From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on ip-172-31-74-118.ec2.internal X-Spam-Level: X-Spam-Status: No, score=0.8 required=3.0 tests=BAYES_50 autolearn=ham autolearn_force=no version=3.4.5-pre1 Date: 3 Feb 93 05:53:05 GMT From: oucsboss!att-out!cbnewsl!willett@sun.com (david.c.willett) Subject: Re: Comments on Ada vs. C++ panel Message-ID: <1993Feb3.055305.16347@cbnewsl.cb.att.com> List-Id: >>From article <1993Feb2.180552.5536@mksol.dseg.ti.com>, by mccall@mksol.dseg.ti .com (fred j mccall 575-3539): Original C++ discussion deleted..... > explain to me why a language restriction is better than, say, an > organizational policy restriction. Personally, I would consider a > language that allows me to do things in special cases to be better > than one which does not, even if the 'things' are not necessarily a > good idea in the usual case. > > [Let's try this one in Email -- I'm *NOT* interested in starting some > kind of language flamewar. I'm simply honestly curious about the idea > that it is somehow 'better' for a language to restrict the developer > rather than for the developer to restrict himself or herself to good > practice while having the ability to step outside that for unusual > cases.] > > -- > "Insisting on perfect safety is for people who don't have the balls to live > in the real world." -- Mary Shafer, NASA Ames Dryden > ----------------------------------------------------------------------------- - Fred, I'll take on your question, don't worry no language flamewar from here. To me the question hinges on whether a language should "assist" the developer or not. It is much the same question as GUI vs. command interpreter interfaces or automatic vs. manual transmissions. Generally, I'd prefer not to have the assistance, so I eschew GUI interfaces, spreadsheets, and automatic transmissions. However, there are situations where I find such things useful. If I'm doing a lot of "application hopping" being able to run 3..5 things in different windows without having to type complicated command scripts is nice. Spreadsheets have their purpose in accounting work, but I generally prefer an "old fashioned" ledger form. Automatics are better for city driving, which is the kind I'm doing most of the time now. How does all this relate to languages? What is "better" depends on the kind of work you do. If I have a one-to-four person-month project which is isolated or well insulated from the rest of the world, I don't care which language is used. Maintenance might become an issue but probably not for a while. At the other extreme is the 24-48 person-month effort which is central to a sophisticated system of software. Here is where I'd want to make very sure that the techniques used were consistent and rigorous. You suggest that ogranizational standards (may I infer software management? ) could enforce such techniques so the language doesn't have to. I submit that compilers are better "enforcers" than people. They are consistent. They are equitable. They do not crack under cost or schedule pressures. The bottom line is that they are generally better at the job. -- Dave Willett AT&T Federal Systems Advanced Technologies attmail!dwillett (AT&T FSAT) If you think it's so !#$%^& easy, You try it!