Re: Building the virtual OTLT

"Christopher Browne" <cbbrowne_at_acm.org> wrote in message news:2qbjj0Fte7usU1_at_uni-berlin.de...
> The world rejoiced as "Laconic2" <laconic2_at_comcast.net> wrote:
> > So the virtual OTLT can be a useful tool, even though it illustrates
bad
> > data management.
>
> Well, this strikes me as being the fashion in which "OTLT" can be
> considered to 'turn out well.'

I guess it should be the "OTLV"
>
> What's interesting is that you _didn't_ fudge things up by creating a
> way to insert bad data. Within the realm of "updatable views," this
> isn't likely to be able to be anywhere _near_ "automatically
> updatable," particularly if many of the tables it references contain
> additional fields (e.g. - updated_on or created_on or such).

In this particular case, I could not have screwed it up, even if I had wanted to. My access to BANNER was read only.

That's all I needed to fulfill my contract.

But regardless of what privileges I might have had, I would not have foisted the OTLT off on an unsuspecting public.

>
> There's a place for text search, and this is it.

There's a place for skunk works, too. And this is it.

In my more theoretical moments, I would note that the OTLV joins metadata (table and column names) with data (code values and descriptions). This, by itself, is at best unaesthetic and at worst irresponsible.

But it works!

(Your signature reminds me that, in about 1970, we would have called this "a neat hack".)

> --
> output = reverse("gro.gultn" "_at_" "enworbbc")
> http://www3.sympatico.ca/cbbrowne/nonrdbms.html
> "A hack is a terrible thing to waste, please give to the
> implementation of your choice..." -- GJC
Received on Thu Sep 09 2004 - 22:00:53 CEST