Path: news.daimi.aau.dk!news.net.uni-c.dk!uninett.no!news.maxwell.syr.edu!feed1.news.rcn.net!rcn!master.news.rcn.net!not-for-mail From: ell@access.digex.net (Ell) Newsgroups: comp.lang.beta Subject: Real OO vs Craftism Date: Fri, 25 Sep 1998 08:53:00 GMT Organization: Universe Lines: 166 Message-ID: <360b59bf.204874@news.erols.com> Reply-To: ell@access.digex.net NNTP-Posting-Host: 207-172-52-83.s83.tnt1.brd.erols.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: winter.news.erols.com 906713737 3786 207.172.52.83 (25 Sep 1998 08:55:37 GMT) X-Complaints-To: abuse@rcn.com X-Newsreader: Forte Agent 1.5/32.452 Xref: news.daimi.aau.dk comp.lang.beta:11715 "Pete McBreen" wrote: >kderrick@my-dejanews.com wrote: >> >> But indeed, the craftsman/guild approach survived the industrial >>revolution in >> the form of industrial apprenticeships. It did so for the simple reason >>that >> *it works*. >> >> This approach recognises and passes on not just the technical skills >>required >> for a trade, but the nuances and advanced thinking which the master has >> acquired over the years. >Agreed! > >Craftsmanship is what we should all strive for. The old masters used to >sign their work for a reason - they were proud of what they had created and >wanted to let posterity know that this was their highest espression of >their craft and themselves. > >The Craftsmanship and apprentice system is necessary to instill the needed >attitudes that result in excellence. Yes, you can always hire assembly >workers to make mass produced items of hopefully defined quality, but for >real excellence you need craftsmanship. We don't always need the quality produced by modern day craftspeople, and artists. Or if we do, we only need their output in very small quantities for limited situations. For the most part, we get the quality we need at much more affordable prices by using mass production. >Remember that continuous improvement is a Craft mindset, not a repeatable >assembly mindset. I beg to differ. Certainly that would be news to Deming. the Japanese and those in competitive fields generally. Improvement is a life or death matter in competitive mass production fields. >And while I am at it, 'Gwoin with your bad self! :-} >a Pragmatic, Empirical mindset is also very useful in >software development. Pragmatism is needed to ensure that we can ship the >system on time, we know we will improve it later, but the first release is >important. Any system that is successful will be evolved over the years to >keep pace with it's environment, and it will always be sub-optimal in some >dimension, there will always be parts we want to rework. But systems are >not works of art, Well then why are you arguing above that only craftspeople should do the work and not mass production *for the most part*? But let's move on. >they are working machines, and the test of a machine is >whether it is more productive to use the machine than not use it. Not sure of your point in that paragraph. But I'll try as best I can to respond. While the craft process produces good products, the mass production process also produces good products for the proper circumstances. And the determination as to whether to use one or another should be based upon an objective assessment of the quality required and the prices involved. So there are 2 issues involved here: 1) The nature of a bad frequently arising mindset from the craft process and craftpeople 2) The choice involved in any given situation between whether it makes sense to go with the output of either the craft, or mass production process. One thing I want to clarify about my position, as I did in another article in this thread, is that I'm not totally against the craft training process and I recognize that product quality is typically better from the craft process than from the mass production process. While having the mass production alternative is significant and a key to having today's and tomorrow's efficient and productive society, that is not generally the point of what I'm getting at when I state that a position, or stance is craftite. What I'm criticizing is the narrow, insular, anti-real world modelling, anti-holistic, anti-objective, often anti-scientific, and regressive mindset/mentality that often flows from those with a craftite outlook. The fundamental underlying aspects of that mentality, and the ones from which a number of other horrible, ancillary ones stem from are: 1) a bad *Pragmatic* stance - "design what works and domain modelling doesn't matter". 2) a bad *Empiricism* stance - "reality doesn't exist, or we can not really know reality using science" >Empiricism is also needed since ours is the people skills business, we >cannot do repeatable scientific experiments to discover the "one true way" >to do something. That's totally untrue and stems from # 2) above. Every domain has objective aspects, facts and laws which using the scientific method, we are generally able to know and understand. While a domain may have many ways to approach and solve a problem, it is usually possible to scientifically find those "approaches" and "solutions". The scientific method involves doing development work, summarizing that work and proposing changes in the way we work when necessary, testing those proposed changes and then eventually settling on new and better ways to do development work. We then continue to do development work and propose changes when necessary in a never ending helix. The iterative and incremental software development (IID) process that I proposed in '89, in comp.object, comp.lang.c++, and elsewhere on Usenet, and which was widely adopted in software engineering (mostly without proper accreditation) is based on the same scientific method. >If we have success with a technique, and more success than >we have had with any other technique, then we will adopt it until we find >something with a better promise of success. But this ignores that we should be systematically trying to apply the scientific method and that we should be basing conclusions not just upon the experience of oneself, or the groups one knows, but upon the experience of all, or a statistically large enuff number, of those having experience in an area. >Yes it is not "Scientific", but it works for us, Ahh, the sad siren of pragmatism-"it works for us". But for how long? To what extent? Is it most efficient? Is it fair? Etc and so on. I can not emphasize enuff that only if we base ourselves firmly upon the scientific method can we most quickly and accurately answer such questions. >and our clients are willing to pay for us to repeat this >unscientific behavior for their benefit. So while being less than completely scientific doesn't get us too far, we should pragmatically rip clients off, because they allow us to? Not me, not many others, not for a million years. The scientific method, pre-code holistic modelling and the recognition of contextual objective facts are 3 key beautiful means to achieve both untold software engineering advances, and a world free from the "Reaganesque twinkle" inspired evils of "ketchup is a vegetable" and the murder of nuns in Guatemala. [ Please. please, please, Vote NO, to oppose the formation of a moderated news group whose lifetime moderators overwhelmingly oppose riding the 3 great vehicles - the scientific method, pre-code holistic modelling, and the recognition of contextual objective facts - into a bright software engineering future. ] Elliott :=***=: VOTE NO TO MODERATION! ALL IDEAS SHOULD BE CRITICIZABLE! :=***=: MODERATORS SHOULD NOT HAVE LIFETIME TERMS! :=***=: Objective * Pre-code Modelling * Holistic :=***=: Hallmarks of the best SW Engineering Check out SW Modeller vs SW Craftite Central : www.access.digex.net/~ell Copyright 1998 Elliott. exclusive of others' writing. may be copied without permission only in the comp.* usenet and bitnet groups.