Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!ames!uhccux!uhunix1.uhcc.Hawaii.Edu!todd From: todd@uhunix1.uhcc.Hawaii.Edu (Todd Ogasawara) Newsgroups: comp.lang.prolog Subject: Re: Arity's so-called upgrade Message-ID: <11987@uhccux.uhcc.Hawaii.Edu> Date: 16 Mar 91 19:25:51 GMT References: <11768@uhccux.uhcc.Hawaii.Edu> <11954@uhccux.uhcc.Hawaii.Edu> <1991Mar15.050522.21309@athena.cs.uga.edu> Sender: news@uhccux.uhcc.Hawaii.Edu Organization: University of Hawaii Lines: 50 In article <1991Mar15.050522.21309@athena.cs.uga.edu> mcovingt@athena.cs.uga.edu (Michael A. Covington) writes: >Of course you have to remember that Borland has a truly *gigantic* >corps of beta-testers. Most companies would be hard put to assemble >such a group. In fact I'll bet Arity is constantly up against the problem >that they can't get enough people to test the product, thoroughly enough, >during the beta phase. I want to make sure that people who may have not read my previous posts understand that nowhere in my posting that started this discussion thread did I imply that Arity Prolog 6.0 was buggy. In fact, I feel that Arity's implementation of Edinburgh style Prolog is as stable as anything else I've seen. My complaints are based on two major points. 1. Arity has not provided new documentation since 5.0 in 1988. Versions 5.1 and 6.0 have made a number of confusing changes and additions to some of their proprietary extensions. There is no source of information besides Arity to go to to clarify these predicate extensions (i.e., you're not going to learn how to use these extensions by reading a Clocksin & Mellish or O'Keefe book). In addition the 5.0 manual was incredibly error ridden. It is not an overstatement to say that there is at least one technical error or on every other page. And, more likely, you are likely to find several errors of a single page that has code samples. 2. Arity claimed in their upgrade notice and implied in their version number shift (from 5.1 to 6.0) that this was a significant upgrade and a major overhaul of the system which has numerous shortcomings in providing a modern programming environments (when compared to Borland C++, Quick C, Actor, etc.) for the MS-DOS platform. They also charged (by PC standards) a large upgrade fee (~$120) for what is essentially a maintenance release. As to having a large number of beta-testers for QA. I don't think that QA in terms of operational bugs is the issue here. The issue, in my mind at least, is that Arity is out of step with current trends in programming environments. Their environment is as out of step with today's standards as a line editor was to full screen editors a few years ago. >But I agree, Borland's reliability and support are exemplary. Agreed.. BTW.. I don't claim that Borland C++ 2.0 is bug free. I just haven't had it long enough to make such a statement. I will say one thing though. I was pleasantly surprised to find that the program examples in Petzold's "Programming for Windows" which was written for use with Microsoft C and SDK work without modifications when compiled using Borland C++. Arity's examples, on the other hand, won't even compile in their own compiler without modifications. -- Todd Ogasawara, U. of Hawaii UUCP: {uunet,ucbvax,dcdwest}!ucsd!nosc!uhunix!todd BITNET: todd@uhunix INTERNET: todd@uhunix.UHCC.HAWAII.EDU