tag:blogger.com,1999:blog-7395977411859619892.post7427059684731953460..comments2024-01-30T09:21:06.579+01:00Comments on About Oracle: Implementing entity rulesRob van Wijkhttp://www.blogger.com/profile/00499478359372903250noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-7395977411859619892.post-20640902700105430312008-09-01T22:19:00.000+02:002008-09-01T22:19:00.000+02:00Robert,No, you did not overlook something. You are...Robert,<BR/><BR/>No, you did not overlook something. You are right, the second column is completely unnecessary. I've edited the post accordingly. Thanks!<BR/><BR/>Regards,<BR/>Rob.Rob van Wijkhttps://www.blogger.com/profile/00499478359372903250noreply@blogger.comtag:blogger.com,1999:blog-7395977411859619892.post-64464478383059404432008-09-01T15:24:00.000+02:002008-09-01T15:24:00.000+02:00Rob, why do you do a two column index?rwijk@ORA11G...Rob, why do you do a two column index?<BR/><BR/>rwijk@ORA11G> create unique index non_processed_orders_uk on orders<BR/>2 ( decode(processed_indicator,'N',customer_id)<BR/>3 , decode(processed_indicator,'N','N')<BR/>4 )<BR/>5 /<BR/><BR/>One column should be sufficient<BR/><BR/>rwijk@ORA11G> create unique index non_processed_orders_uk on orders<BR/>2 ( decode(roberthttps://www.blogger.com/profile/01724179181550310220noreply@blogger.comtag:blogger.com,1999:blog-7395977411859619892.post-78255094561193670622008-08-13T22:28:00.000+02:002008-08-13T22:28:00.000+02:00Brian,Yes I was already aware of this, thanks to y...Brian,<BR/><BR/>Yes I was already aware of this, thanks to your comments on Lucas' AMIS-post. And indeed I decided not to bring it up because it is so rare.<BR/><BR/>However, I did not read the entire AskTom thread before, and I also wasn't aware of how a serializable transaction can be detected. So I just did some catching up.<BR/><BR/>After all the reading, my first conclusion is that we Rob van Wijkhttps://www.blogger.com/profile/00499478359372903250noreply@blogger.comtag:blogger.com,1999:blog-7395977411859619892.post-1559488594213465682008-08-13T21:05:00.000+02:002008-08-13T21:05:00.000+02:00Bruno,Thanks for your comment.I agree that English...Bruno,<BR/><BR/>Thanks for your comment.<BR/><BR/>I agree that English error messages and feedback messages would be more appropriate for this blog. However, I prefer Dutch messages for everything else I do on my databases.<BR/><BR/>I have just experimented a bit, and it seems that if I change my NLS_LANG registry setting prior to opening my database, this results in English messages. Now I only Rob van Wijkhttps://www.blogger.com/profile/00499478359372903250noreply@blogger.comtag:blogger.com,1999:blog-7395977411859619892.post-87384302379810478952008-08-12T19:21:00.000+02:002008-08-12T19:21:00.000+02:00I had a comment about the "other entity rules"...I...I had a comment about the "other entity rules"...<BR/><BR/>It would not surprise me if you were aware of this already and chose not to bring it up for the sake of clarity, but...<BR/><BR/>DBMS_LOCK will not protect <B>serializable</B> transactions against updates or inserts by other transactions that committed <B>after</B> the serializable transaction started. This is an issue for any approach Brian Camirehttps://www.blogger.com/profile/08041233250752852225noreply@blogger.comtag:blogger.com,1999:blog-7395977411859619892.post-67265341159627442362008-08-12T04:16:00.000+02:002008-08-12T04:16:00.000+02:00I discovered your blog via The Database Programmin...I discovered your blog via The Database Programming blog, and follow it ever since. Thanks for your carefully written posts. My only complain is that Oracle's messages are in Dutch (I suppose). Can it be configured to output in English?br1https://www.blogger.com/profile/16868211253715877783noreply@blogger.com