HPE NonStop Solutions: Always on and always adapting
Be aware, complexity is your enemy, the industry could learn from that!
There is something about having a long experience in this industry that is of high value to us because we get to witness how different technologies are born, going through a hype cycle, being adopted in very specific solutions or getting wider adoption or in some cases just being forgotten. For customers it may be a daunting task to decide where to invest. So many new technologies simply claim to solve world hunger. In general it is rare that a technology becomes universal across all vendors. TCP/IP can claim that, for languages C/C++ and Java have had success, for operating systems, is Linux the standard now? But what else? When it comes to programming for example, not long ago, multi-threading was the common way to scale an application. Then BigData proved that “scale out” may be a better way to address massive scale for data access. I think we knew this already in our NonStop world but the industry didn’t.
While it is very hard to predict which technology will succeed and to what extent, I believe there is one pattern that we can observe and learn from which is that, too much complexity is likely to come back with a revenge at some point in life of a product or technology. There are many examples that come to my mind. Corba and OpenStack for example had very ambitious goals, yet they failed, I believe partly because of complexity. Python is making the headlines recently as reaching the top 3 of most used languages, even though it was known as slower and single threaded but its simplicity to the end user makes the difference. We also have seen many new languages trying to better address concurrent programing such as Node.js or Erlang. What is their claim? Thread programming is too complex. Maybe too early to say but I wonder if Blockchain as a business technology may face similar problems and risk being doomed by complexity.
In NonStop it is part of our DNA to strive to make things simpler for our end users. It starts at the Operating System level where very complex problems such as using a distributed file system just works out of the box. Seeing the relational database as a single entity just works out of the box. Why is that important? I would quote one of the biggest brains of our time for tackling complex programming concepts (such as semaphores for example):
“Simplicity is a prerequisite for reliability”
-Edsger W. Dijkstra (1970)
I cannot agree more. I think these days that some of this complexity could be inherent to Open Source based projects. Yes, having thousands of “free” contributors makes miracles to create new technology but they typically solve one problem in isolation and often overlook the bigger picture complexity that in many cases eventually shows as too high of a price for the given benefit.
In this edition of The Connection, you will see how much we work on simplifying the customer experience. Should it be SQL/MX “Just add nodes”, with Database Services “One click deployment”, with high integration of security or replication solutions, we continuously work to try to take ownership of the complexity so that our end users don’t have to . . .
Roland Lemoine
NonStop Product Management Database
Edsger Wybe Dijkstra (May 11 1930 – August 6 2002) was a Dutch computer scientist and an early pioneer in many research areas of computing science. Widely considered as one of the most influential members of computing science’s founding generation, Dijkstra helped shape the new discipline from both an engineering and a theoretical perspective.
Be the first to comment