Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!rice!uw-beaver!uw-entropy!dataio!bright
From: bright@Data-IO.COM (Walter Bright)
Newsgroups: comp.std.c
Subject: Re: Vendor representatives on committee
Message-ID: <2222@dataio.Data-IO.COM>
Date: 22 Nov 89 19:41:41 GMT
References: <11134@riks.csl.sony.co.jp> <15217@haddock.ima.isc.com> <1643@crdos1.crd.ge.COM> <1989Nov17.205004.19236@cs.rochester.edu> <1653@crdos1.crd.ge.COM> <11645@smoke.BRL.MIL> <1989Nov20.124013.28617@algor2.algorists.com> <1662@crdos1.crd.ge.COM>
Reply-To: bright@dataio.Data-IO.COM (Walter Bright)
Organization: Data I/O Corporation; Redmond, WA
Lines: 25

In article <1989Nov20.124013.28617@algor2.algorists.com> jeffrey@algor2.ALGORISTS.COM (Jeffrey Kegler) writes:
| A number of postings has suggested that there is something unethical
| or immoral about advocating the particular interests of a specific
| vendor on the ANSI C committee, as opposed to the general public
| interest.

I've also seen a lot of opinion that pushes the idea that compiler
implementation issues are irrelevent, the only thing that counts is
programmer issues. To present a counter argument:

If implementing a language is excessively complex:
1. Fewer vendors will get into the business.
2. There will be long delays before compilers get implemented.
3. Fewer implementations directly translates into lower quality implementations,
   because of less competition.
4. Fewer implementations mean higher compiler prices.
5. Compile times will be slow.

The programmer's interests are not well served by that scenario. I, for
one, would prefer a high-quality implementation of a simple language than
a poor-quality version of a complex one.

For an example of the above 5 points, see Ada.

P.S. I'm a compiler vendor (Zortech).