RE: Counterquestion - is Oracle a He or a She - or an It or a hermaphrodite
- From: "Lex de Haan" <lex.de.haan@xxxxxxxxxxxxxx>
- To: VerreyB@xxxxxxxxxxxx
- Date: Mon, 1 Aug 2005 13:15:54 +0200 (CEST)
OK, there we go :-)
there is no such thing as a NULL numeric, or a NULL string, or a NULL date
-- and also, NULL is *not* the same as empty. NULL is a marker to indicate
the absence of an attribute value. So NULL represents "missing". that's
why the two-valued logic is not good enough anymore; we need three-valued
logic.
if you look at the substr function, it returns a string, right? by
decrementing one of the arguments, the result gets shorter and shorter --
until the last character is taken away. the result is *still* a string,
and as such a value, in my opinion.
or, to give you another example, what should be the result of instr(s,s)
where s represents a string? 1 (one), right? I would argue that this
should also be the case if s is the empty string.
By the way, the ISO/ANSI standard thinks along the same lines, and as such
Oracle is *not* compliant with the ANSI/ISO SQL standard in handling empty
strings.
cheers,
Lex.
> Lex de Haan [mailto:lex.de.haan@xxxxxxxxxxxxxx] wrote:
>
>> well, I don't want to start a religious thread here, but I think it
> is a
>> problem that Oracle treats empty strings as nulls. an empty string
> is a
>> string, allbeit a short one (just like an empty set is a set, though
> a
>> small one) and thus has a datatype. a null is a marker, not a value,
> and
>> has no datatype. therefore, there is (or should be) an important
>> difference between empty strings and nulls.
>
> Lex, then why is a NULL numeric not confusing then?
>
> Can you get an "empty" numeric variable? No! It is a null variable.
>
> Why treat a string variable differently then?
>
> A numeric variable either has a value (-1, 0 or billions). Or not.
> When it does not have a value it is considered a NULL.
>
> A string variable is no different. There is no such thing as an
> "empty" string. Just as there is not such thing as an "empty" numeric,
> or an "empty" date, or an "empty" bool, or an "empty" pointer, or an
> "empty" <insert your favouriote datatype here>. It is either NULL or
> it has a value. AS SIMPLE AS THAT!!
>
> So why treat string variables differently than all other data types!?
> Magically we now have "empty" strings despite the fact that an "empty"
> variable in ALL other data types are considered null. Does not
> compute.
>
> Nothing religious about it. Unless common sense and logic are
> considered a religion?
>
> --
> Billy
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> This e-mail and its contents are subject to the Telkom SA Limited
> e-mail legal notice available at
> http://www.telkom.co.za/TelkomEMailLegalNotice.PDF
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> --
> //www.freelists.org/webpage/oracle-l
>
--
//www.freelists.org/webpage/oracle-l
Other related posts:
