During today’s session, many of you were sufficiently interested in the “versioning of database components” trick that I offered to post the deck from last year’s session here on the blog.
Well, I’ve looked at the deck and it’s frankly a little thin. What I need to do is take the scripts from that demo (which were run on a copy of a fairly large development database) and work up a compact demonstration that I can upload here along with the deck, so that you get the full picture of the techniques without being distracted by other issues.
That’s something that will take me a little bit of time to accomplish. I hope to have the demo and the deck uploaded here for you sometime next week.
Please accept my apologies for the delay, along with my thanks for your patience and your enthusiastic participation in today’s session.
You did a great presentation!
I have one question though about the versioning of database components. If you make separate tabels for the different version how do you handle inserts and updates?
If UserA uses tables from V1 and UserB uses tables from V2 how do you accomp+lish updates in both tables? I guess you can write the update (insert, delete) procedures to take care of that, but them the procedures seems to be very complicated.
It's an interesting idea, and your presentation made me think differently on some projects I'm in.
Thank you very much for your question and your kind words, Marcus. When you say that these ideas have you rethinking your projects, you say the most fulfilling words that this particular presenter can ever here.
The short answer to your question is that UserA and UserB pull different versions of data from the same table. We accomplish this through some set-based sleight-of-hand in the schema, implemented along with the parlor trick I showed in Barcelona this year.
The long answer to your question will have to wait for the post I'm preparing.
This topic was a presentation all its own in 2007 in Barcelona. Based on the response it got in both "SQL Tricks" sessions this year, I'm hoping it will surface on the agenda in Berlin next year.
Thanks again for your interest, Marcus. I'll get a proper discussion of this posted just as soon as I can.
Fulfilling my promise at TechEd , I’ve finally completed a self-directed demonstration of the database