[lug] LinuxJournal: Introducing the Open Cluster Framework

[Alan Robertson]

J. Wayde Allen wrote:
> On Thu, 5 Sep 2002, John Dollison wrote:
> 
> 
>>Linux should be an abysmal failure. It breaks every single known rule
>>of software development."
>>
> 
> Does it really ... I'm not so sure?  Perhaps the emerging rules of
> intellectual property are breaking the rules of meaningful human
> interaction ...?

Rich did a pretty good job of quoting me, compressing down many hours of 
interviews into a relatively short article.  He didn't always do a perfect 
job of capturing my thoughts or words though.  And even when he did a good 
job, I might not have stated it completely clearly.

In this case what I meant was more like if you look at what is taught in 
software engineering classes, the vast majority of formal methods taught as 
being necessary for quality software development are wildly at variance with 
normal Linux practices.  Since I had seen a lot of really good software 
developers and even a few great ones, NONE of them followed any of these 
formalisms.  Yet they produced much better code than others - even those who 
followed more rigorous formal procedures.  Linux captures the simple idea 
that good people when properly motivated and left to their own devices 
produce good work.

The "surgical team" model is the only exception I know of.

It also borrows from more normal human endeavors things like 
apprenticeships, and cooperation, etc.

	-- Alan Robertson
	   alanr at unix.sh