ada vs C benchmarks

ada vs C benchmarks

[John Sissler]

hi,

	does anyone have either the data or ability to produce 
meaningful benchmarks evaluating ada vs C.  i think such a
comparison would be of vast interest.  specifically, we are interested
in VADS vs pcc on a VAX.

				thanks,

				john sissler - HHB-Softron

		{decvax|allegra|ihpn4}!philabs!hhb!anwar!jon

ps. much thanks to all those who provided me with ada compiler
    info.

Re: ada vs C benchmarks

[Eugene D. Brooks III]

> in VADS vs pcc on a VAX.

I would suggest VADS vs tcc (Tartan's C compiler) on a VAX.

Re: ada vs C benchmarks

[Jan Steinman]

In article <175@anwar.UUCP> jon@anwar.UUCP (John Sissler) writes:
>	does anyone have either the data or ability to produce 
>meaningful benchmarks evaluating ada vs C.

To be fair, turn those checks off when benchmarking Ada against anything,
except "safe" dialects of Pascal, etc.  When comparing a Porche to a Sherman
Tank, are you looking for the fastest and lightest, or the most bomb-proof!
The term "meaningful" as used above, will obviously be weighted by what you
will be measuring.

Ada was designed to be used for applications where exceptions are intolerable.
(You don't want to launch a nuclear weapon because some array overflowed!)
The constraint and allocation checking will cause Ada to measure slower and
larger than languages like C that let you do as you damn well please.
-- 
:::::: Jan Steinman		Box 1000, MS 61-161	(w)503/685-2843 ::::::
:::::: tektronix!tekecs!jans	Wilsonville, OR 97070	(h)503/657-7703 ::::::

Re: ada vs C benchmarks

[Larry West]

In article <781@mako.UUCP> jans@mako.UUCP (Jan Steinman) writes:
>In article <175@anwar.UUCP> jon@anwar.UUCP (John Sissler) writes:
>>	does anyone have either the data or ability to produce 
>>meaningful benchmarks evaluating ada vs C.
>
>To be fair, turn those checks off when benchmarking Ada against anything,
>except "safe" dialects of Pascal, etc.  When comparing a Porche to a Sherman
>Tank, are you looking for the fastest and lightest, or the most bomb-proof!
>The term "meaningful" as used above, will obviously be weighted by what you
>will be measuring.

And yet simple tests can be meaningful if one keeps such issues in
mind.   For example, if you were designing a screen editor, one might
well want the best security possible (don't want to lose anyone's
data)... until one discover that this comes at the cost of making
even the simplest commands take seconds to complete.    Or that the
running code won't fit within an externally-imposed memory limit.

In short, direct comparisons of single aspects like execution speed
or code-size or "security" are not likely to be useful in choosing
a language/system.   But you may be able to establish that one
language/system or another simply cannot meet one of your basic
requirements, no matter how great its advantages in other aspects.
-- 

Larry West			Institute for Cognitive Science
(USA+619-)452-6220		UC San Diego (mailcode C-015) [x6220]
ARPA: <west@nprdc.ARPA>		La Jolla, CA  92093  U.S.A.
UUCP: {ucbvax,sdcrdcf,decvax,ihnp4}!sdcsvax!sdcsla!west OR ulysses!sdcsla!west