December 6th, 2012 — Software
We recently wrote about a disruptive trend we are following along with cloud computing, devops and open source software in the enterprise. Our 451 Research subscribers also got a preview of our findings in a recent spotlight report.
Although there is added pain in programming with multiple languages, benefits such as scalability, interoperability and concurrency increasingly necessitate it for optimal efficiency and quality.
Now we are pleased to present our latest special report, ‘The Rise of Polyglot Programming.’ The report investigates the drivers, disruption, challenges and opportunities from the trend. We also present market sizing and growth implications for polyglot programming, drawing on data and analysis from our Market Monitor service to show how polyglot programming will be part of a growing opportunity worth more than $35bn by 2015.
May 22nd, 2012 — Software
451 Research was pleased to collaborate on the Future of Open Source Survey 2012 with North Bridge Venture Partners and Black Duck Software. This year’s survey garnered 740 responses from a variety of vendors and non-vendors in the industry. Overall, the survey highlighted some subtle and sometimes dramatic changes in what is driving open source software. It also made clear that while there is still a good degree of education and awareness yet to occur around open source software, there is a large amount of open source code making its way into today’s enterprise, webscale, consumer and other computing environments.
Some of the key findings:
*The survey reinforced the prominence and influence of open source software in the enterprise and in key trends driving it, as we and others have highlighted for some time with reports such as Seeding the Clouds and Mobility Matters. When asked which technology areas would see the most significant open source software community innovation from, respondents ranked ‘cloud’ highest at 40%, then ‘mobile apps’ (19%) and ‘mobile enterprise’ (15%) for a combined 34%, then ‘analytics’ with 10%. These areas are indicative of where we see open source software projects, communities, vendors and consortia continuing to broaden use of open source software.
*The survey asked what are the top barriers to selecting open source software when compared with proprietary alternatives, resulting in unfamiliarity (48%), lack of internal technical skills (47%), lack of vendor support (35%) and legal concerns about licensing (33%) as the top answers. Although this indicates there is still some trepidation and lack of awareness around open source and commercial options for support, other survey responses indicate open source software is still spreading to new industries and customer categories. When asked about the most important trend for open source software over the next two to three years, respondents identified the top choices as: adoption in non-technical segments such as government or healthcare (42%); enterprise adoption (40%) and growth in industry-specific communities (10%).
*The survey also showed there is a heavy volume of new, meaningful code coming out of open source software’s many communities. When asked what share of their deployed code they anticipate will be open source software over the next five years, about one third of survey respondents (32%) reported open source had already reached major deployment at 75% or more of their code. Another one third of respondents (30%) said open source will make up half to 75% or more of its deployed code. About a quarter of respondents (23%) indicated open source would make up 25-50% of their deployed code over the next five years, while 15% of respondents said the open source share of deployed code would be a quarter or less.
*We also saw a high rate of open source participation from the survey. When asked about community engagement with open source and their preferred method, 49% of respondents said consuming code, 36% said reporting patches or fixes, 31% said contributing new features, 28% said initiating new projects, 25% said contributing through partners or industry alliances. We believe this shows a high rate of open source participation beyond using code, which is also a meaningful contribution. This also indicates a greater willingness to get involved with open source projects and to start new projects.
*The survey also highlighted the changing drivers of open source software in the enterprise. When asked what are the top factors that make open source software attractive, respondents identified freedom from vendor lock-in (60%), lower acquisition and maintenance cost (51%), better quality (43%) and access to source code (42%) as the top answers. While we had seen vendor lock-in fade as a factor and cost as paramount two or three years ago, today vendor lock-in has become much more of a factor for customers. We believe this has to do wtih cloud computing and customers’ desire to maintain flexibility as they figure out how to best leverage cloud resources. The survey also showed that cost, which we also equate to time and efficiency, is always a strong factor, with 62% of respondents identifying reduced cost of development and maintenance as the main reason they use open source or initiate projects.
*The survey also reinforced our belief that while open source software lays the groundwork and underlies much of cloud computing, the cloud is also giving back to open source by providing vendors a way to differentiate free downloads from paid, cloud-based services. In fact, it seems support and services subscriptions are a much higher priority for open source software vendors than so-called ‘open core’ models that provide software for free and certain extensions, features or support as paid. When asked which revenue generation strategies are likely to create the most value for open source vendors over the next two years, respondents ranked an annual, repeatable support and service agreement as the top answer (52%). Other open source revenue models, such as ad-hoc services and support (41%), value-add subscription (40%), hosted or cloud software services (38%) all ranked higher than a closed-source license or open core model (12%).
For our full analysis on the results of the 2012 Future of Open Source Survey, see our Spotlight report. The results were also presented this week on a panel at the Open Source Business Conference and that presentation is available at the Open Source Delivers blog.
