$99 C++ compiler

[Wolf N. Paul]

In article <8805051248.AA20720 at decwrl.dec.com> nadkarni at erlang.dec.com (Ashok P. Nadkarni DTN 226-7448) writes:
>
>	According to PC Week, Zortech has announced a C++ compiler
>	for $99. Note this is a compiler, not a preprocessor like
>	Guidelines and some others that require MS C to do the actual
>	compilation. Anyone ever heard of this company ? The price
>	sounds REAL GOOD to me. Maybe they're planning to do with C++
>	what Borland did with Turbo-Pascal.

The Zortech C++ compiler is a descendant of the Datalight C Compiler.
Zortech used to market Datalight C in Britain, and have now acquired
marketing rights for the US as well.

I understand that they are rewriting the documentation, and are a bit
behind in the actual release of their version.

Zortech C++ is a true compiler, not a pre-processor, and is written to
the specs in the C++ book.  I will also be ANSI C compatible, and the
object files are MSC compatible (unlike the old Datalight object files
which were Lattice compatible).

>	Also, how about POWER C from MIX software ? $19.95 for the compiler
>	$19.95 for the debugger, and, get this, $10 for the library (450
>	functions) source. Again sounds too good to be true. They claim
>	to beat Turbo-C and MS C in the standard benchmarks.

I bought Power-C because it claims Turbo-C source code compatibility
and was too cheap to pass up.  It produces medium-model code (i.e. 64K
code and unlimited data); it uses non-standard object and library file
formats.  There is a utility for converting MS object files into MIX
object files; however there is no utility for going the other
direction.

A recent posting seemed to imply that Power-C was simply a new version
of the old MIX compiler; I don't think that's true -- it is a
completely re-written product.  Note that MIX still sells the old
compiler as well, for the same price.

There are a number of errors in the manual (i.e. it seems they planned
to support multiple memory models and a conversion utility to convert
MIX files into OBJ files and then changed their minds), and a few bugs
in the library (i.e. the functions in the ctime(3) group are confused
about whether tm.tm_year should contain the year, or the year-1900).

If you buy the library source for $10, you not only get that, but also
a library manager (MERGE) and an assembler (ASM), not compatible with
MASM.

Sorry I am not very concerned with benchmarks of either speed or .exe size,
so I cannot comment on those.

Hope this is helpful anyway.
-- 
Wolf N. Paul * 3387 Sam Rayburn Run * Carrollton TX 75007 * (214) 306-9101
UUCP:     ihnp4!killer!dcs!wnp                 ESL: 62832882
INTERNET: wnp at DESEES.DAS.NET or wnp at dcs.UUCP   TLX: 910-280-0585 EES PLANO UD