🚀 go-pugleaf

In article <1992Apr6.203655.5105@webo.dg.com> holmes@mrx.webo.dg.com writes:
>I'm probably about to open a great big can-O-worms and generate a 
>heated discussion, but which produces more efficient (size or speed, I'm
>looking
>for info on both) code, Microsoft's or Borland's C compilers??  I'm all for
>Borland simply because of the better user interface.  Anyone care to comment?
>-- 
>-----------------------------+------------------------------------------------
>Chris Holmes                 | Deliberate misuse by concentrating and inhaling
>Terminals Group	             | fumes may be harmful or fatal. 
>Data General Corporation     |
>Westboro, MA                 | Internet: holmes@mrx.webo.dg.com                 

It generally depends on how you write your code. IF you hand
optimize your C code, you can get either compiler to generate ASM
code that is just as fast as you having written it your self. BUt,
if you want to take the position of some "standard" tests, then
Microsoft gets the nod for speed, but not by much. In reality I
have found the Borland Compiler to produce code that acts more
stable in the environment than Microsoft, but there are things I
will go to Microsoft code for that it tends to do better, such as
segment placements and so forth.

Borland and Microsoft both produce good products, each with their
own drawbacks. I prefer Borland, until something better comes
along. One other thing to consider: Technical SUpport is a big
issue with compilers, and Borland wins that game hands down. They
have always been able to answer my questions and if they did not,
they duplicated the problem and called me with a workaround or
shipped me a bug fix for whatever I found. That I my book rates
higher than faster code. With machines getting quicker by the day,
wasted cycles are not as critical as they used to be.