Re: How to signal a program that a table has changed?

From article <1992Mar9.190219.1_at_vxcrna.cern.ch>, by roeber_at_vxcrna.cern.ch:
> In article <2260_at_keele.keele.ac.uk>, csa09_at_seq1.keele.ac.uk (P. Singleton) writes:
>> [...]
>> If you polled the database, how would you tell whether a table has changed?  

> All of our data is timestamped. A client would check that the timestamp
> (of the row with the greatest timestamp, if it's a logging table) is
> different than what it remembers.

OK but that requires the application to make provision for the polling.

It would be neat if Oracle kept some per-table sequence no. which it incremented whenever the table was altered. Like file-modification-times on UNIX files. Then we could cheaply check whether _any_ table has changed.

Surely, finding the greatest timestamp in a table involves a search of the whole table? Or do you index on the timestamp?

---
  __   __    Paul Singleton (Mr)           JANET: paul_at_uk.ac.keele.cs
 |__) (__    Computer Science Dept.        other: paul_at_cs.keele.ac.uk
 |  .  __).  Keele University, Newcastle,    tel: +44 (0)782 621111 x7355
             Staffs ST5 5BG, ENGLAND         fax: +44 (0)782 713082