Re: here's a good one from dizwell on the recent product launch

"Cristian Cudizio" <cristian.cudizio_at_yahoo.it> a écrit dans le message de news: 1185185715.306332.247300_at_k79g2000hse.googlegroups.com... On Jul 20, 5:26 pm, "Michel Cadot" <micadot{at}altern{dot}org> wrote:
> "Mark Townsend" <markbtowns..._at_sbcglobal.net> a écrit dans le message de news: 46A0C746.3000..._at_sbcglobal.net...| Noons wrote:
>
> | > On Jul 19, 12:35 pm, Mark Townsend <markbtowns..._at_sbcglobal.net>| > wrote:
>
> | >>> Nuno gave me bug 5092688.8
> | ...
> | >>
> | >> I don't think this is what Nuno is talking about
> | >
> | > Let me see, Mark: I provide the bug
> | > number to Jonathan and "this is not what
> | > I'm talking about"? Are you even
> | > READING?
> | >
> |
> | Yes.
> |
> | You claimed that "Oracle continues to deliver new releases without
> | fixing first what are absolutely egregious bugs."
> |
> | You also stated that the database (?) "fails consistently for nearly 7
> | years and the "dedicated" folks do preciously NOTHING to fix problems
> | that have seen three major releases without a permanent fix"
> |
> | This concerns me. I asked you for some pointers to such problems so I
> | could investigate them.
> |
> | You told me offline to go pound sand.
> |
> | However, based on a posting on your blog, Jonathan offered 5092688.8 as
> | a potential example. You seemed to indicate that this was a correct example.
> |
> | However when I looked at the specific problem, it has been fixed within
> | 7 days of it being reported, it was fixed before 11g, it was not 7 years
> | old, and it has not been outstanding for three active releases.
> |
> | Nuno - you have our attention. I would love to have concrete examples of
> | what you are reporting in the forum. I can then assure you that any
> | such examples will be actively reviewed at the highest levels in
> | development.
> |
> | But you need to throw me a bone - give me a list of specific SRs or Bugs
> | that you think are indicative. I would love to remove as much as the
> | invective from this conversation as possible to get to a description of
> | a problem that is actionable.
>
> As you talk about a bone, here's one I found 2 days ago (SR6427362.992).
> It is closed as a duplicate of bug 2499608 opened on 06-AUG-2002 and
> never fixed.
>
> The description is the following one:
>
> <quote>
> When NLS_NUMERIC_CHARACTERS is not set to the default '.,'
> we get the following errors:
> SQL> alter session set nls_numeric_characters='!:';
>
> Session altered.
>
> SQL> select to_number('1,234','9,999') from dual;
> select to_number('1,234','9,999') from dual
> *
> ERROR at line 1:
> ORA-01722: invalid number
>
> SQL> select to_number('1,234','9,999.9') from dual;
> TO_NUMBER('1,234','9,999.9')
> ----------------------------
> 1234
>
> 1 row selected.
>
> First question: Why the error is reached in the first case?
> Second question: Then why it is not in the second one?
> </quote>
>
> OK, it's not a terrific bug, easy to workaround, but
> why 5 years without being fixed?
>
> Regards
> Michel

I found this example more an example of bad use of to_number (that should not be permit in my opinion), because i think it is ambiguous. I think that it should be used only with syntax select to_number('1,234','9G999') A FROM DUAL; as said on metalink note about the bug. I say this because i'm writing from italy where the comma is used to separate decimals and so the ambiguity is frequent. So i think it is normal to request at least to correct the problem with new release but to hope Oracle will release a patch only for this is too much.

Bye
 Cristian Cudizio

http://oracledb.wordpress.com
http://cristiancudizio.wordpress.com

But there were 3 major versions since the problem was raised. So it is not a matter of patch.

As for the matter of knowing if it is a bad use or not, this is irrelevant to the fact that this MUST be corrected. In France, we also use comma to separate decimals and dot to separate groups but we get data for USA that use comma to separate groups, this is why we have this statement for these data and not 'G' that we use for ours.

Regards
Michel Cadot Received on Mon Jul 23 2007 - 10:43:08 CDT