Marcus/Ray -
I agree wholeheartedly with both of you. I've encountered more than
enough of both NIH and stubborn re-invention, especially in the Academia
and the National Lab context.
This has been one of my biggest challenges as a mentor of young people
who have *plenty* of book learning but not much practical experience...
learning the balance between (motivated) learning and wasteful
re-invention.
I am *not* a trained engineer, though a few phases of my career I have
in fact done systems and software engineering. I am *most* interested
in exploration, discovery, and innovation and try to arrange my work
life so that such activities *are* appropriate.
Many of my (re)inventions are relatively subtle (I think). I was once
trying to expand on a set of graph analysis tools to do some consistent
comparison of complex graphs from a wide variety of disparate sources.
No one graph library really had the full suite of tools I needed and
even within a given library, the actual execution complexity (space and
time) were not consistent. Where I had source to review, I discovered
(unsurprisingly) that the libraries were collections thrown together
from more than one source (multiple graduate projects by different
students?). I even found some latent bugs in a few of them. The
only way to get consistent results was to in fact re-implement these
algorithms. This was well over 10 years ago before graph analysis
became so "en vogue".
What I tripped over in the process was the "Tree Heap" or "Treap"...
but I didn't trip over it by doing research... i tripped over it by
*needing* a data structure that had those properties, so I ended up
building one from whole cloth... only to discover months later that such
an data structure had already been devised. To add insult to injury, a
few years later, I was relating the story to a woman who had come to
work at LBL while I was there, and *she* was one of the original
discoverers/inventors of the Treap! It is truly a small world.
To make this relevant to the discussion... I don't think I could ever
have come to recognize the value of such a data structure if I *hadn't*
felt obliged to re-invent (re-implement?) a number of algorithms that
had already been implemented by others... to differing degrees of quality.
- Steve
On 2/20/14, 5:47 PM, Parks, Raymond wrote:
However, the practical engineer in me wants to scream whenever
someone reinvents stupid ways to do things.
It is indeed infuriating when someone makes no effort to learn about
what the state-of-the-art is and imposes their ignorance and
incompetence on other people. Being curious and fearless (motivated
learning) is not the same thing as not-invented-here syndrome and
being stubbornly illiterate.
Marcus
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
