home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
tsql
/
doc
/
tsql.mail
/
000093_@ICINECA.CINECA…s64.cineca.it _Tue Apr 27 12:05:17 1993.msg
< prev
next >
Wrap
Internet Message Format
|
1996-01-31
|
1KB
Received: from icineca.cineca.it by optima.CS.Arizona.EDU (5.65c/15) via SMTP
id AA22605; Tue, 27 Apr 1993 03:03:35 MST
Received: from deis64.cineca.it by ICINECA.CINECA.IT (IBM VM SMTP V2R2)
with TCP; Tue, 27 Apr 93 12:03:31 SET
Received: from [137.204.57.79] (deis79) by deis64.cineca.it (4.1/SMI-4.1)
id AA13128; Tue, 27 Apr 93 12:05:43 +0200
Date: Tue, 27 Apr 93 12:05:17 +0100
From: (Fabio Grandi) <fabio@deis64.cineca.it>
Message-Id: <43524.fabio@deis64.cineca.it>
To: tsql@cs.arizona.edu
Subject: TSQL benchmark. Re: Key time-invariance
As to the problem of Key time-invariance
----------------------------------------
We agree with Jim Clifford on the importance
of the "grouping" issue. Our position can be
summarized as follows:
1) the benchmark model SHOULD support the grouping
(e.g. via surrogates) and, thus, keys are
not required to be time-invariant;
-- or, alternatively --
2) if the model does not support the grouping,
the keys SHOULD be time-invariant, so that
the user can explicitly support the grouping
via the key.
However, we prefer the first solution and
strongly disagree with a ungrouped model
together with time-variant keys.
Sincerely,
F. Grandi, M.R. Scalas and P.Tiberio.