[Openmcl-devel] prompt-for-file (embarrassing question)

Paul Krueger plkrueger at comcast.net
Fri Mar 31 07:46:30 PDT 2017

So it wasn’t my intent to contribute to any sort of inter-generational squabble. Each generation has its own opportunities and challenges to deal with. About the only thing that I might have that my younger colleagues lack is some sense of the history to date of the software industry. I wrote my first real computer program almost exactly 50 years ago (it immediately went into an infinite loop). I have a doctorate in computer science (AI), have written real code using about 30 different languages (Lisp is still my favorite) and was eventually a corporate executive in the industry. I’m now retired with an opportunity to write whatever code I want. All that is unimportant except that it provides me with a particular perspective on how the industry has evolved.

One of the things that I’ve witnessed is a truly enormous amount of creative energy devoted to redeveloping the same basic things over and over again on newer hardware with newer interfaces. That’s not to say that nothing new is being done. On the contrary there are some fantastic new things being developed every day. To a large extent that is due to the rapid improvement in hardware capability and lots of well-trained young developers. There were so many things that we dreamed of doing that were just impossible on the equipment that was available. So I’m truly envious of younger developers and the capabilities that are at their disposal.

At the same time I know why some things ultimately failed. Open source common interface efforts failed because they ultimately couldn’t keep up with the changing capabilities of hardware and new vendor-specific api’s. Some seemed to be well on the road to more general use before they lost momentum and just faded away. Projecting just a little, I expect that common or non-vendor-specific interfaces developed today will likely meet the same fate. That’s not to say that they can’t be useful for some period of time, but if they do ultimately fail, as I suspect they will, then the useful things that were developed using them will have to be re-developed once again. I know that some of the things that I’ve done have been ported multiple times just in my lifetime.

I don’t have a fix for the problem, all I have is information about what seems to me to be the underlying reason for it. Things change and people change and fresh ideas may ultimately provide an answer that my generation missed. Or perhaps there is some value in each generation having to re-solve the problems that were dealt with by their predecessors. Maybe that’s the only way to really learn. I also have a theory about why technologies like Lisp are less successful than abominations like C++, but perhaps that’s best left for another time.

Good luck to all!

More information about the Openmcl-devel mailing list