Comments on: To RR or to YY ? http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html Oracle Certified Master Sun, 07 Sep 2008 06:09:28 +0000 http://wordpress.org/?v=2.0.5 by: Laurent Schneider http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6181 Sat, 12 Apr 2008 05:52:24 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6181 yes, right : <a href="http://download.oracle.com/docs/cd/B19306_01/server.102/b14220/datatype.htm#sthref3818" rel="nofollow">Oracle® Database Concepts 10g Release 2 (10.2)</a> <i>Zero and positive and negative infinity (only generated on import from Version 5 Oracle databases) are stored using unique representations. Zero and negative infinity each require 1 byte; positive infinity requires 2 bytes.</i> yes, right :
Oracle® Database Concepts 10g Release 2 (10.2)
Zero and positive and negative infinity (only generated on import from Version 5 Oracle databases) are stored using unique representations. Zero and negative infinity each require 1 byte; positive infinity requires 2 bytes.

]]> by: David Aldridge http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6179 Fri, 11 Apr 2008 20:31:59 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6179 Wasn't there plus and minus infinity in version up to up to Oracle v5? I think I rememeber _reading_ about that. (I'm not that old). Wasn’t there plus and minus infinity in version up to up to Oracle v5? I think I rememeber _reading_ about that. (I’m not that old).

]]> by: Log Buffer #92: a Carnival of the Vanities for DBAs http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6177 Fri, 11 Apr 2008 16:39:41 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6177 [...] Laurent Schneider is looking at the right-or-wrongness of date-format. He writes, “What is worst? To use DD-MON-RR or to use DD-MON-YY?”. Probably neither, as the comments point out. Chen Shapira says, “FXYYYY rules indeed.” [...] […] Laurent Schneider is looking at the right-or-wrongness of date-format. He writes, “What is worst? To use DD-MON-RR or to use DD-MON-YY?”. Probably neither, as the comments point out. Chen Shapira says, “FXYYYY rules indeed.” […]

]]> by: Laurent Schneider http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6170 Thu, 10 Apr 2008 08:48:49 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6170 dear, it is not inf, it is nan :D <code>select cast(utl_raw.cast_to_number('7E') as binary_double) from dual; CAST(UTL_RAW.CAST_TO_NUMBER('7E')ASBINARY_DOUBLE) ------------------------------------------------- NAN </code> as NaN does not exist as a Number, the only purpose to have such a number is annoying your colleagues. Of course it is not supported. And operations involving such an unsupported thing are, well, surprising :twisted: <code>select cast((-utl_raw.cast_to_number('7E')) as binary_double) from dual; CAST((-UTL_RAW.CAST_TO_NUMBER('7E'))ASBINARY_DOUBLE) ---------------------------------------------------- Inf </code> dear, it is not inf, it is nan :D 

select  cast(utl_raw.cast_to_number('7E') as binary_double) from dual;

CAST(UTL_RAW.CAST_TO_NUMBER('7E')ASBINARY_DOUBLE)
-------------------------------------------------
                                              NAN

as NaN does not exist as a Number, the only purpose to have such a number is annoying your colleagues. Of course it is not supported. And operations involving such an unsupported thing are, well, surprising :twisted: 

select cast((-utl_raw.cast_to_number('7E')) as binary_double) from dual;

CAST((-UTL_RAW.CAST_TO_NUMBER('7E'))ASBINARY_DOUBLE)
----------------------------------------------------
                                                 Inf

]]> by: Chen Shapira http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6169 Thu, 10 Apr 2008 03:40:12 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6169 Hey Laurent, Forget about Gary's users. I want an explanation for 7E as number. It displays as -inf in sqlplus, but it clearly doesn't behave as -inf: SQL> select UTL_RAW.CAST_TO_NUMBER('7E') from dual; UTL_RAW.CAST_TO_NUMBER('7E') ---------------------------- -~ SQL> select UTL_RAW.CAST_TO_NUMBER('7E')+1 from dual; UTL_RAW.CAST_TO_NUMBER('7E')+1 ------------------------------ 1 Other single byte numbers also display the same: SQL> select UTL_RAW.CAST_TO_NUMBER('7F') from dual; UTL_RAW.CAST_TO_NUMBER('7F') ---------------------------- -~ We can also get inf: SQL> select UTL_RAW.CAST_TO_NUMBER('81') from dual UTL_RAW.CAST_TO_NUMBER('81') ---------------------------- ~ And of course, 0 is between :-) (but this is trivial) SQL> select UTL_RAW.CAST_TO_NUMBER('80') from dual UTL_RAW.CAST_TO_NUMBER('80') ---------------------------- 0 Steve Adams wrote some stuff about infinite numbers: http://www.ixora.com.au/notes/infinity.htm but I suspect that this is unrelated. Please explain! Hey Laurent,

Forget about Gary’s users. I want an explanation for 7E as number.

