mwm@ea.UUCP (12/01/84)
> Should a properly limited government in a free society > include organizations equivalent to today's National > Bureau of Standards? No, it should not. Since NBS standards are *not* enforced by law, but by practice, there isn't any point in having the standards defined by a government agency (with people from concerned companies making the decisions :-). > However, this organization occupies a unique niche: > it DEFINES terms that affect the meaning of many laws > and contracts: seconds, meters, volts, and so on. No, it standardizes what was common practice, and chooses a (hopefully) happy medium when there is more than one definition for the term in use. When was the last time NBS (or any other standards organization, for that matter) changed the definition of something enough to matter in practice? > The difficulty with leaving the definition of weights and > measures to private industry is that here is one place > where I do not want multiple competing definitions. So what do you do when you buy 1000 tons of something, and get 2,000,000 lbs, instead of the 1,000,000 kgs you expected? > If there were multiple standards organizations, then every > time I signed a contract to buy N pounds of something, > I would have to specify, at least in principle, the standard > that defined 'pound' in that contract. But what if that > standards company goes out of business? What if they decide > to change their standard? An industry would generally have a "common definition" for the units they use. If you are going to be doing business in an industry, it behooves you to know what that definition is. There are lots of examples of this in CS. Lots of good examples of how not having a standard causes problems, too. However, as soon as there is *some* document - even if it isn't a standard yet - people start asking "how does it compare to XXX.YYY?" > It seems to me that since the decisions of a standards organization > effectively have force of law, these organizations cannot be > left uncontrolled. If they were run by industry (are you *sure* they aren't now?), they would probably behave much as they do now - a group decides they need a standard for something, lets others know, then they meet and work out a compromise that hurts the members of the committee least. [Classic case: there was one LAN standard (forgot which) wherein some of the members had broadcast systems in the field, whereas others had token passings systems in the field. Result: the first draft specified the *hardware*, but the software could be either of the above. Cute.] > This, anyway, is my current opinion. Anyone care to try > to talk me out of it? Yup. I doubt if I did, but I tried. However, to exemplify what I mean, suppose that someone bought 120 ft of X from me. I sell them 120 ft, but my ft are only 10 inches long (my inches are standard inches). Now, they are going to object, and probably sue me. The judge wouldn't bother trying to find a law that defined a "foot" (does such exist? If so, where?), but would tell me to give the guy his missing 20 ft of X. <mike
nrh@inmet.UUCP (12/07/84)
>***** inmet:net.politics / alice!ark / 10:30 am Nov 22, 1984 >Should a properly limited government in a free society >include organizations equivalent to today's National >Bureau of Standards? > >The arguments against it are obvious: it does nothing >that cannot, in principle, be done by private industry. > >However, this organization occupies a unique niche: >it DEFINES terms that affect the meaning of many laws >and contracts: seconds, meters, volts, and so on. > >The difficulty with leaving the definition of weights and >measures to private industry is that here is one place >where I do not want multiple competing definitions. > >If there were multiple standards organizations, then every >time I signed a contract to buy N pounds of something, >I would have to specify, at least in principle, the standard >that defined 'pound' in that contract. But what if that >standards company goes out of business? What if they decide >to change their standard? > >It seems to me that since the decisions of a standards organization >effectively have force of law, these organizations cannot be >left uncontrolled. Therefore they should be part of the >government, since this is preferable to having a government- >controlled organization that is otherwise private. > >This, anyway, is my current opinion. Anyone care to try >to talk me out of it? >---------- > Sure! Two important points: 1. Contracts are in general subject to arbitration in the event of re-definition of something like the "pound". Ordinarily, it is enough to agree on "The ANSI pound, as defined on Jan 1, 1970", if you're feeling paranoid. 2. There's no reason to think that the government will do any better than private industry in terms of keeping the "pound" standard. A recent example of this sort of fiddling was the re-definition of the cost-of-living index by the administration in (I think) 1981. A better example is the fact that it became illegal to ask whether one was to be paid in silver assignats or in paper assignats just before the French Revolution. That standards companies compete with each other doesn't mean that one would have conflicting definitions -- the organizations involved cooperate a good deal, for obvious reasons. ANSI, the American National Standards Institute, does a lot of computer standards work, and they're a private organization.