[PRL] Benjamin Pierce: Types Considered Harmful

[Felix S Klock II]

On May 30, 2008, at 3:05 PM, Richard Cobbe wrote:

> On Fri, May 30, 2008 at 02:39:43PM -0400, William D Clinger wrote:
>> Dave Herman wrote:
>>> The weirdest part was when he claimed that one of the big issues  
>>> people
>>> have with type systems is performance. Does anyone actually say  
>>> this?
>>
>> Yes, and this view was extremely influential for about
>> 15 years.  See Brian Kernighan's essay on "Why Pascal
>> is Not My Favorite Programming Language", online at
>> http://pingu.ii.uj.edu.pl/~armo/homepage/amideveloper/dev/ 
>> nopascal.pdf
>
> That 404s.  When I googled for the article title, I found
> <http://www.lysator.liu.se/c/bwk-on-pascal.html>.
>
> Perhaps this is not in fact the article that you're referring to,  
> but I
> didn't really see anything in there that addresses the claim that Dave
> mentioned above.  The closest I could get was a discussion of how  
> Pascal's
> type system prevents the programmer from writing an application- 
> specific
> alternative to "new", but nothing that I saw specifically links this  
> to
> performance concerns.
>
> On several occasions, Kernighan presents the lack of an "escape" from
> Pascal's type system as a criticism of the language, but he does so  
> largely
> as an issue of expressibility (in an informal sense only, of course)  
> rather
> than of efficiency.
>
> Which part of the article did you mean?

+1

-Felix, who was in the midst of writing a response that analyzed the 9  
point summary at the end of Kernighan's essay to argue that each of  
the 9 points were not about both types and performance (in fact few of  
the points were about either of them), but aborted upon reading  
Richard's response, and can now return to "real work."

p.s. (i'm still curious to see an explanation of slide 29.)