May 11th, 2006
The top ten most influential (Unix) technical strategies today
As lots of regular readers will know there’s a contagious, and mutating, top ten disease going around randomly picking off zdnet bloggers. Just last week, for example, Richard Koman 
fell prey to a list of most influential people in American government IT - and I can hear coughing going on all round Management issues like licensing still trump technology issues. that ring of serious looking faces at the top of the blogs page.
I’m not immune either; in fact I was going to produce a list of the top ten most needed technology words - with "oudiocy" in the number one spot - but got side tracked into thinking about the ten most influential people currently working in Unix.
Unfortunately, I haven’t a clue who they are - and neither can you because I can pretty much guarantee that at least half the people who should be on that list are complete unknowns whose current contributions won’t be understood or appreciated for another five to ten years.
Oddly, I think I know who number two will prove to be and why: Marc Rochkind - for his role on the SCO lawsuit. But that leaves nine, and so I decided to go with the top ten most influential technologies or technical trends in Unix today instead.
You could argue (and please do) about omissions or placement on this list, but here are my nominees in reverse order:
- IBM’s takeover of Linux and the consequent diminuation of innovation in that "space";
- The migration of open source ideas to business data management and applications;
- Mactel - whatever happens: a Microsoft takeover, gradual self extinguishment, or enormous success, Mactel’s fate will affect Unix development, Unix use, and Unix credibility;
- Dtrace (it’s already rewriting application development processes);
- The security tools, ideas, and management processes in use within the openBSD community;
- Plan 9 (a continuing happy hunting ground and test bed for ‘network generation’ Unix ideas);
- SOX (a major weapon in data processing’s continuing attempt to surpress Unix based computing.)
- IBM’s Cell Architecture including the physical grid on a chip and the OS accessible microcode layer;
- (Yes, a tie) "Coolthreads" and the whole set of related SMP on a chip implementation ideas;
- ZFS (like Dtrace, this is a technology that’s affecting business and development decisions throughout the industry);
- And number one? Sun’s CDDL license and the mutual non aggression pact among patent holders signing it.
Bottom line: there’s more change going on now than we’ve ever seen before - but management issues 
like licensing still trump technology issues like Microsoft’s move to PPC or Intel’s 64bit Itanic failure.
P.S. Note that I have intentionally omitted virtualization because I regard it as a step in the wrong direction - a bad idea to begin with and one that’s now being re-adopted for the same reason it became popular in the sixties: as a palliatve for shoddy systems design and sloppy thinking.
Paul Murphy (a pseudonym) is an IT consultant specializing in Unix and related technologies. See his full profile and disclosure of his industry affiliations.
Subscribe to Managing L'unix via Email alerts or RSS.
| 05/11/06
