NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/alt/msdos/programm/3002

Internet Message Format | 1992-12-23 | 7.9 KB

Xref: sparky alt.msdos.programmer:3002 comp.os.msdos.programmer:11600 comp.sys.ibm.pc.programmer:668 comp.lang.c++:18384 comp.lang.c:18849
Newsgroups: alt.msdos.programmer,comp.os.msdos.programmer,comp.sys.ibm.pc.programmer,comp.lang.c++,comp.lang.c
Path: sparky!uunet!ferkel.ucsb.edu!taco!lea.csc.ncsu.edu!bank
From: [e-mail] (Belgarath the Sorcerer)
Subject: Re: Newbie Wants Advice on C-Programming
Message-ID: <[e-mail]>
Sender: [e-mail] (USENET News System)
Organization: /etc/organization
References: <[e-mail]>
Date: Thu, 24 Dec 1992 03:46:56 GMT
Lines: 158

In article <[e-mail]> [e-mail] (Jon Noring) writes:
>Hello,
>
>My wife, who is a database analyst working with DB2, etc. (all IBM mainframe
>stuff), is now interested in learning C programming because of pending changes
>at the company she works.  I, too, have an interest in learning C - I'd like to
>be able to write programs to solve a variety of problems (I've done a lot of
>fairly complex scientific programming with, heaven-forbid, Fortran, so I do
>have some general programming experience).
>
>Since both of us are newbies in this area, I'd like some advice as well as
>answers to the following questions:
>
>1) There seems to exist C, as well as C++, and possibly other variants.  What
>   are the differences between these?  Which should we learn first?
>
>2) We want to do most of our programming on our 386-33 PC running MSDOS 5.0.
>   What C-type compilers exist for the PC?  Which ones are better?  How much
>   do they cost and where can one get them at a discount?
>
>3) How robust is C compared to Fortran for purely number crunching, scientific
>   type applications on the PC?
>
>4) What books would you recommend in learning C (my wife might want to take
>   classes - I do better teaching myself)?  And of course, since I will be
>   doing programming on a PC, what books would you recommend that are
>   specific to that platform, especially w.r.t. memory management as well
>   as graphics?
>
>
>Much thanks for your help/advice.
>
>Jon Noring
>
>=============================================================================
>| Jon Noring          | [e-mail]        | I VOTED FOR PEROT IN '92 |

    Welcome to the 20th century of computing! :-) I suspect I know

which company your wife works for, but I won't go into that. :-)

    Your questions may qualify as FAQ - I dunno. But I'll be happy
to give some answers.

    1) Grammatically speaking, C++ is an extension of C. A C++ compiler
        will happily compile regular old C source code; naturally, doing
        so wastes all the extra power in C++

       C++ is an OOP language; that is, it was designed for Object
        Oriented Programming. All of its extensions are designed to
        let you do OOP.

       C, on the other hand, is a traditional, top-down programming
        language. It was designed for portability and the maximum
        in extensibility. You can judge for yourself how well it
        achieves those goals.

       Which should you learn first? Well, if you want to make the
        leap into OOP, start with C++.

       HOWEVER, I suspect that simply making the jump to C will be
        difficult enuf. You may not want to try to leap into OOP
        at the same time. OOP and traditional programming are two
        totally different mindsets. Unless you are schizo, its
        extremely difficult to switch back and forth. If you're
        not interested in getting into OOP, then I'd go with C.

   2) Your platform sounds good. On the PC, the major players in
       the C compiler market are Borland, Microsoft, and Zortech.
       There ARE others...these are just the ones that are
       usually regarded as 'major'. Whatever that means.

      I'm not going to come right out and say 'X compiler is better
       than Y compiler'...I'm not here to start a flamewar.

      Borland compilers tend to compile and link faster than
       the Microsoft ones. Microsofts optimizations tend to be
       better. Borland's floating point emulation code runs faster
       than Microsoft's, but Microsoft tends to be faster in the
       integer arithmetic arena. I have never used any Zortech
       compilers, but once bought some C-code libraries from
       them and was EXTREMELY dissatisfied.

     You will prolly pay on the order of several hundred dollars
      for a C compiler. I would look in Computer Shopper to get
      the best prices. I don't have one handy. Last time I looked
      at Microsoft C was Version 6.01a (it is up to Version 7
      now) and it was about $350 then.

  3) Unlike FORTRAN (and Pascal and some other languages), C have
      (excuse me, has) very little in the way of built-in
      number-crunching functions. Some compilers include math
      libraries that do a decent job...but these are extensions,
      not part of the C language itself, so the functions and
      their usages vary from compiler to compiler.

     You see (no pun intended), in order to give C the maximum in
      flexability, its designers (Kernigan and Ritchie) only put
      a minimum of I/O and math functions in the language. To
      do things like complex math and string ops and so forth you
      get libraries of routines. If the numeric-processing library
      that comes with your compiler doesn't pack the oompf you
      need, all you have to do is go get another, more powerful
      one and link it in and you can use it instead.

     This contrasts with languages like FORTRAN, where all this
      stuff is built-in. And if your particular implementation
      of FORTRAN has any shortcomings, too bad. You can't switch
      parts out. C lets you select the library of routines that
      best fits your needs.

     Also, C is MUCH more memory efficient than FORTRAN. But then
      FORTRAN is an old has-been. ANSI blew FORTRAN's chance to
      be a real language when the dropped pointers from its
      revision (altho some compiler-makers have wisely decided
      to ignore ANSI in this case and added pointers anyway).

  4) Hmmm. Part of this depends on what compiler you choose. I have
      done 80% of my work in C using Microsoft C (first Version 5.1,
      later 6.0). I found "The Waite Group's Microsoft C Bible"
      (2nd Ed.) to be a very good reference. ISBN 0-672-22736-3.
      They also put out another good book, called "MS-DOS Developer's
      Guide" which had lotsa stuff on memory management and graphics.
      ISBN 0-672-22630-8. As for learning how to program in C,
      a book titled "Variations in C" (2nd Ed.) seemed to be pretty
      decent. ISBN 1-55615-285-X. Two other books you might find
      useful are "DOS Programmer's Reference", ISBN 0-88022-458-4
      and "The Programmer's PC Sourcebook" (2nd Ed),
      ISBN 1-55615-321-X.

   A course at a local college or university is prolly a good place
to start off learning C. Make sure the course on C is a regular
part of their curriculum and not just taught occasionally by some
guy who usually teaches BASIC.
   Depending on how fast you pick things up, you can expect to
spend 6 months to a year programmin C (or C++, for that matter)
before you reach any real level of proficiency. I look back at the

code I wrote during the first 6 months I spent writing C and I
shudder in revulsion.
   If you have been a mainframe-type for many years, do not expect to
pick C up in a week. Pointers are likely to be a major stumbling
block - I doubt DB2 has anything like them. C is going to be a whole
new way of thinking - it is a language that assumes you know what you
are doing and gives you plenty of rope to do whatever you want,
including hanging yourself.

   Good luck.

Belgarath the Sorcerer
[e-mail]

DISCLAIMER: The above message constitutes an honest effort by the
             author to impart information he knows, or reasonably
             knows, to be true. All other interpretations are in
             error.