March 22nd, 2012 — Software
Last year, I wrote about the key pillars of openness in today’s enterprise IT industry, highlighting open source software, real open standards, open clouds, and open data as the ‘Four Pillars of Modern IT Openness.’
More recently, I wrote about what I now consider to be the fifth pillar, which is open application programming interfaces (APIs). Of course, when we talk about ‘open’ anything — open source, open standards, open clouds, open APIs — there tends to be debate about what is really open, how we should define open and who should or should not be able to carry the phrase. My focus on open APIs and on APIs in general generated some good discussion, as well as some pushback, regarding the value of APIs compared to open source software, which APIs are open, and how open is open enough?
I want to make clear I am not saying open APIs are better than open source. The real point is that the activity, development and innovation happening around APIs — particularly as cloud computing and hybrid public-private use continues to evolve — is reminiscent of the way open source software began disrupting the industry some two decades ago.
The other point is that while customers are typically interested in open source software for flexibility, cost savings, mitigating vendor lock-in, performance, ROI or other reasons, my conversations with both vendors and customers indicate much of the integration in the cloud centers on the openness of the APIs. When customers have stable, documented APIs, it is often more conducive and effective to work there, rather than on the source code. If code, development and deployment are disrupted by closed, changing, weak or undocumented APIs, then developers, customers and the market are likely to quickly move on to other APIs, perhaps ‘open APIs’ that are well documented and include examples. Similar to the other pillars of modern IT openness — open source software, open standards, open clouds and open data, open APIs are most effective and efficient when combined with the others.
Let’s not let open APIs become another version of ‘open standards’ that were anything but 10 years ago. Instead, we should seek to use and call out truly open APIs, which would typically mean connection to open source software, open standards, open clouds and open data as well. However, we must also be aware of the threat, competition and pressure from APIs such as Amazon’s EC2 and AWS interfaces, which are not open source nor open standards, but nonetheless may be open enough for a majority of developers and market.
March 3rd, 2010 — Software
We continue to see and hear signs of a new movement in enterprise IT: devops. The term has for some time been a reference to the blending roles and benefits of an integrated approach when it comes to the ‘development’ of an application through testing and QA to production and ‘operations.’ We cited this as the focus of SpringSource’s acquisition of Hyperic and then VMware’s acquisition of SpringSource last year. Still, we see the trend today on a few different levels, all of which are drawing in vendors eager to meet the opportunity.
On its most basic level, devops is about people. Developers and admins/operations have historically been separate camps within enterprise organizations. While there has been some integration and collaboration, helped by open source software and social networking trends, there seemed to be a persistent disconnect between these two parts of typical enterprises. Similar to how open source software developers have come a long way in considering usability, user interfaces and users in general, we now see enterprise developers taking into consideration the deployment and use of the software. I’ve also actually met some of these devops, hired to strattle the software from creation to consumption, and I believe it is a growing job title for enterprise IT. Further reinforcing the devops people factor, CollabNet indicated a need to manage not only code and applications with agility, but also to similarly manage people and teams as a driver of its acquisition of Danube, a project management company.
Of course, most of the people involved in devops come from the software development and IT operations worlds. At the developer level, we see open source software tools and practices, Web and agile development all contributing to devops, whereby development and production are getting aligned.
On a more technical level, we again see open source and agile development practices, but with the addition of cloud computing as the delta of all of these trends.
Yet another level of devops is the business level, whereby not only developers and operations folks are getting involved, but the executives and people representing business and application requirements, needs and realities are also finding they have a stake in devops. This is something we’ve already seen in open source software thanks to vibrant communities of not only developers, but users too, particularly in specialized fields where their stake in the development and deployment of the software is more critical. For example, we’ve heard from many vendors how their open source software for the healthcare and government sectors is pulling in new stakeholders and experts, including doctors and nurses, who are directly reporting their needs, issues and realities to developers, operations and in the best case scenario, to devops.
One thing all of these different levels of devops have in common: they are all driving activity and strategy among a wide array of vendors and projects. After all, at yet another level, devops represents the promise of cloud computing, elasticity, doing more with less and driving synergy and efficiency through people and technology.
We will be delving much deeper into this subject for a special report from our Commercial Adoption of Open Source (CAOS) and Infrastructure Computing in the Enterprise (ICE) practices planned for this summer. We look forward to speaking with more devops players and hearing what they think of the trend.
January 28th, 2009 — Software
To the disappointment of some and perhaps the relief of others, this blog is not about extra-terrestrials. No, instead it’s about a different ET, one that has far more relevance in the technology industry and open source software: energy technology. While current global economic conditions may have dimmed the brightness of ET and the business to burgeon around it, at least for now, it appears there is still ample opportunity for energy and technology to serve as solutions to many of today’s problems. The new U.S. administration under Obama, which we reported is seeking counsel on use of open source software, continues to keep energy high on its agenda as one prominent example.