It displays as -inf in sqlplus, but it clearly doesn’t behave as -inf:

SQL> select UTL_RAW.CAST_TO_NUMBER(’7E’) from dual;

UTL_RAW.CAST_TO_NUMBER(’7E’)
—————————-
-~

SQL> select UTL_RAW.CAST_TO_NUMBER(’7E’)+1 from dual;

UTL_RAW.CAST_TO_NUMBER(’7E’)+1
——————————
1

Other single byte numbers also display the same:

SQL> select UTL_RAW.CAST_TO_NUMBER(’7F’) from dual;

UTL_RAW.CAST_TO_NUMBER(’7F’)
—————————-
-~

We can also get inf:

SQL> select UTL_RAW.CAST_TO_NUMBER(’81′) from dual

UTL_RAW.CAST_TO_NUMBER(’81′)
—————————-
~

And of course, 0 is between :-) (but this is trivial)

SQL> select UTL_RAW.CAST_TO_NUMBER(’80′) from dual

UTL_RAW.CAST_TO_NUMBER(’80′)
—————————-
0

Steve Adams wrote some stuff about infinite numbers:
http://www.ixora.com.au/notes/infinity.htm
but I suspect that this is unrelated.

Please explain!

]]> by: Laurent Schneider http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6164 Wed, 09 Apr 2008 11:33:10 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6164 Gary, your users will love to have in UTL_RAW.CAST_TO_NUMBER('7E') in the number columns :) Gary, your users will love to have in UTL_RAW.CAST_TO_NUMBER(’7E’) in the number columns :)

]]> by: Laurent Schneider http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6163 Wed, 09 Apr 2008 09:50:20 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6163 dates in BC, sounds fun ! well, truncate may reveal some inconsistencies ... <code>select column_value,trunc(column_value,'CC') from table(sys.odcidatelist(date '2008-04-09',date '-2008-04-09')) ; COLUMN_VA TRUNC(COL --------- --------- 09-APR-08 01-JAN-01 09-APR-08 01-JAN-00 </code> :) dates in BC, sounds fun !

well, truncate may reveal some inconsistencies …

select column_value,trunc(comµçVÖå÷fÇVRÂt42r’g&öÒF&ÆR‡7—2æöF6–FFVÆ—7B†FFRs#‚ÓBÓ’rÆFFRrÓ#‚ÓBÓ’r’’°Ð¤4ôÅTÔåõdE%Tä2„4ôÀТÒÒÒÒÒÒÒÒÒÒÒÒÒÒÒÒÒÐУ’Ô"Ó‚Ô¤âÓУ’Ô"Ó‚Ô¤âÓРÐ

:)

]]> by: Chen Shapira http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6161 Wed, 09 Apr 2008 03:41:06 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6161 FXYYYY rules indeed. Thanks for the insight about limitations of YYYY. FXYYYY rules indeed.
Thanks for the insight about limitations of YYYY.

]]> by: Gary http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6160 Wed, 09 Apr 2008 01:03:58 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6160 select date '08-12-31' from dual; gives the year 0008, so FXYYYY still wins out. If you really want to confuse co-workers, switch a bunch of dates from AD to BC. They'll never work it out. Why Oracle (or SQL) thought it would ever be necessary to store an exact date (and time) two thousand plus years ago is stunning. Who thought that would ever be useful ? Please, give me a setting so that database-wide, Oracle would reject dates outside sensible boundaries. Default it to 1-jan-1800 to 31-dec-2199 and 99% of the date problems would disappear. select date ‘08-12-31′ from dual; gives the year 0008, so FXYYYY still wins out.
If you really want to confuse co-workers, switch a bunch of dates from AD to BC. They’ll never work it out. Why Oracle (or SQL) thought it would ever be necessary to store an exact date (and time) two thousand plus years ago is stunning. Who thought that would ever be useful ?
Please, give me a setting so that database-wide, Oracle would reject dates outside sensible boundaries. Default it to 1-jan-1800 to 31-dec-2199 and 99% of the date problems would disappear.

]]> by: Laurent Schneider http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6157 Tue, 08 Apr 2008 12:50:03 +0000 http://laurentschneider.com/wordpress/2008/04/to-rr-or-to-yy.html#comment-6157 yes, YYYY-MM-DD is pretty, it even sorts right as string :mrgreen: yes, YYYY-MM-DD is pretty, it even sorts right as string :mrgreen:

]]>