info-cpm (12/04/82)
>From RMS.G.HNIJ.MIT-OZ@Mit-Mc Fri Dec 3 23:42:45 1982
To: info-cpm@BRL
Via: Mit-Mc; 4 Dec 82 0:56-EST
Via: Brl; 4 Dec 82 1:05-EST
Via: Brl-Bmd; 4 Dec 82 1:44-EST
Has anyone had any experience with the Aztec C compiler for either
CP/M or HDOS? If so, what's your opinion on:
* Speed. We need to use it for communication programs, so we
need it FAST! How does it compare with BDS C?
* M80/L80 compatibility. We have some assembly code that's
written for M80/L80.
* Library. Is the library full C (such as the file I/O).
* Compiling time. Is it as slow as Whitesmiths?
Thanx very much.
John Labovitz
(RMS.G.HNIJ@MIT-OZ)
-------info-cpm (12/05/82)
>From dag@UCBARPA.BERKELEY.ARPA Sun Dec 5 03:31:04 1982
Received: from UCBARPA.BERKELEY.ARPA by UCBVAX.BERKELEY.ARPA (3.227 [10/22/82])
id A02504; 5-Dec-82 01:42:24-PST (Sun)
Received: by UCBVAX.BERKELEY.ARPA (3.227 [10/22/82])
id A04017; 5-Dec-82 03:31:53-PST (Sun)
To: RMS.G.HNIJ.MIT-OZ@Mit-Mc, info-cpm@BRL
Via: Ucb-C70; 5 Dec 82 4:43-EST
Via: Brl; 5 Dec 82 4:46-EST
Via: Brl-Bmd; 5 Dec 82 5:27-EST
Aztec C is a good overall C Compiler, though BDS C is better
for those who like all sorts of neat goodies like a symbolic debugger
and a package that is easier to use. Aztec is slightly faster in execution
to BDS C and can work with M80/L80 stuff. It compiles MUCH faster
than Whitesmiths (A monkey could hand-compile code faster than
whitesmiths.) For further time information on both compilation
and interpretation, see my articles in November/December 1981
and Jan/Feb 1982 for reviews of BDS C and Whitesmiths, and Nov/Dec 1982
for reviews of Aztec and C/80.
Overall, I prefer to program in BDS, but if I needed full C usage
including register variables, I would use Aztec without a second
thought.
Davidinfo-cpm (12/05/82)
>From dag@UCBARPA.BERKELEY.ARPA Sun Dec 5 03:45:57 1982
Received: from UCBARPA.BERKELEY.ARPA by UCBVAX.BERKELEY.ARPA (3.227 [10/22/82])
id A02519; 5-Dec-82 01:43:22-PST (Sun)
Received: by UCBVAX.BERKELEY.ARPA (3.227 [10/22/82])
id A04177; 5-Dec-82 03:47:39-PST (Sun)
To: RMS.G.HNIJ.MIT-OZ@Mit-Mc, info-cpm@BRL
Via: Ucb-C70; 5 Dec 82 4:45-EST
Via: Brl; 5 Dec 82 4:58-EST
Via: Brl-Bmd; 5 Dec 82 5:28-EST
Oops..those articles are in Microsystems Magazine.
Sorry for the omission.
Davidinfo-cpm (12/07/82)
>From teklabs!tekcrd!tekmdp!kennethw@Ucb-C70 Mon Dec 6 21:40:47 1982
Received: by UCBVAX.BERKELEY.ARPA (3.227 [10/22/82])
id A00698; 6-Dec-82 21:46:30-PST (Mon)
To: tekcrd!teklabs!ucbcad!ucbvax.info-cpm@Ucb-C70
In-Reply-To: Your news article ucbvax.208 of Sat Dec 4 01:22:24 1982
Via: Mit-Ai; 6 Dec 82 17:22-EST
Via: Brl; 6 Dec 82 17:39-EST
Via: Brl-Bmd; 6 Dec 82 17:45-EST
I have the AZTEC C compiler for the HDOS system. The
compiler itself seems to be quite good and generates fast code.
My benchmarks show that it is faster than BDS in running the BYTE
prime number benchmark. The library is quite complete and does
support all the normal file I/O features. The output source code
of the compiler can also be assembled using an M80 compiler if desired.
Now for the bad news! The support of this product stinks.
There are several small problems in the library (you can't write to the
line printer, for example) and I have been totally unable to get
any help out of MANX software on the problems. I am also unable to fix
the problems since there library program does not work.
I have called and written to MANX several times and have received
either no answer or a very nasty one. They just are not willing to support
there compiler, which is too bad because it is basically a good product.
I would like to hear any other comments on the AZTEC compiler myself,
especially if someone has managed to get the compiler working.
Please post any useful information to the net. Thanks very much.
Ken West
TEKTRONIX
Box 4600
Beaverton Oregon 97075
(503) 629-1533