home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!torn!cunews!revcan!software.mitel.com!murphyg!murphyg
- From: murphyg@Software.Mitel.COM (Gary Murphy)
- Newsgroups: comp.lang.forth
- Subject: Re: Free Forth
- Message-ID: <MURPHYG.92Aug13140514@murphyg.Software.Mitel.COM>
- Date: 13 Aug 92 18:05:14 GMT
- References: <3958.UUL1.3#5129@willett.pgh.pa.us>
- <BEVAN.92Aug6173810@otter.cs.man.ac.uk>
- <BEVAN.92Aug8212328@jaguar.cs.man.ac.uk>
- <1992Aug10.153601.26952@crd.ge.com>
- <BEVAN.92Aug11091421@jaguar.cs.man.ac.uk>
- Sender: murphyg@Software.Mitel.COM
- Organization: Applications Software II, MITEL Corporation, Kanata, Ontario,
- Canada
- Lines: 75
- In-reply-to: bevan@cs.man.ac.uk's message of 11 Aug 92 08:14:21 GMT
-
- In article <BEVAN.92Aug11091421@jaguar.cs.man.ac.uk> bevan@cs.man.ac.uk (Stephen J Bevan) writes:
-
- In article <1992Aug10.153601.26952@crd.ge.com> eaker@ukulele.NoSubdomain.NoDomain (Charles E Eaker) writes:
- I'm curious: can you comment on (1) why C++ was proposed instead of
- C?
-
- "Because C++ is `objected oriented' and so it must be better"
- Seriously that is the answer I got when I asked the same question.
- At this point I just gave up and walked away.
-
- Not at all. Speaking as a reluctant C programmer since the first
- versions came out, C++ is light-years ahead of C for large products.
- Most of what you do in C++ is just convenient notation for the uglier
- parts of C, like the 'hungarian notation' Microsoft is touting. C++ grew
- out of the user-feedback from C, and if it were truly portable to as
- many platforms as C, I might get more opportunity to use it.
-
- (2) how you got free C++ compilers for both UNIX and PCs (and
- maybe Atari STs)?
-
- Free FORTH, from what I've seen so far, should be compared to Small-C
- for the unix/pc side --- they give you the basic idea, but to write
- anything worthwhile, you will spend hundreds of times the cost of a real
- C in labour reinventing wheels. This was true until GNU, and the GNU
- option is one I would highly recommend for FORTH.
-
- I currently use the Pygmy FORTH, which is a good engine, but it is sadly
- only that, and for the above reason spends most of its time on a
- diskette. I learned all my other languages either by the seat of my
- pants (because the client insisted) or through studying sample programs
- in the public domain and stealing bits of them to fill out what I didn't
- have the skill or the time to implement. This I see as the real trouble
- in FORTH: I look up in the simtel archives, and there isn't a single
- truly meaty forth program available in source. I've ordered a few items
- through the FILE ON-LINE offers here, but again, either what I get does
- not run in pygmy, or it is too simplistic to bother with. I like FORTH
- and would love to get to know it better, but I have never had the time
- to learn by painstaking trial and error; I also cannot afford to spend
- hundreds to get a full-featured package for entertainment purposes only.
-
- A few years ago, I found most of my clients would not trust 'free'
- software because they thought paying for software guaranteed support.
- In the GNU project, the software (and source) is free, but you must pay
- for support, and the result has been a set of utilities that so surpase
- what the vendors sell that I have been able to move through three
- corporate contracts and still be encouraged to use GNU emacs, GNU C/C++,
- GNU make, even GNU smalltalk! When I finally upgrade my PC to a 386,
- you can bet my borland compiler will join the forths on the shelfs, as
- will my epsilon editor, leaving only my prolog compiler as unsurpased by
- free software (I use precious little MsDOS on that machine, having
- replaced and extended the system with GNUish MsDOS programs)
-
- A classic chicken and the egg problem if ever I saw one.
-
- Not really, I think it is irrelevant: the real need is to have an
- accessible archive of sample FORTH code to the same degree as we see for
- C, Pascal and assembler --- I can get btree database,
- dBase/Lotus-interfaces, windowing systems, menu systems, communications
- modules, data transfer protocols, TCP/IP drivers, in short, all of the
- specialized non-task parts that make a program useful. In C++, for the
- 386 and unix users, there is an even more amazing library of what FORTH
- would call 'words' available for free --- not the greatest, but then the
- C library shipped with most C compilers can contain a lot of junk, too.
- The point is, within hours of opening the box, you can demonstrate the
- language, and this has not been the case with the free FORTH. If the
- resource is there, most people, even students, might be willing to shell
- out $100 or so for a 'production-quality' engine --- this was certainly
- true for TurboPascal where Borland was the driving force behind
- releasing free sample code for their (then) $69 competitor to the
- stagnant $400 IBM Pascal.
-
-
- --
- Gary Lawrence Murphy - Gary.Murphy@software.mitel.COM - (613) 592-2122 x3709
- Real generosity toward the future lies in giving all to the present.
-