My interview with Borland's Tod Nielsen on the future of development producitvity
Our customers are asking us to pursue innovation on the pre-deployment side ... By automating the menial tasks, individual developers can now spend more time innovating, delivering value and being a better partner to the business.
Borland Software Corp. has well-earned its place in the pantheon of venerable innovators of software development productivity. The company's sterling product heritage is outdone only by its record of trying to reinvent itself, of making bold moves into new markets. The process continues with the Silicon Valley company's recent news of IDE product spin-offs and acquisitions. So I thought the time was auspicious to pose some questions to Borland CEO Tod Nielsen, to plumb the depths of Borland's new strategy and direction of broadly improving how software is created. The timing is particularly relevant, given Borland's active role in the Eclipse Foundation, with noted irony, and with the EclipseCon event beginning March 20 in Santa Clara, CA. Hope to see you there, BTW. The following is an email interview I conducted this week with Nielsen. Full disclosure: Borland is not a client of Interarbor Solutions. Dana Gardner: Borland has been assembling and delivering an application lifecycle management (ALM) solution, what you refer to as software delivery optimization (SDO), for several years. What prompted the timing now to distance yourselves from the integrated development environment (IDE) vendor role? Tod Nielsen: It’s an issue of focus. SDO is a long-term vision for how we help make the overall process of software more manageable and predictable than it is today. It’s no longer just about individual productivity tools. It’s about how teams work together and how they work within the larger organization. We’ve been investing in growing our ALM portfolio to include areas such as software process improvement, IT governance, requirements definition and management, modeling, change management. Now, with the pending acquisition of Segue Software, we’ll be able to offer testing and lifecycle quality management to our ALM portfolio.
We believe that spinning out our IDE product lines will do two things. First, it will enable us to focus solely on ALM and make quicker progress on our SDO vision. Secondly, it will allow our IDE customers to get the level of attention and investment they require. Both the IDE and ALM markets are important, but each requires a unique strategy, and dedicated attention and resources. Borland couldn’t do both well.
Gardner:This seems to mean that Borland is IDE ecumenical and will therefore innovate and add value above, below, and beyond any tooling environment. As you focus on applications management and deployment efficiencies, what then are the boundaries of what Borland's SDO value offers? Please put a "box" around what it is you will do, so we get a sense of what you won't be doing as well. Nielsen: Yes, we believe this will allow us to move “up the stack” to not only be platform-independent, but also language and IDE-independent. The average developer today has many IDEs on their desktop. As such, it makes sense to enable multiple tooling environments.
Moving forward, we will continue our focus on helping organizations streamline the processes that support the delivery of quality software. This includes giving management the technology and information they need to have more visibility and control over this process. Our focus is on the pre-deployment side of the house versus the post-deployment/operations side. There are many vendors working to make operations more efficient. Our customers are asking us to pursue innovation on the pre-deployment side.
What we won’t be doing is alienating developers. Developers have and will continue to play an essential role in the software delivery process. Our heritage is obviously with developers, and even though we are spinning out our IDE product lines, we want to continue to have tight relationships with this community.
Gardner:Getting your requirements right before needing tooling seems obvious in development but remains treacherous. Do you see advancement in requirements optimization a function of products or methodologies/professional services? Do you expect over the next two years that Borland's professional services revenue as a portion or total revenues will grow and by how much?
Nielsen: It isn’t a question of one or the other, mastering requirements requires a combination of people, process and technology. Without process, defining and managing requirements is undisciplined. Without people trained to apply those processes, it’s unreliable. And without the technology to automate and enforce those processes, it’s inefficient. Our solution includes a combination of process improvement services, skills training and our ALM technology. By packaging it and delivering it in a disciplined and personalized way, we hope to make it easier and quicker for customers to see success. Too often in the past, vendors have addressed requirements in a one-size-fits-all and technology-centric way. We want to change that. I don’t expect a significant change in the mix of services and ALM license revenue, although eventually I expect we’ll see a higher concentration of license revenue as more customers learn to adopt and use our products.
Gardner:SOA management and governance is a hot area now, rife with consolidation. How will Borland SDO embrace and extend into SOA governance, which has a possible future as encompassing more overall IT and process governance?
