RE: Separate cursor for bind value assignment, is this normal
Date: Mon, 8 Oct 2018 17:26:15 -0400
Message-ID: <028b01d45f4d$8b8a3c80$a29eb580$_at_rsiz.com>
hehe, which started as the ufi (user friendly interface), which some customers called the uvi (user viscious interface) for things like its lack of a history buffer and so forth.
You are correct that it was never intended as a programming tool. Oracle had Pro*C for amateurs and the C call libraries for professionals. And at 300 baud, you had better figure out a way to avoid telecom line turn-arounds in the middle of a transaction, too, and with 25 MHz cpus and table locking you better avoid parses in the middle of transactions, too.
UFI was competitive with its competition as a quick way to ask the database quick questions.
mwf
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Dragutin Jastrebic
Sent: Monday, October 08, 2018 2:18 PM
To: contact_at_soocs.de
Cc: krishsingh.111_at_gmail.com; Oracle - L
Subject: Re: Separate cursor for bind value assignment, is this normal
Hello
In this old thread=>
https://asktom.oracle.com/pls/asktom/f?p=100:11:0::::p11_question_id:113012348062
Tom Kyte explained
==>
.... in sqlplus the exec :x := &x will be parsed each time,
those will be unavoidable (sqlplus is not a programming environment, it is just a simple command line tool)
Yes, I wish sqlplus had a way to initialize a bind variable -- it would be nice.
<===
So possibly that it was event a design choice of Oracle development team, since sqlplus was more seen a command line tool, not a programming tool.
Dragutin
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Oct 08 2018 - 23:26:15 CEST