recent roundtable organised by The Banker looked at issues surrounding the non-proprietary Linux operating system. A group of senior bankers took part in a group session discussing Linux adoption and prospects, chaired by The Banker’s technology columnist Chris Skinner.
Q Is Linux a technology that is the next bandwagon and why should it be adopted? Clive Whincup, chief information officer at Banca Popolare di Milano: Linux for me is part of an overall issue, which is to do with open standards. The primary driver for me is choice: to have the possibility of choosing suppliers and to know that the decisions that are made today are not going to lock me into inevitable decisions tomorrow. The initial debates we had going back four years were about open standards as opposed to proprietary solutions. As a consequence, we arrived at Linux. Hence, the prime business issue for me is choice – both choice today and choice tomorrow. Suresh Viswanathan, senior vice-president IT, at Citigroup: The solution we are using to replace our legacy systems is Flexcube. We started off initially by running this on Hewlett Packard (HP) hardware. Over a period of time, as the application has matured and Linux has taken off, we are just about at the stage where we are evaluating what is going to happen with our programme, and what happens in the next 18-24 months. Whether, from a total cost of ownership perspective and a maintenance facility and connectivity perspective, Linux is the area for us to explore – it is at a very early exploratory stage of our experience with Linux. Q Is the total cost of ownership (TCO) of Linux more expensive, when you wrap it all together over five years, than Windows? P J Di Giammarino, chief operating officer at Barclays Capital:I will choose my words carefully here. Depending on your relationship with Microsoft, that equation can be swung to an extent. Clive Whincup: The issue of TCO is extremely specific. Our experience has been that in some cases, there is a powerful argument for Linux, in other cases there is not; it depends on circumstances. This is why I find a lot of confusion with the benchmarks that are quoted and the comparisons that are made, some of which are obviously quite provocative. The key to effective decision-making is to understand your own context, understand your own business and understand where that is going – and even the areas of business – and then to make the decision on the figures that you have. At the end of the day, when we looked at our specific circumstances and we looked at the overall costs of Microsoft and Linux, we were convinced that Linux was an advantage from a TCO perspective. Many of the issues concern the desktop. Some of the issues are about the server. However, the overall issue is how compliant Microsoft is with a fairly rigorous open standard architecture – because we found many instances where we were being forced to compromise on that merely by choosing the Windows desktop. That has a whole series of cost implications that are not normally factored into TCO calculations. Bob Fuller, director of IT at Dresdner Kleinwort Wasserstein: We are not only early adopters of Linux, we are also early adopters of open source. I think that one of the problems that Microsoft had in both the server market and the desktop market is that it has been in a situation where it has tried to provide huge amounts of functionality. Microsoft now wishes to go down to the server end of the market and it has some very good technologies. However, fundamentally, it did not develop its operating systems with security bottom-up, as that was not its main gravitational pull for marketing. Reverse engineering that security aspect is now challenging. You only have to see how many downloads you have to do for Windows XP and any of the other Microsoft operating systems to find that is the case. Moving from Windows 95 to Windows 98 to Windows NT and to Windows XP is never a painless experience. Hence, the number of packages and the difference in the operating systems – not necessarily from the user perspective, but how it works underneath the surface – has been a major issue. Linux probably does not suffer from that. Linux shows no characteristic currently, to me, that indicates that it will become a different operating system or that there will be a Linux II. Linux will just migrate into what it is going to be. That has its own issues, but they are not quite as bad as just taking a leap from Windows XP to whatever comes next. One of the reasons for Linux’s slow adoption is that it is not in the hardware suppliers’ best interests for Linux to be all-encompassing, as you can use your old kit for longer. You definitely do not want that. The other interesting thing that has happened to Linux, which has given it a badge of responsibility that it did not have before, is that IBM has decided that it is its way to compete with Sun. The only reason it did it was that it wanted to take Sun on head-to-head. I believe that there are a lot of things going for Linux. It is one of the few operating systems still around. Ten years ago, you needed all fingers and toes to count all the operating systems; now there are just over a handful and there will be even fewer within the next 10 years. Therefore, it has to be always on the table. It has to be there as a choice. At the desktop level, I think the big problem that Linux has is to make itself more and more like Windows without running into legal problems. At the moment, you would have to say that the desktop is owned by Windows, for a whole host of reasons. Q What role will Linux play in the future, especially in relation to Microsoft? Clive Whincup: At the end of the day, for us, Linux has now ceased to be a primary issue of debate, given that the decision was taken on the service side over two years ago and on the appliance side earlier this year. Linux itself has become a non-issue. It has gone into that space where it is almost an appliance; it is almost like MS-DOS underneath Windows. Who cares? It is just something that is there. The real issue is about what is on the top. Therefore, when you are talking about the server side, it is a management issue and it is an IT issue. David Hanley, director, financial solutions, Oracle EMEA: From Oracle’s viewpoint, we have Linux to enable us, within our own organisation, to take advantage of low-cost computing power. Our founder, Lawrence Ellison, would say that we bet our business on Linux. I think that, from a practical aspect, we have. The reason that we have done it is because we see good price performance. Allied to that, we have been able to cluster those low-cost devices on Linux. We support a breadth of operating systems. Q What will be your main infrastructure in 2009? Suresh Viswanathan: On the server side, my view is that Linux – or an evolved Linux by that stage – stands a very good chance of being our chosen option. On the desktop side, there are no bets at this stage. It is a long time before I can really talk about eliminating Microsoft. My guess is that it will be a Microsoft front engine at that stage. P J Di Giammarino: In 2009, will the desktop have opened up and will it be less proprietary and will we be able to get past layering issues? I think the answer to that is: absolutely. In terms of how exactly we will have got there, I do not know. The market forces will begin to take us that way on the desktop. V Senthil Kumar, chief executive officer at I-flex Solutions: I think Linux is the only one that qualifies. On the desktop, I think there will be alternative devices available for access. The desktop itself will change into something else. On the application side there will be more open standards, as we have already seen in the last five years. The classic example is of Oracle running on any platform provider. Chris Skinner: My perspective in terms of five years ahead is the same as I was saying 15 years ago, which is that IT is like electricity – I do not care how it is generated. The point is that you are able to switch it on and use it. As long as it works, that is all I care about. Bankers probably have the same attitude. It does not matter if it is Linux, Sun, Microsoft or IBM. Nobody cares as long as it works and does the job and, moreover, that it does the job more competitively and better than everyone else’s systems. The roundtable was organised in association with i-flex and Oracle
