Re: Extending my question. Was: The relational model and relational algebra - why did SQL become the industry standard?
Date: Tue, 04 Mar 2003 08:32:03 +0200
Jan Hidders wrote:
>>OK, point being that if SQL had not had duplicates from the start the
>>motivation for such research would have been much stronger, and we would
>>have got results sooner.
>Could be, but note that this would then be because you made things *harder*
>for the implementors and forced them to think about these issues.
Exactly, and as a result we would have better products. I would also affect how people thought and
used products. We would probably utilize views much more than now. Updatable views, anybody?
>>>So it could very well that for this particular problem you are right.
>>Meaning that "you can do more optimisations with set's"?
>Meaning that in this case you might be right that for this particular query
>the bag-based approach makes optimization a bit harder. However, after
>giving it a bit more thought I doubt that even this is true. The problem is
>just as difficult in a set-only as in a bag-based approach. In both cases
>you can optimize a join followed by a project that projects out a certain
>table by using only the index for the "invisible" table.
Or ignoring it completely...
>Ah, wait, I hadn't really seen the appendix in part II. It's really Chris
>Date at his worst. After reading such nonsense it always takes me a day or
>two to take him serious again. For starters he seems to miss the essential
>point that the presented algebra is an internal algebra. When I read his
>remark about "classical relational algebra" I can only conclude that he is
>the one who doesn't know the literature. On all the conferences I have been
>I have heard this phrase used and everybody knew exactly what it meant. But
>no, Chris Date is going to explain to everyone what these words really mean.
>I can only describe that as hubris.
OK, no comments on that.
Lauri Pietarinen Received on Tue Mar 04 2003 - 07:32:03 CET