[PRL] Why knowing Haskell is bad

Matthias Felleisen matthias at ccs.neu.edu
Fri Oct 6 10:50:57 EDT 2006


On Oct 6, 2006, at 10:39 AM, William D Clinger wrote:

>> If the cost is reasonable (constant factor), it'll be okay in 18 or
>> 36 months.
>>
>> -- Matthias
>
> I've been hearing that old saw for a lot longer
> than 36 months.  It might be true if you could
> hold ambitions constant.
>
> Will

I am sympathetic to this view though read on:

It was the guideline for delivering DrScheme to schools, and it  
worked like a charm. After five years we switched it to "stay smaller  
than Explorer" and that worked, too, because it leaks/leaked so much.

I know you will respond with "yes, but schools are the one place  
where ambitions are held constant and possibly even sink on a yearly  
basis." But this is not true either. When people switch to Java from C 
++ (and they do so a lot), they take a performance hit to this day.  
And not a small one at that. So what gives?

-- Matthias






More information about the PRL mailing list