"CIOs will never get fired for using open source"

by Sharon Lobo    Nov 02, 2012

Harish Pillay Red HatAt every IT forum, open source is sure to a topic of discussion. Though it has gained considerable traction over the last few years, especially in the consumer space, open source still has to see mass adoption in the enterprise segment. At Interop 2012, CXOtoday spoke to Harish Pillay, Global Head - Community Architect & Leadership, Red Hat, to understand more on the adoption of open source in the enterprise space, and what makes it so unique when compared to proprietary systems. Excerpts

What message do you intend to deliver to the CIOs present here?
I am here to talk on how business can maximize their IT investments. For instance, best practices indicate that you need to replace your systems when it reaches end-of-life (EOL). But what does EOL actually mean? Usually it is when a vendor decides not to support an existing version of a product, thereby compelling customers to upgrade to the latest version. But is this the right thing to do? Just because the vendor tells you a product has reached EOL, why are you forced to move. However, businesses do ask such questions, and the answer they always receive is to upgrade.

My contention is that moving to the next version should not be the default answer every time. Businesses should always review what they have currently and see how it can continue to work. I am aware of a particular business who are running systems that had reached EOL twelve years ago. This business runs an OS which is officially not available or supported today, a middleware, whose different version is available today. To top it all this business uses customized apps, whose vendor doesn’t exist anymore! Now this organization is in a situation where even if it upgrades its systems, OS and middleware, it is not sure if its custom-built apps will continue to work. Additionally, the company would have to go through a new UAT process and also ensure they comply with their pre-defined best practices. And all this leads to additional costs. So here the business needs to ask if there is a better way to do things.

My proposal is that such a business should use virtualization, so now they can virtualise their existing environment and put it on a brand new platform, without having to worry about UAT (user acceptance testing) or compliance. Also, one should understand that best practices are good guidelines and not a mandatory thing. Just because particular set of best practices works for one organization doesn’t mean it works for another. So CIO’s should avoid going behind the wall of best practices and instead evaluate their existing IT infrastructure and find the best possible way of doing things.

Most IT decision makers prefer proprietary systems over open source for the sheer skepticism surrounding the latter? This skepticism also stems from the fact that as students or as a non-business users people prefer using pirated versions of Windows than use Linux. Don’t you think if this perception is changed at an early stage, open source would find much more acceptance in enterprises?
This is not a path you should be going down, instead you should go down a path where in you create new avenues of doing things. And this will change the way how you consume technology. For instance, we have OpenShift, a free, auto-scaling Platform as a Service (PaaS), which is a new avenue for creating new things using old stuff. Also, it doesn’t matter what OS you use on your desktop, cause desktops will change in the next 3-5 years.

Today, you don’t even need a laptop to do things any more, theoretically you can do all that stuff even on a hand phone or a tablet. And the next generation of users, probably those who are in school today, they will be much comfortable on using such devices than we do today. When these young users see a CRT monitor they think it so ancient as they are used to LCD panels. This shows how differently they perceive technology. So as hardware changes, you need to deploy software which provides the experience and usability accordingly. Similarly, a CIO needs to take a futuristic approach when deploying solutions both for the businesses internal use as well as for their customers.

I do agree we need engage with students so that we get them to work on open source projects. At Red Hat we have POSSE (Professor’ Open Source Summer Experience), where we engage with faculties across universities to show them how the open source community does what it does. We explain how they collaborate, share and engage with one another with tools such as blogs, repositories, IRC, RSS etc. Here the challenge is that the faculty members have graduated many years ago and these technologies were either not present then or were in different forms. So we want the faculty to use these technologies in their classrooms and this way the students start getting engaged.

As you said, CIOs need to look beyond the present. However, for most it is solving the present day issue, which is more critical. In such scenario, how can one expect the CIOs to focus on the future?
Preparing for the future is as critical as solving today’s problems. The CIO is someone whose mandate is to plan ahead. Solutions used today are something that was perceived years ago. Today’s CIO should strive to make their IT environment ubiquitous, where the work should not halt just because a user device has broken down or stolen. Instead a user should be able to take another device irrespective of the form factor and carry on with the work. As technology evolves CIOs need to be open to exploring newer ideas or else there will be challenge.

Don’t you think open-source has always been a victim of “too many cooks spoil the broth”. For instance, when a particular Linux distribution gains good traction, the community decides to do something radically different and as a result there is an exodus of developers and users leading to another distribution gaining traction and this cycle continues. For instance, Ubuntu was the most preferred desktop Linux OS, till they decided to switch to the Unity user interface and since have steadily lost users. When non Linux users, who have been stuck to single OS see such trends it gives them jitters to try Linux. Also don’t you think a gradual change in software will keep users at ease. Your views.
I have seen users ranting on this (Unity) issue and this is what the whole democratic process open source is all about. This could also happen in the proprietary world too, but the beauty of open-source is that you are not locked to a single OS and so you can switch to what you prefer. The same thing also happened with GNOME 3. When I first started using GNOME 3 my productivity went down. However, gradually my productivity rose to pre-GNOME 3 levels.

I agree there is a resistance to change especially when it is unfamiliar, and proprietary systems don’t want to upset users with a radical change. Today when I look at GNOME 2, I find it old and cluttered. Open source gains acceptance as it matures. I also agree that people are put off when drastic changes are made in the software, but then the question arises how fast is too fast and how slow is too slow. In open source there is no user survey or study done on an user’s expectations or experience, it is just what the community says. In most cases it is the noisy minority that decides changes, whereas the silent majority continues to remain silent. It’s only when the silent majority open their mouths only then will the accepted pace of making changes be understood.

What is the one thing you would say to IT decision makers so they are convinced to adopt open-source?
“You will never get fired for using open source”, that is the one message I would give to CIOs. Whatever CIOs do, they need to ensure it is open standards based. It not only eliminates the possibility of being locked in to a particular vendor. But also if something goes wrong, they have the entire code that they can rework on with a different vendor.