Related articles |
---|
Benchmark information for C compilers ananda@sequent.com (Ananda Babu) (2001-07-23) |
Re: Benchmark information for C compilers anton@mips.complang.tuwien.ac.at (2001-07-23) |
Re: Benchmark information for C compilers joachim_d@gmx.de (Joachim Durchholz) (2001-07-23) |
From: | "Joachim Durchholz" <joachim_d@gmx.de> |
Newsgroups: | comp.compilers |
Date: | 23 Jul 2001 23:22:27 -0400 |
Organization: | Compilers Central |
References: | 01-07-106 |
Keywords: | benchmarks |
Posted-Date: | 23 Jul 2001 23:22:27 EDT |
Ananda Babu <ananda@sequent.com> wrote:
>
> Does any know of any public domain benchmarks that can be used to test
> a compiler and language against other compilers and languages?
The efficiency of a language depends heavily on the abilities of the
programmer, and the time budget that's used for improving
efficiency. A language isn't efficient if it's easy for a compiler to
optimize it, it's efficient if it allows the programmer to write
efficient code with little extra effort. The problem is that a
benchmark alone isn't enough to measure this, you'd have to take the
effort that went into the software into account. The problems here is
that the effort will vary greatly according to programmers'
backgrounds. This makes benchmarking much more expensive since you
have to ask lots of people to program for an extended period of time,
so few studies are available.
Note that efficiency isn't everything, if that were the case we'd all
program in assembly and never use macros. The trade-offs between
efficiency and the other factors is difficult to judge since these
other factors are very difficult to measure and weigh.
It's no surprise that hard information on the quality of programming
languages is scarce and usually not backed by hard data.
Regards,
Joachim
Return to the
comp.compilers page.
Search the
comp.compilers archives again.