[PRL] The Importance of Excel

[Matthias Felleisen]

When I first received this link early in the week, someone suggested that
our unit checking system for Excel (ICSE 2004) could have caught the bug. 

But if I understood the paper correctly the formula should have been 

     N 
------------
average(K,L)

but was entered as 

     N 
------------
   K + L 

So this is neither a type nor a unit bug but a truly logical bug. If this
model had been treated as a real program, we might have been able to specify
an expected interval and the factor of (K+ L) vs average(K,L) might have been
caught.  

-- Matthias




On Feb 14, 2013, at 9:20 AM, Mitchell Wand wrote:

> http://baselinescenario.com/2013/02/09/the-importance-of-excel/
> 
> See also the HN discussion, linked at the end.
> 
> Yes, I know we've done type-checking for Excel.  What else might we be able to do to help?  (See the HN discussion for a discussion of the real difficulties).
> 
> --Mitch
> 
> 
> _______________________________________________
> PRL mailing list
> PRL at lists.ccs.neu.edu
> https://lists.ccs.neu.edu/bin/listinfo/prl