home *** CD-ROM | disk | FTP | other *** search
- Path: yama.mcc.ac.uk!dmu!usenet
- From: Graham Perkins <grp@dmu.ac.uk>
- Newsgroups: comp.lang.c++,comp.lang.eiffel,comp.lang.c,comp.object,comp.software-eng
- Subject: Portability of code & skills (Beware of "C" Hackers etc)
- Date: Fri, 15 Mar 1996 10:57:45 +0000
- Organization: De Montfort University
- Message-ID: <31494D29.4D4B@dmu.ac.uk>
- References: <4i9lbm$pmh@news.interpath.net>
- NNTP-Posting-Host: mkcsst22.mk.dmu.ac.uk
- Mime-Version: 1.0
- Content-Type: text/plain; charset=us-ascii
- Content-Transfer-Encoding: 7bit
- X-Mailer: Mozilla 2.0 (Win16; I)
-
- Too much stuff on the endless looping of "Beware of "C" Hackers..."
- thread so I'll try and split it. Or should we "fray" a thread?
-
- Scott McMahan - Softbase Systems wrote:
- > What people don't realize about C is that it's a self-preservation
- > phenomenon. Moving from computer to computer, you can bring your C
- > skills and have a lot less to learn when you sit down in front of a new
- > box, a new OS, etc.
-
- Too true, too true. So if you want your new language to be a success,
- then give it away for ten years, preferably with a free O.S. and all source
- code as well. It's difficult to see many other reasons for widespread
- adoption of C and Unix.
-
- However, if C really is as portable as some seem to be saying, then it
- can be used as a universal assembler language enabling other systems to
- be portable. This does not seem to have happened, and I don't understand
- why. Most of the CASE tools which generate C or C++ seem to be working
- as "programmers' assistants" for C/C++ programmers. IMHO they should
- present one or more high level views with their own direct manipulation
- models, dynamically altering wadges of C code behind the scenes. The
- raw C code should be just as invisible as the byte-files and btree indexes
- managed by an RDBMS.
-
- > Writing COBOL on a mainframe, Delphi in Windows, REXX in OS/2,
- > and perl in UNIX isn't as good as writing C everywhere.
-
- I suspect you'll find Cobol available more widely than C, especially if
- you only count internationally standardised and benchmarked compilers.
-
- If I need to use an indexed file, with a primary index of unique composite
- keys and two secondary indexes, one of which allows duplicates, and I want
- the program to work on IBM, ICL, DEC mainframes and all DOS, Windows, Unix,
- OS/2 boxes without editing one line of code or writing a make file, then
- Cobol might be a better choice than C. And there are quite a few other
- "courses" for which Cobol is the appropriate "horse".
-
- I've got a sneaking suspicion that a major market pressure for OODBMS
- is the demand from C programmers who eventually realise that data storage
- is more than text files and bytes streams, but refuse to use Cobol ;-)
-
- > C isn't perfect -- but it can be
- > implemented easily on new platforms.
-
- An unfortunate aspect of market pressure which ensures software development
- technology will always operate at a chronically low level.
-
- I'm sorry to report that acadaemia is also following this route of sacrificing
- power and quality for cheapness. Unfortunately we are never in a position
- to go back to our managers and say "Look here, this cheap compiler we got is
- just a compiler that translates my source code into bits. If you want me
- to develop good source code then we need to invest in a good development kit".
- [departmental head: "sorry, no capital budget .. staff budget also ]
- [ cut so if you can't teach with limited resources then we'll sack ]
- [ you and get someone else who will do it on a lower salary ]
-
- > C supports
- > just about everything Modula-2 does
-
- Oh come on now! You may have a case, in which case don't fiddle the evidence!
-
- --
- person: Graham Perkins paper: School of Computing
- voice: +44 (0)1908 834936 De Montfort University
- dots: +44 (0)1908 834948 Milton Keynes MK7 6HP
- bits: grp@dmu.ac.uk United Kingdom
-