db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jayaram Subramanian <rsjay1...@gmail.com>
Subject Re: [jira] [Commented] (DERBY-5276) Reference manual has wrong length for BIGINT columns in system tables
Date Thu, 21 Jul 2011 12:47:38 GMT
Hi All,
The dita configuration link in derby mention about updating config file in
subversion with dita file... But i am not able to find the config file for
subversion in my local pc... could you help me in identifying the locatio of
the file.

With Regards
jayaram

On Thu, Jul 7, 2011 at 11:27 AM, Rick Hillegas (JIRA) <jira@apache.org>wrote:

>
>    [
> https://issues.apache.org/jira/browse/DERBY-5276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13061416#comment-13061416]
>
> Rick Hillegas commented on DERBY-5276:
> --------------------------------------
>
> I vote for consistency: bump all of the BIGINT lengths to 20. Certainly the
> SYSSEQUENCES columns can hold negative BIGINTs. For numeric columns in the
> catalogs, length seems to refer to the max number of characters needed to
> display the column value.
>
> > Reference manual has wrong length for BIGINT columns in system tables
> > ---------------------------------------------------------------------
> >
> >                 Key: DERBY-5276
> >                 URL: https://issues.apache.org/jira/browse/DERBY-5276
> >             Project: Derby
> >          Issue Type: Bug
> >          Components: Documentation
> >    Affects Versions: 10.9.0.0
> >            Reporter: Knut Anders Hatlen
> >            Priority: Minor
> >
> > The description of the SYSCOLUMNS table at
> http://db.apache.org/derby/docs/dev/ref/rrefsistabs22441.html says that
> the BIGINT columns AUTOINCREMENTVALUE, AUTOINCREMENTSTART and
> AUTOINCREMENTINC have length 19. The maximum length of a BIGINT is 20 (when
> including the sign for negative values). Here's an example that shows that
> these columns can hold values whose length is 20:
> > ij> create table t(x bigint not null generated always as identity (start
> with -1234567890123456789, increment by -1234567890123456789));
> > 0 rows inserted/updated/deleted
> > ij> select autoincrementvalue, autoincrementstart, autoincrementinc from
> sys.syscolumns where columnname='X';
> > AUTOINCREMENTVALUE  |AUTOINCREMENTSTART  |AUTOINCREMENTINC
> > --------------------------------------------------------------
> > -1234567890123456789|-1234567890123456789|-1234567890123456789
> > 1 row selected
> > Other places where BIGINT columns are listed with length 19:
> > SYSCONGLOMERATES -
> http://db.apache.org/derby/docs/dev/ref/rrefsistabs39391.html
> > SYSFILES - http://db.apache.org/derby/docs/dev/ref/rrefsistabs40972.html
> > SYSSEQUENCES -
> http://db.apache.org/derby/docs/dev/ref/rrefsistabssyssequences.html
> > SYSXPLAIN_RESULTSET_TIMINGS -
> http://db.apache.org/derby/docs/dev/ref/rrefsysxplain_resultset_timings.html
> > SYSXPLAIN_STATEMENT_TIMINGS -
> http://db.apache.org/derby/docs/dev/ref/rrefsysxplain_statement_timings.html
> > SYSCS_DIAG.SPACE_TABLE -
> http://db.apache.org/derby/docs/dev/ref/rrefsyscsdiagtables.html
> > I haven't verified if negative values are allowed in all of these
> columns. If they are only allowed to hold non-negative values, should we
> keep the value 19? Or should we change all to 20 for consistency? I'm
> assuming that "length" in these tables refers to the display size of the
> columns.
>
> --
> This message is automatically generated by JIRA.
> For more information on JIRA, see: http://www.atlassian.com/software/jira
>
>
>

Mime
View raw message