[PRL] reliability vs market forces

[Matthias Felleisen]

The paper bubbled to near the top of my stack a number of times.

1. It is good to read that the systems people are as far removed from  
the market as the PL people.

2. It is also good to see that someone in the systems community  
public says how far removed from reality -- to the point of complete  
irrelevance -- the Theory of Dist Comp people were. I tended to sense  
this watching numerous people at Rice, and my colleagues would mumble  
"toy mathematics" after such talks, but nobody had the guts to say  
this aloud.

3. It appears to me that the authors fail to expose the obvious  
dilemma of the discipline in general:

The parallel development of CS as an academic discipline and Software  
as probably the largest manufactured product slice of the economy has  
left us with a deep split between the two. Academia is over-populated  
by toy mathematicians who produce occasionally nifty philosophical  
insights but by and large are unaware and uninterested in the  
problems that emerge from the world of "practicing software  
engineers." (Instead they read books for the working category  
theoretician -- an oxymoron if there is one). Industry is over- 
populated by get-it-done-no-matter-what programmers who aren't worth  
their salary. They do come up with products that people want to use  
and some of them have enough reflective power to pose problems to  
academia but by and large they don't care what comes out of academia  
and they don't care to talk to us. As Will once says "the idea that  
'test first' constitutes progress to industry says a lot about them."

I suspect that the discipline would be much healthier if "they" were  
able to articulate real problems and pose them to us and "we" would  
listen and tackle them. Conversely,  "we" should be able to routinely  
demonstrate how our work applies to theirs with artifacts and  
processes. Until they learn to stop to think and reflect and until a  
large majority of us builds products for real users, nothing will  
happen though.

-- Matthias







On Jun 28, 2008, at 6:29 PM, Dave Herman wrote:

> I don't understand everything in this paper but it's interesting:
>
> How the Hidden Hand Shapes the Market for Software Reliability
> by Birman, Chandersekaran, Dolev, and van Renesse
> http://www.cs.huji.ac.il/~dolev/pubs/MarketFailure.pdf
>
> Abstract:
>
> Since the 17th century, economists have recognized that absent
> government intervention, market forces determine the pricing and
> ultimate fate of technologies. Our contention is that the "hidden  
> hand"
> explains a series of market failures impacting products in the  
> field of
> software reliability. If reliability solutions are to reach mainstream
> developers, greater attention must be paid to market economics and  
> drivers.
>
> Dave
>
> _______________________________________________
> PRL mailing list
> PRL at lists.ccs.neu.edu
> https://lists.ccs.neu.edu/bin/listinfo/prl