Most systems for no holes use an array of available numbers. Once you
commit that number is no longer in the array, otherwise it stays in
and is available for use.
- James Cicenia
On Feb 6, 2006, at 10:18 AM, Helge Staedtler wrote:
oh, I must have overlooked that there was a restriction existing
which said
"without any holes". sorry. I would say, that a sequence without
any holes
can only be provided by a queuing machanism. I think a timestamp with
milliseconds would do this in 99.99% of all cases. just an idea.
have fun playing around with this.
just my 2 cent.
regards,
helge
Am 06.02.2006 16:32 Uhr schrieb "Guido Neitzer" unter
<[EMAIL PROTECTED]>:
On 06.02.2006, at 16:08 Uhr, Helge Staedtler wrote:
hmm, I would have a look at the primary keys.
those are ensured by EOF and the DB to be unique, so any number
which is
derived from the primarykey (you might e.g. transform an
alphanumeric PK
into a pure number key) should meet your criteria.
No. Primary keys are not guaranteed to be in sequence without
"holes". You may use a stored procedure or something like that to
increment a counter - beware of rollbacks.
If there is no requirement for a real sequence without holes, pk may
work. Or, depending on the database, any mechanism that can be used
for integer pk generation.
cug
--
Momo was there and joined in, that was all, but for some reason her
mere
presence put bright ideas into their heads. -- from Michael Ende's
"Momo"
Expect and Respect.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/james%
40jimijon.com
This email sent to [EMAIL PROTECTED]
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to [email protected]
