[PRL] Benjamin Pierce: Types Considered Harmful

John Clements clements at brinckerhoff.org
Fri May 30 15:24:25 EDT 2008


On May 30, 2008, at 7:15 AM, 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? A
> couple weeks ago Steve Yegge argued against what he saw as the  
> principal
> complaint about *dynamic* types, and it was also performance.  
> Everyone's
> beating on the same straw man!

I agree that slide 19 erects a straw man, but I can see where he's  
coming from. Note that slide 19 doesn't directly suggest that types  
slow things down, but rather that types lead to safety, and that  
safety slows things down.  I'm not so sure about the causality between  
types and safety, but there are plenty of people who feel that safety  
isn't worth the time penalty.  On the other hand you have people like  
me, who can say "30x slowdown versus C?  Sure, that sounds tolerable."

You're right, though, he's not really comparing safe and unsafe  
systems; the rest of his talk appears to take safety as a given, so I  
agree that the talk would have been better without this slide, except  
that it appears to unintentionally highlight the much larger war of  
safe vs. unsafe systems.

John Clements

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2484 bytes
Desc: not available
Url : https://lists.ccs.neu.edu/pipermail/prl/attachments/20080530/d6608da9/attachment.bin 


More information about the PRL mailing list