Nielsen: SDO addresses the biggest issues of improving IT performance and maximizing the business value of software. This includes SOA management and governance. SOA is an initiative within many of our customer’s IT plans.
For example, our IT Management & Governance (ITM&G) solution helps organizations adopt processes and technology to automate areas such as demand management, portfolio management, project and program management, as well as resource, financial and asset management. These asset management capabilities can help organizations better manage and govern SOA initiatives, and we’re looking to expand these SOA management capabilities throughout our ALM platform in the future.
Gardner: Are we headed toward a day when management and governance in IT merges with and perhaps dominates general business management and governance? What is the relationship between IT governance and general business governance for technology-heavy-use industries? Nielsen: With technology impacting — and often driving — nearly every key business decision made today, I think it’s likely that we will see IT and general business management and governance merge in the near-term. You can already see how federal mandates, such as Sarbanes-Oxley, have not only impacted the financial departments in organizations, but also IT as IT departments strive to document and ensure business processes are repeatable and auditable. Gardner:As more applications, especially on legacy platforms, are sun-set and/or ushered into services, shouldn't SDO soon refer instead to "services" delivery optimization? Specifically, what does the Borland product roadmap hold for SOA and BPM? Nielsen: Behind every good service lies good software. SOA and BPM are definitely key factors in our decisions about future product functionality. For example, our Together modeling products already support service and business process modeling. Companies using Together can define their business processes and generate Business Process Execution Language for Web Services (BPEL4WS) as well as architect the components used in SOA. Our requirements definition and management, change management and lifecycle quality management solutions also support the definition, management and testing of SOA components and integrations. But ultimately we want to serve our customers by solving the broader business challenges that are hindering the success of software development. Gardner:You've been deeply involved with application development and deployment strategies for many years. What if anything is different about the current trends and direction of the technology and industry? Nielsen: Automation is contributing to many of the productivity gains we’ve seen in the last few years. I’ve heard a lot of people worried about the effect that automation is having or going to have on development -- that it’s taking away the creativity of the process and shifting power away from the individual. In reality though, I think the opposite is true. By automating the menial tasks, individual developers can now spend more time innovating, delivering value and being a better partner to the business.
Open source is also a big change. The commercial software industry can learn a lot from the transparency and individual accountability that’s such a vital part of the open source development process.
Gardner:Anything to add to the Eclipse and/or NetBeans discussion now that you're moving away from IDEs?
Nielsen: One of the things Eclipse struggles with is that most people believe Eclipse is only an IDE. In fact, Eclipse is much more. This community is trying to create an open and non-proprietary framework for handling important lifecycle issues such as business intelligence reporting (with the BIRT project) or performance management (with the TPTP project). There will always be more than one IDE out there -- this community thrives on choice and there continues to be a need to improve individual productivity as the world of development changes. Yes, there will continue to be consolidation and commoditization within this space. There’s nothing new here. This process allows vendors to add value over and above the basic level, which is what ultimately drives innovation. Gardner:As you forecast what is best for TCO and operational efficiency for your customers, what do see as imperative for server-delivered client GUIs? Are you a sympathizer to the AJAX and rich Internet applications hoopla? Nielsen: Borland supports choice. We always have, we always will. There are many reasons to go thin client and many reasons to build fat clients -- every customer is different. What organizations need is a consistent process to maintain visibility and control for all of their software development efforts. This goes beyond desktop applications, but also includes integrating services and implementing and customizing applications. Gardner:What's the best advice you could give an enterprise IT strategy executive for what will best prepare their organization for SDO efficiency and competitiveness over the next three years? Nielsen: My advice to CIOs who want to be a strategic business partner is to fix one of the weakest links in IT -- the chaotic and unpredictable way that software is developed and delivered today. Every CIO knows that software, if it’s done right, not only supports the business but it drives innovation and helps build leaders.
Start by taking a look at the maturity of your existing development organization and find the pain points. Is it your ability to respond to external changes or compliance mandates? Is it your ability to balance a more strategic portfolio of projects? Is it the growing amount of rework or poor quality that you’re suffering from? Once you identify the pain, you then need to make a choice to ease that pain. A little bit of investment in your processes, people, and technology will go a long way and make an enormous difference in short-term efficiencies and long-term capabilities.