T.R | Title | User | Personal Name | Date | Lines |
---|
1445.1 | Waiting patiently | SMAUG::SPODARYK | Scaring the pedestrians... | Tue Sep 20 1988 12:37 | 18 |
|
I talked to Lattice this morning, and they told me that C++ will
basically be a C++ pre-processor packaged with their latest C compiler.
There *will* be an upgrade path for people like me, who are using
C v3.03. (Got it in 1.1 developers kit - all for $50, call me cheap.)
I've been using v3.03 for just a short while, to get my feet wet
in the Amiga waters.
I know it is easier for Lattice to produce a pre-processor, and then
run the code through their standard compiler, but how does this
affect efficiency? Hopefully not too much. I have seen PC C++
offered as a C++ compiler. Less time/work for the programmer, it
seems. Hopefully, it won't be *too* buggy.
Shipment is scheduled for late October, early November. I've seen
it advertised for some time now. Could anyone actually have it?
Steve
|
1445.2 | Too Bad It's a translator | CIMAMT::CROMACK | | Tue Sep 20 1988 17:36 | 8 |
|
I have found it a pain to use a C++ translator on VMS.
If you use a symbolic debugger then it probably won't do
you much good since that C source is probably unreadable.
Compilation takes longer and is more complicated.
Dean Cromack
|
1445.3 | Slow, but it's real C++ | TLE::RMEYERS | Randy Meyers | Tue Sep 20 1988 20:00 | 33 |
| The usual way of implementing C++ is as a "preprocessor" that reads the
C++ program and produces regular C which is then compiled by a compiler.
The Bell C++, which is still the standard for C++, is a preprocessor.
C++ compilers are still very rare: I believe that there is only one
or two shipping in the clone world, compared to hundreds of C compilers.
Most C++ systems are still the preprocessor, usually a port of the Bell
Labs C++. The disadvantages are as pointed out in .1: you lose source
level debugging and the compilation process is slower. The advantage
is that you get real C++ with the latest improvements to the language.
C++ is still a rapidly evolving language. It lacks a standard definition
(if you disallow the code of the preprocessor itself). There is a book
on the language from the language's inventor, but it is not comprehensive
enough to serve as a language definition to a compiler writer. The
inventor of the language is still working on improvements to the language.
Those improvements show up first in the Bell Labs C++ that he distributes.
So, C++ systems that are posts of his preprocessor are next in line
for updates.
You may have noticed I put quotes around "preprocessor" the first time
I used it (I dropped the quotes subsequently for readability). The
C++ "preprocessor" is really a compiler. It does extensive type checking
and transformations on the program far above what a simple preprocessor
could do. If the C++ translator compiles the program, the resulting output
file is guaranteed to compile successfully by the C compiler. There is
just a great temptation to call the C++ translator a preprocessor because
it "just" translates the C++ program into different ASCII text instead of
binary.
However, in C++ talk, a C++ compiler goes from C++ to binary where a C++
translator (or "preprocessor") goes from C++ to C.
|
1445.4 | is it worth the extra compile time etc..? | WJG::GUINEAU | Just a Window in Time | Wed Sep 21 1988 08:35 | 8 |
|
What is C++?
What new constructs are available?
John
|
1445.5 | Book on C++???? | ANT::GERBER | Robert Gerber | Wed Sep 21 1988 09:52 | 4 |
| Re: .3
I have not seen the book for C++ yet...could someone post some more
information please.
|
1445.6 | Some C++ references | BARDIC::RAVAN | | Wed Sep 21 1988 12:08 | 25 |
| There are two books I know of on C++. The only one I have at my desk
currently is the one written by the original designer of the language.
The C++ Programming Language
Bjarne Stroustrup, AT&T Bell Laboratories
Addison-Wesley Publishing Company
Reprinted with corrections July, 1987
This is not a very good introduction to the language, since it was
written by someone obviously VERY close to his subject matter. There
is a better text available, but I'll have to hunt for the reference.
There are also the proceedings of the USENIX C++ workshop:
USENIX Proceedings and Additional Papers, C++ Workshop
Sante Fe, NM, 1987
$20, (+$25 air postage or $5 surface postage for overseas)
write:
USENIX Association
P.O. Box 2299
Berkeley, CA 94710
There is also a C++ notesfile, CADSYS::C_PLUS_PLUS.
|
1445.7 | Other references | SMAUG::SPODARYK | Scaring the pedestrians... | Wed Sep 21 1988 13:30 | 14 |
| I have a book from the DEC Library that is entitled "An Introduction
to Object Programming and C++". It seems pretty good but haven't really
looked at it too much because I don't have C++, or time. (yet)
Check out the Library under subjects "Object Oriented" and "C++".
There is a series of articles in August's Byte on C. Basically
How it started, How it has evolved (C++), and How to implement
OOP using standard C. I know its a PC mag, but grin and bear it. (^:
Of course the C_plus_plus notesfile has information, pointers also.
It seems like C++ would fit in very nicely with the ObjOriented Amiga.
--Steve
|
1445.8 | | SMAUG::SPODARYK | Scaring the pedestrians... | Wed Sep 21 1988 13:31 | 2 |
| That title should be "An Introduction to Object Oriented Programming
and C++" - but you probably knew what I meant. (^:
|
1445.9 | That's the title... | BARDIC::RAVAN | | Wed Sep 21 1988 14:49 | 7 |
| > That title should be "An Introduction to Object Oriented Programming
> and C++" - but you probably knew what I meant. (^:
This is the title of the book that is supposed to be much better
than the Stroustroup book. I have read this book cover to cover.
I recommend it. I tried to read the Stroustroup book, but stopped
after a couple of chapters.
|