* Re: A Simple struct vs a convoluted class object.
[not found] ` <35B48E17.12228895@oma.com>
@ 1998-07-21 0:00 ` John Goodsen
0 siblings, 0 replies; only message in thread
From: John Goodsen @ 1998-07-21 0:00 UTC (permalink / raw)
Tim Ottinger wrote in message <35B48E17.12228895@oma.com>...
>> You have me at an advantage here, are you suggesting that there exist
>> today optimizing compilers for Smalltalk or Eiffel or Java that outdo
>> C++ ? My comment was meant to indicate that while one may reasonably
>> make many critisisms of C++ I had just never heard anyone claim it was
>> slow ...
>
>I've heard it, but I've not heard it justly applied. Usually when C++ isnot
fast
>enough, it's because it was written inefficiently or (more often)
>it's a complaint against new/delete as provided by the compiler vendor.
>Admittedly, copying huge objects is slow (that's why we don't do it),
>and the malloc implementation of new is typically slow. One issue is
>a quality of code issue, another a quality of implementation issue.
>
There were some studies done a few years ago back on Ada compiler
optimization and it showed that Ada compilers have the ability to
optimize *MUCH* better than a C compiler, because of the strong
language definition. I cross-posted this to comp.lang.ada, in hopes
that somebody from that zone might respond...
I suspect that Eiffel and Java have the same potential. I'm not sure
about Smalltalk, with it's dynamic typing and all ... what does it mean
to "compile" smalltalk - I can't imagine the dynamic typing optimizing
faster than a statically typed language...
--
John Goodsen Saguaro Software, Inc. / Training & Consulting in:
jgoodsen@saguarosoft.com - User Centered Systems Analysis
http://www.saguarosoft.com - Object Oriented Architecture and Design
602.283.0142 / 888.298.2566 - Rapid Incremental Delivery S/W Process
"Example isn't another way to teach, it is the only way to teach. "
- Albert Einstein
^ permalink raw reply [flat|nested] only message in thread