Open Source – World Biggest Software Company?

Over the last couple of years open source software has gradually shifted from developer’s desktop to corporate infrastructure. Born out of university labs, this is quite a commendable journey. While this trend is captured in quite a few studies around that capture the mood and sentiment of CIOs, Architects; this particular post from silicon.com caught our attention sometime back. Sure, most of us were convinced about the economic impact of open source besides the freedom part, but we perhaps weren’t thinking about the magnitude of that impact. As the post and the quoted study suggest the economic impact – in this case lost sales opportunity of proprietary software to open source is estimated to be $ 1 trillion per year. How does that compare with the revenues of some of the largest software companies in the world, say Microsoft at $ 60 billion? There you go.

Palle Pedersen’s blog Inside Open Source too provides some interesting insights. With an estimated 1.7 billion lines of code developed each year in the open source world, the estimated cost of development works out to $ 30 billion per year. Compare that to Microsoft’s $ 11.5 billion as cost of revenue or development costs. There you go again.

So, if the scope is so devastating, then what are the industry watchers thinking? Here is Gartner Research’s take on open source . By 2012 , 80 per cent of all commercial software will include elements of open-source technology. By Apr 2008 Gartner revised their earlier estimate and suggested the figure to be 90%. Again later in the year 2008, Gartner published that over 85% of the companies are already using open source and the remaining 15% plan to do so in the next 12 months.

What does this phenomenal growth suggest? Lower Total cost of ownership for the enterprises, opportunity for many developers and overall ability to choose and perhaps modify based on requirements? The benefits of open source as a list is going to be fairly long, however one of the prominent reasons sighted by the opposition of open source is that open source does not amount to free but instead sums up to higher costs if one were to take maintenance and support into account. That may be a lame excuse if it comes to consumer software such as firefox, openoffice etc, but may hold some ground if one were to consider enterprise deployment of applications and infrastructure. These are far more complex and enterprises value business continuity along with saved costs and therefore typically look for maintenance and support. Besides a big reason why people look for open source is the ‘open source’ part where one can adapt or extend code to suit business needs.

So, the key question is how does one compare the cost of maintenance or ownership in a broader sense? Better still how does open source as a community work towards bringing down such costs for its own benefit and growth? Be it maintenance and support or development extensions or integration and customization pieces, how does one promote reuse here as well, keeping the spirit of open source while driving greater efficiency? Is that really possible or is some of that already happening?

Looking at some of the emerging trends in IT consulting and services sector, one of the key players is India with over $ 50 billion of IT exports in 2008. This industry has been mostly supporting and supplementing the enterprise product deployments, developing extensions, training, support etc – considered a complementary eco-system in many cases to proprietary enterprise software. For example companies have SAP, Oracle, Microsoft practice etc. So, then how is this industry poised in terms of open source adoption? One would expect it to be a spontaneous transition given the open nature, but a look at the participation and contribution numbers kind of tells a different story. With over $ 50 billion in IT exports, India is definitely a key player. Compare that to a ranking of 7th in terms of open source contribution from India amongst Asian nations in 2006. Ranking at a world level would obviously be much lower. So, clearly something is missing and we (now the enthusiasts behind Kenfuse) started pondering sometime back if anything can be done to kick start higher contribution levels.

Some high level understanding coupled with our own experience working in this part of the world as well as exchanging thoughts with some of the industry leaders and open source project leads, few factors emerged as likely influencers.

1. Lack of awareness: The education system in most technical colleges in India are still focused around theoretical concepts and fundamentals with little emphasis on usage etc. Quite naturally this limits the exposure of students to a vast and growing open source space.
2. Lack of role models: Unlike in some of the US , European universities where open source participation is common place, Indian campuses lack that. This leads to little or no peer pressure effect – often considered one of the key motivators to open source contribution
3. Lack of mentorship: Technical training institutes in India are big. They range from Java courses to embedded system and more. Mostly favored by professionals without a computer science background, this segment is also quite large and growing. However, open source does not have such training institutes where one can get mentorship – barring of course Linux training which again just focus on usage and not so much on development.
4. Open source culture and etiquettes: This is perhaps one of the biggest inhibitors. More than the student community, the large IT services companies employing over millions of developers are doing very little to promote contribution even though many having been benefiting from open source. Taking to a few senior leaders it was evident that given the transient nature of their business it is not yet very strategic and therefore not recognized as important. This needs to change. It would be useful to get into a situation where companies build practices around open source much like they do today around SAP, Microsoft, Oracle etc. It is going to be important for developers to have them on their resume as well. This may not be huge in terms of open source community, but still a key factor for driving adoption. The reason is simple – companies try and position technology usage more around the available skill set rather than pure technical merit.

Of course there were many more factors that emerged (we will share them subsequently) but these were sort of identified as key and therefore folks decided to work on these and see if anything can be done to bootstrap and facilitate ways of overcoming some of these challenges for open source and the knowledge industry in general – sort of knowledge fusion and hence the name Kenfuse. The sole goal being, how can one engage a larger community beyond the development group; drive awareness and make participation easier and make 2009 a year where open source reaches dizzy heights.

Kenfuse is a modest beginning to provide a platform for students and mentors to engage with the limited understanding that we have at this time. We obviously as open source enthusiasts would continue to listen and figure out the better ways and also look towards the open source community, the consumers, the contributors to help grow this initiative. We see this as ‘win all’ – a ‘positive sum’ game where everyone benefits. Lets make it spectacular!