Re: MV Keys

From: Jon Heggland <heggland_at_idi.ntnu.no>
Date: Fri, 3 Mar 2006 09:15:09 +0100
Message-ID: <MPG.1e721958c1f3b02c98977f_at_news.ntnu.no>


In article <MPG.1e720e3f9f7c590398977e_at_news.ntnu.no>, heggland_at_idi.ntnu.no says...
> In article <1141372781.078243.197630_at_e56g2000cwe.googlegroups.com>,
> marshall.spight_at_gmail.com says...
> > Jon Heggland wrote:
> > >
> > > As long as the type of a bunch of values is well-defined, there is
> > > indeed no problem---indeed, that makes the two alternatives I outlined
> > > equivalent. However, if any attribute is a set (or list), you probably
> > > want (for convenience) to treat it as a single value of its constituent
> > > type when it contains only one value, I.e. "Jon" = { "Jon" }. A language
> > > that does this may exhibit undesirable behaviour in other places; I
> > > believe SQL has trouble with table literals because of this.
> >
> > Can you say more about this? What kind of trouble?
>
> I think Hugh Darwen has written a piece on this. I'll try to dig it up.

http://www.dbdebunk.com/page/page/2928212.htm discusses this, among other things (row types and user-defined types are also mentioned, in connection with the main point: NULLs). It's quite amusing, in a scary (for SQL users) way. It's mostly Darwen; Pascal-boycotters need not fear. The point I was referring to is in #3 in the list of SQL mistakes in Darwen's second reply; it may not be all that significant in our current discussion, though.

-- 
Jon
Received on Fri Mar 03 2006 - 09:15:09 CET

Original text of this message