Subj : Re: Compiler and an interpreter
To   : comp.programming
From : Jon Harrop
Date : Sat Aug 06 2005 02:24 pm

Gerry Quinn wrote:
> In article <42f394ef$0$91510$ed2e19e4@ptn-nntp-reader04.plus.net>,
> usenet@jdh30.plus.com says...
>> Gerry Quinn wrote:
>> The real issue is that the ratio of OCaml to C++ code size increases as
>> the absolutely code size increases. On 100 LOC programs I typically find
>> C++ to be twice as verbose. On 1,000 LOC programs C++ is more like 5x as
>> verbose. The largest project of my own projects replaced ~100kLOC of C++
>> with 4kLOC of OCaml.
> 
> 25X seems a bit extreme.  That 100KLOC was your own C++, right?

No, that was the part of Mathematica that I reimplemented in OCaml.

>> > I doubt that, unless he really hates short functions!  Maybe we could
>> > say that writing such functions, even if they are one in a hundred,
>> > might well take a significant fraction of the programmer's time, if the
>> > task is math-intensive.
>> 
>> What exactly do you mean by "math intensive", e.g. are algorithms and
>> data structures "math"?
> 
> I don't include data structures and simple algorithms like iteration. I
> mean stuff like your example of manipulating periodic graphs.

Ok, then I'll object - languages like OCaml are much better at data
structures and algorithms than languages like C++. This is why I find them
useful. Indeed, this is why I first switched from C++ to OCaml for my work
on vector graphics.

-- 
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com

.