Re: Speedy compilers (Michael Wolfe)
19 Nov 1998 23:22:32 -0500

          From comp.compilers

Related articles
Speedy compilers (Kirk Abbott) (1998-11-06)
Re: Speedy compilers (Janusz Szpilewski) (1998-11-15)
Re: Speedy compilers (1998-11-19)
Re: Speedy compilers (Jack W. Crenshaw) (1998-11-19)
Re: Speedy compilers (Toon Moene) (1998-11-21)
Re: Speedy compilers (Joachim Durchholz) (1998-11-24)
Re: Speedy compilers (Andrew Fry) (1998-11-24)
Re: Speedy compilers (Robert Bernecky) (1998-11-24)
Re: Speedy compilers (1998-11-30)
[14 later articles]
| List of all articles for this month |

From: (Michael Wolfe)
Newsgroups: comp.compilers
Date: 19 Nov 1998 23:22:32 -0500
Organization: The Portland Group, Inc. Wilsonville, Oregon U.S.A.
References: 98-11-047 98-11-086
Keywords: performance

> So compile speed is a very important factor and compiler designers
> should treat it seriously...

One lesson I learned from too many years in graduate school and many
more in industry is it is commercially infeasible to try to sell a
slow compiler. Multiflow had a wonderful, highly optimizing,
well-structured, but slow compiler, and it was a factor in losing
machine sales. We still see research papers claiming "...users will
be willing to spend the additional compile time to benefit from the
obvious run-time improvements..." Sorry, folks; your research may be
interesting and even important, but commercial compilers have to be
fast. Of course, researchers don't depend on people actually buying
their compiler, so it's hard to be convincing.

A study in human psychology might decide when a compiler is 'fast
enough'; I suspect users would claim a compiler is 'fast enough' if
the response time is close to instantaneous, regardless of the program
being compiled.

-Michael Wolfe
[Given the relative speeds of Turbo C and Visual C++, I expect that this
lesson may no longer apply. -John]

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.