Well-known American journalist and author Thomas Friedman wrote much about ET in his recent work, ‘Hot, Crowded and Flat,’ which covers the threats of a warming Earth and growing population, as well as the opportunities in solving the dilemmas, particularly energy. Friedman cites open source software as one of his ‘flatteners,’ a factor or impact that will work to level the competitive playing field on a global basis. Friedman demonstrates a decent understanding of open source, but I wonder what the role and impact of open source practices and strategies of sharing and openness will be within the emergent energy technology industry.
Free and open source software has certainly had a far-reaching and deep impact on enterprise software development and business and arguably on the IT industry as a whole. Will any of the entrants into ET see the potential to take this tool of development and distribution to help spread the next best forms of power generation, distribution and use?
There are some precedents to look at: open source has for some time played a role in biotechnology, where researchers see more benefit in collaborating than in competing on many aspects of their sophisticated research.
We’ve also reported on Collaborative Software Initiative, the startup from former OSDL director Stuart Cohen that is finding a place for open source style collaboration among giants in finance, public health and safety and other areas.
With open source software and the principles of transparency, collaboration, flexibility, innovation and quality that go along with it finding their way into so many verticals and industries, it seems logical open source will have a role to play in ET. The extent of that role may help determine the winners and losers in this emerging space.
November 4th, 2008 — Software
I wanted to write an Election Day blog to coincide with the presidential contest playing out in the U.S. today, but I must admit to some apprehension. You see, like many in the tech industry, I have seen the predominately negative reaction of mixing politics and technology.
I was therefore not too surprised to see the same type of thing playing out with Tim O’Reilly, his blogged, personal endorsement of Barack Obama and the reaction.
Tim leads us to a good question to ponder and discuss: is a political endorsement appropriate for a technical site. I’d take it a step further: should we be mixing politics and technology? We must first acknowledge the presense of politics in all things from families to governments to communities to commercial products. However, given the bitter reaction to small doses of political perspective offered by myself, O’Reilly or any other number of people, I have to wonder why people are so vehement about the need for a separation between politics and tech?
The negative reactions, which are typically peppered with supportive reactions that are admittadly politically suspect, do not typically come from a political counterpoint. Instead, people just seem pissed off to the point of threatening not to visit your site or buy your stuff at the very idea you have introduced politics into their world of tech. Maybe people just want to avoid politics in a place — whether it’s the latest software project or code they are working on, the new architecture model they’re pondering or the best way to get ahead technically — where they are not necessarily required to factor in or fiddle with politics.
I may be able to relate. When I began writing more specifically about technology and the IT industry (circa 1998 and the eve of Microsoft desktop dominance, the birth of Linux, open source and much of what I still cover today), I was glad to be out of the traditional news business, which was then focused on the Clinton impeachment. Technology was a way to step completely outside of that drama and noise, and still write about important, interesting stuff.
I’m not really advocating a separation of politics and tech or between tech and state, but I think I may understand why there is such strong resistence of comingling the two. As Tim O’Reilly asks, what do you think?
July 3rd, 2008 — Software
Score one for open source software in the history and record books. Mozilla announced its Firefox open source Web browser set a Guinness World Record for number of downloads in a day, notching more than eight million downloads in 24 hours June 17-18, 2008. The day also marked the release of Firefox 3, the latest version of the browser that has steadily built its user base up from a fraction of the browser market to a solid, near 20% share today.
There were a few hiccups during Firefox’s so-called Download Day, a PR stunt that could have easily been dubbed Delay Day by those undable to obtain the latest version. I think it’s interesting that even in this day and age, even when you are geared up for spikes and infrastructure is far more flexible and agile, there’s still downtime with the rush. Nevertheless, it successfully drew attention to Firefox and more importantly, the browser has been praised for its security and appeal to many users.
However, Mozilla has also been questioned for not focusing more on the enterprise opportunities for Firefox. Given its security, flexibility and customization advantages, I see Firefox as a perfect fit for enterprises, and I’m aware of several vendors and companies in tech and other industries that encourage or enforce use of Firefox. However, Mozilla seems content to allow consumer and home use to seep into the enterprise.
I believe it is wise for Mozilla to focus on user features and innovations, to look to advertising for its revenue and to refrain from enterprise sales departments and staff. However, the company should be doing all it can to provide the features and functionality that will ease Firefox use in the enteprise. This is a matter of both perception (‘That browser isn’t made for use in our business, is it?) and reality (I can’t push proxy settings across numerous installs), but they are issues that can be addresssed by the involvement, speed and variety of open source software development.