For some 20+ years ago, the “standard best practice” of our industry
was to develop everything in-house. That is, organizations used their
own resources to develop not only the core services/capabilities/
functionality of the products they were selling/providing, but also
most if not all of the contained components, h/w as well as s/w.
In addition to that, many companies also embarked upon developing
their own tools, methods, processes, and some even the operating
systems, communication stacks and languages. The slogan of that era
was “if not invented here, then it’s no good”.
Everything, from “components”, e.g. special purpose, very expensive
rugged hardware for DoD/mil applications, or custom OS:es, to the
processes, methods and tools used for development were in-house custom
Then, sometime early to mid 90ies, came the “focus on core business”
wave, followed by the COTS trend. At the business level, the new
slogan was “focus on core business”, meaning that everything that was
not part of the core business should be insourced from outside, i.e.
everything below the top level of the capability/functionality stack
should now be obtained from outside suppliers. So, all of sudden
companies developing systems and software now turned to commercial
suppliers of boards, chips, operating systems, databases, development
tools, processes, etc (as well as luncheon canteen’s, janitor and
cleaning services 🙂
This core business & COTS trend has been pretty persistent even since
then, but now, with Open Source gaining momentum in the software
dimension, it seems that the pendlum is about to swing back: now it
appears that many organizations are starting to
look upon the “ultimate customizability of Open Source” as a potential
competitive edge: “with Open Source, we can customize all our tools/
technology exactly according to our needs, and thereby become more
efficient and gain a competitive edge” is a comment I’ve heard a
couple of times lately.
With a bit of historical perspective, it’s interesting to see how the
pendlums of our industry swing back and forth: until recently, it
would have been a career suicide to suggest going back towards the
“not invented here” approach, now it seems that people are building
their careers on exactly that approach, and the enabling technology
for this movement is Open Source.
If this trend towards “not invented here” really takes off, expect to
see a huge increase in in-house “methods,tools and tool integration
departments”, as well as a proliferation of diversity in in-house
technologies,methods, processes and ways-of-working.
The jury will be out for years to come on whether this new trend will
pay the dividends.