Another gem from JP, he has blogged about Alan Kay’s talk at a CIO conference, and his interpretations on them. Very interesting and insightful thoughts. I will take a while for me to digest them, but this question did tempt to expand on it:
How come there isn’t a Moore’s Law for software?
In case you are not aware of it, Moore’s Law is a prediction for advancements in the hardware technology, which has held up quite good. Its impact is not limited to the technology, but to the entire economy around it.
So, why is there no Moore’s Law for software?
I think it is the nature of software that actually does not allow for it. If you think about it, the Moore’s law targets a component of hardware technology, the transistor. The hardware world has a better understanding of components and assembly. The software world has managed to find only more than a hundred ways of decomposing and integrating. Another factor which has helped the economic cause in the hardware world is the ability to produce and implement at a high scale. I do not see the software world doing that, every problem is unique enough to stop us from doing it.
And this makes it extremely difficult, if not impossible, to predict in the software world. After all, decades later even a single software project has started to use short cycles and iterations to avoid predicting (or estimating in this context) for a longer term.
November 6th, 2007 at 12:50 pm
I believe that software is an art form, every new release adds some extra strokes and ideas to an existing canvas of ideas. Moore’s law is more or less applicable for industrialisation, just a thought, can software ever be industrialised. We may call the current SOA principles as a path to it, but is it really leading us there, or are we creating new art forms with various brushes(web services)?