As it it was not enough that I had to read about the fun (technical and otherwise) of OOW in San Francisco, a mere 1,300 mile journey that I could probably cover in slightly more than a day with the wind in the right direction, but Peter Scott also sees fit to remind me that we’re teetering on the brink of UKOUG 2006.
Alas, despite appropriate budgeting and extensive plans for a little vacationing around “the old country” to round the experience off, I shall not be attending — SWMBO is in Sicily performing services presumably vital to the national well-being of the nation, and has no fixed return date. Maybe she’ll be back by Thanksgiving on November 23rd but until the ticket is in her hand there’s really no knowing.
I could of course take the three boys with me to UKOUG but I doubt they’d go unnoticed during a one hour technical presentation on an Oracle subject-of-interest. If people thought that cell phone usage at OOW was a rude inconvenience then this would really recalibrate their interruptionometer. They would certainly be more than a little underfoot during the off-duty hours as well.
Tsh, can’t be helped. I have RMOUG in February — I’ll be presenting a quick half-hour on Linux 2.6 I/O Scheduling, and my 90 minutes on materialized views is on the “reserve” list in case a slot becomes available.
March will bring the Hotsos Symposium, which reminds me that I must finalise my bio and get my abstract submitted. I had a minor epiphany the other day when I realised that my discomfort with the term “summary tables” (which took over from the term “materialized views”) can be resolved by thinking more in terms of “precalculated result sets”. This was brought about by considering the case of using a materialized view or other data set to improve performance on a year-to-date type query (which I wrote about a little while ago) or on complex financial reports that use other forms of non-equijoin, by allowing them to be resolved to a more efficient (hash join compatible) equijoin. Whether it is a summary table or a materialization of a non-equijoin that provides the performance improvement the common element is the precalculation, storage and re-use of a final or intermediate result set. Does that make sense? Maybe you’ll have to go to Hotsos to find out. Anyway it’s good to get those minor nagging doubts out of the way.
There is one teensy-weensy fly in the ointment — at the moment we face the possibility of having to sell the house and move from Colorado Springs to somewhere in the Arlington area, a trot of around 1,600 miles, so that SWMBO can report no later than the end of February to a new job at the Air Force Office of Scientific Research. That has the period from ten days after RMOUG to a couple of days prior to Hotsos neatly blocked off. On the other hand, we may end up moving to Monterey in mid-2007 instead. As you might imagine it is a little tense not knowing. Either way I have the other temptation of looking for a new job at whatever destination is chosen for us. I’m thinking “large corporation with international prospects”. Or whatever seems best at the time.
Well enough of that — development seems to be ramping up on my client’s project with a 10gR2 upgrade in the next month or two so there’s plenty to be getting along with. The rest I’ll worry about later on.